Wireless Communication Technology

Fundamental Concepts 1
Objectives
After studying this chapter, you should be able to:
( Explain the nature and importance of wireless communication.
( Outline the history of wireless communication.
( Explain the necessity for modulation in a radio communication system.
( Outline the roles of the transmitter, receiver, and channel in a radio
communication system.
( Describe and explain the differences among simplex, half-duplex, and
full-duplex communication systems.
( Describe the need for wireless networks and explain the use of repeaters.
( List and briefly describe the major types of modulation.
( State the relationship between bandwidth and information rate for any
communication system.
( Calculate thermal noise power in a given bandwidth at a given
temperature.
( Explain the concept of signal-to-noise ratio and its importance to
communication systems.
( Describe the radio-frequency spectrum and convert between frequency and
wavelength.
2 ! CHAPTER 1
' 1.1 Introduction

This is a book on wireless communication. That usually means communica-
tion by radio, though ultrasound and infrared light are also used occasion-
ally. The term “wireless” has come to mean nonbroadcast communication,
usually between individuals who very often use portable or mobile equip-
ment. The term is rather vague, of course, and there are certainly borderline
applications that are called wireless without falling exactly into the above
definition.
Wireless communication is the fastest-growing part of the very dynamic
field of electronic communication. It is an area with many jobs that go un-
filled due to a shortage of knowledgeable people. It is the author’s hope that
this book will help to remedy that situation.
' 1.2 Brief History of Wireless Telecommunication

Most of this book is concerned with the present state of wireless communica-
tion, with some speculation as to the future. However, in order to under-
stand the present state of the art, a brief glimpse of the past will be useful.
Present-day systems have evolved from their predecessors, some of which
are still very much with us. Similarly, we can expect that future systems will
be developed from current ones.
The Beginning Wireless telecommunication began only a little later than the wired variety.
Morse’s telegraph (1837) and Bell’s telephone (1876) were soon followed
by Hertz’s first experiments with radio (1887). Hertz’s system was a labora-
tory curiosity, but Marconi communicated across the English Channel in
1899 and across the Atlantic Ocean in 1901. These successes led to the wide-
spread use of radio for ship-to-ship and ship-to-shore communication using
Morse code.
Early wireless systems used crude, though often quite powerful, spark-
gap transmitters, and were suitable only for radiotelegraphy. The invention
of the triode vacuum tube by De Forest in 1906 allowed for the modula-
tion of a continuous-wave signal and made voice transmission practical.
There is some dispute about exactly who did what first, but it appears likely
that Reginald Fessenden made the first public broadcast of voice and music
in late 1906. Commercial radio broadcasting in both the United States and
Canada began in 1920.
Early radio transmitters were too cumbersome to be installed in vehicles.
In fact, the first mobile radio systems, for police departments, were one-way,
FUNDAMENTAL CONCEPTS ! 3
with only a receiver in the police car. The first such system to be considered
practical was installed in Detroit in 1928. Two-way police radio, with the
equipment occupying most of the car trunk, began in the mid-1930s. Ampli-
tude modulation (AM) was used until the late 1930s, when frequency modu-
lation (FM) began to displace it.
World War II provided a major incentive for the development of mobile
and portable radio systems, including two-way systems known as “walkie-
talkies” that could be carried in the field and might be considered the dis-
tant ancestors of today’s cell phones. FM proved its advantages over AM
in the war.
Postwar Soon after the end of World War II, two systems were developed that pres-
Expansion aged modern wireless communication. AT&T introduced its Improved
Mobile Telephone Service (IMTS) in 1946, featuring automatic connection
of mobile subscribers to the public switched telephone network (PSTN).
This was an expensive service with limited capacity, but it did allow true mo-
bile telephone service. This system is still in use in some remote areas,
where, for instance, it allows access to the PSTN from summer cottages.
The next year, in 1947, the American government set up the Citizens’
Band (CB) radio service. Initially it used frequencies near 460 MHz, but in
that respect it was ahead of its time, since equipment for the UHF range was
prohibitively expensive. Frequencies in the 27-MHz band were allocated in
1958, and CB radio immediately became very popular. The service was short-
range, had no connection to the PSTN, and offered users no privacy, but it
was (and still is) cheap and easy to set up. The popularity of CB radio has de-
clined in recent years but it is still useful in applications where its short
range and lack of connectivity to the rest of the world are not disadvantages.
For example, it serves very well to disseminate information about traffic
problems on the highway.
Meanwhile another rather humble-appearing appliance has become
ubiquitous: the cordless phone. Usually intended for very short-range com-
munication within a dwelling and its grounds, the system certainly lacks
range and drama, but it does have connectivity with the PSTN. Most cordless
phones use analog FM in the 46- and 49-MHz bands, but some of the latest
models are digital and operate at either 900 MHz or 2.4 GHz. Cordless
phones are cheap and simple to use, but their range is limited and, except for
the digital models, they offer little privacy.
Pagers were introduced in 1962. The first models merely signaled the
user to find a telephone and call a prearranged number. More recent models
can deliver an alphanumeric message and even carry a reply. Though rela-
tively limited in function, pagers remain very popular due to their low cost
and small size.
4 ! CHAPTER 1
The Cellular The world’s first cellular radio service was installed in Japan in 1979, fol-
Revolution lowed in 1983 by North American services. Cellular systems are quite differ-
ent from previous radiotelephone services such as IMTS in that, instead of
using a single powerful transmitter located on a tall tower for wide coverage,
the power of each transmitter is deliberately kept relatively small so that the
coverage area, called a cell, will also be small. Many small cells are used so
that frequencies can be reused at short distances. Of course, a portable or
mobile telephone may move from one cell to another cell during the course
of a conversation. In fact, this handoff may occur several times during a
conversation. Practical cellular systems had to await the development of
computers fast enough and cheap enough to keep track of all this activity.
Theoretically at least, the number of users in a cellular system can be in-
creased indefinitely, simply by making the cells smaller.
The first cellular systems used analog FM transmission, but digital mod-
ulation schemes, which provide greater privacy and can use bandwidth
more efficiently, are used in all the new systems. These personal communi-
cation systems (PCS) usually operate in a higher frequency range (about
1.9 GHz compared with 800 MHz for North American cellular service).
Current cellular systems are optimized for voice but can also transmit
data. In the near future, high-speed data transmission using PCS is expected
to become a reality. At this point, however, the past merges into the future,
and we’ll resume the discussion later in this book.
' 1.3 Elements of a Wireless Communication System

The most basic possible wireless system consists of a transmitter, a receiver,
and a channel, usually a radio link, as shown in Figure 1.1. Since radio can-
not be used directly with low frequencies such as those in a human voice, it
is necessary to superimpose the information content onto a higher fre-
quency carrier signal at the transmitter, using a process called modulation.
The use of modulation also allows more than one information signal to use
FIGURE 1.1 Elements of a communication system

the radio channel by simply using a different carrier frequency for each. The
inverse process, demodulation, is performed at the receiver in order to re-
cover the original information.
The information signal is also sometimes called the intelligence, the
modulating signal, or the baseband signal. An ideal communication sys-
tem would reproduce the information signal exactly at the receiver, except
for the inevitable time delay as it travels between transmitter and receiver,
and except, possibly, for a change in amplitude. Any other changes consti-
tute distortion. Any real system will have some distortion, of course: part of
the design process is to decide how much distortion, and of what types, is
acceptable.
Simplex Figure 1.1 represents a simplex communication system. The communica-

and Duplex tion is one way only, from transmitter to receiver. Broadcasting systems are
Communication like this, except that there are many receivers for each transmitter.
Most of the systems we discuss in this book involve two-way communi-
cation. Sometimes communication can take place in both directions at once.
This is called full-duplex communication. An ordinary telephone call is an
example of full-duplex communication. It is quite possible (though perhaps
not desirable) for both parties to talk at once, with each hearing the other.
Figure 1.2 shows full-duplex communication. Note that it simply doubles
the previous figure: we need two transmitters, two receivers, and, usually,
two channels.
FIGURE 1.2 Full-duplex communication system

6 ! CHAPTER 1
Some two-way communication systems do not require simultaneous

communication in both directions. An example of this half-duplex type of
communication is a conversation over citizens’ band (CB) radio. The opera-
tor pushes a button to talk and releases it to listen. It is not possible to talk
and listen at the same time, as the receiver is disabled while the transmitter
is activated. Half-duplex systems save bandwidth by allowing the same
channel to be used for communication in both directions. They can some-
times save money as well by allowing some circuit components in the trans-
ceiver to be used for both transmitting and receiving. They do sacrifice some
of the naturalness of full-duplex communication, however. Figure 1.3 shows
a half-duplex communication system.
FIGURE 1.3 Half-duplex communication system
Wireless The full- and half-duplex communication systems shown so far involve
Networks communication between only two users. Again, CB radio is a good example
of this. When there are more than two simultaneous users, or when the two
users are too far from each other for direct communication, some kind of
network is required. Networks can take many forms, and several will be ex-
amined in this book. Probably the most common basic structure in wireless
communication is the classic star network, shown in Figure 1.4.
The central hub in a radio network is likely to be a repeater, which con-
sists of a transmitter and receiver, with their associated antennas, located in
FIGURE 1.4
Star network
a good position from which to relay transmissions from and to mobile radio
equipment. The repeater may also be connected to wired telephone or data
networks. The cellular and PCS telephone systems that we look at later in the
book have an elaborate network of repeater stations.
' 1.4 Signals and Noise

The communication systems described in this book differ in many ways, but
they all have two things in common. In every case we have a signal, which is
used to carry useful information; and in every case there is noise, which en-
ters the system from a variety of sources and degrades the signal, reducing
the quality of the communication. Keeping the ratio between signal and
noise sufficiently high is the basis for a great deal of the work that goes into
the design of a communication system. This signal-to-noise ratio, abbrevi-
ated S/N and almost always expressed in decibels, is an important specifica-
tion of virtually all communication systems. Let us first consider signal and
noise separately, and then take a preliminary look at S/N.
Modulated Given the necessity for modulating a higher-frequency signal with a

Signals lower-frequency baseband signal, it is useful to look at the equation for
a sine-wave carrier and consider what aspects of the signal can be varied. A
general equation for a sine wave is:
e(t) = E c sin(ω c t + θ) (1.1)

8 ! CHAPTER 1
where
e(t) = instantaneous voltage as a function of time
E c = peak voltage of the carrier wave
ω c = carrier frequency in radians per second
t = time in seconds
θ = phase angle in radians
It is common to use radians and radians per second, rather than degrees
and hertz, in the equations dealing with modulation, because it makes the
mathematics simpler. Of course, practical equipment uses hertz for fre-
quency indications. The conversion is easy. Just remember from basic ac the-
ory that
ω = 2πƒ (1.2)
where
ω = frequency in radians per second
ƒ = frequency in hertz
A look at Equation (1.1) shows us that there are only three parameters of
a sine wave that can be varied: the amplitude E c , the frequency ω, and the
phase angle θ. It is also possible to change more than one of these parameters
simultaneously; for example, in digital communication it is common to vary
both the amplitude and the phase of the signal.
Once we decide to vary, or modulate, a sine wave, it becomes a com-
plex waveform. This means that the signal will exist at more than one
frequency; that is, it will occupy bandwidth. Bandwidth is a concept that
will be explored in more detail later in this chapter and will recur often in
this book.
Noise It is not sufficient to transmit a signal from transmitter to receiver if the

noise that accompanies it is strong enough to prevent it from being under-
stood. All electronic systems are affected by noise, which has many sources.
In most of the systems discussed in this book, the most important noise
component is thermal noise, which is created by the random motion of mol-
ecules that occurs in all materials at any temperature above absolute zero
(0 K or −273° C). We shall have a good deal to say about noise and the ratio
between signal and noise power (S/N) in later chapters. For now let us just
note that thermal noise power is proportional to the bandwidth over which
a system operates. The equation is very simple:
P N = kTB (1.3)
where
P N = noise power in watts
k = Boltzmann’s constant, 1.38 × 10 −23 joules/kelvin (J/K)
T = temperature in kelvins
B = noise power bandwidth in hertz
Note the recurrence of the term bandwidth. Here it refers to the range of fre-
quencies over which the noise is observed. If we had a system with infinite
bandwidth, theoretically the noise power would be infinite. Of course, real
systems never have infinite bandwidth.
A couple of other notes are in order. First, kelvins are equal to degrees
Celsius in size; only the zero point on the scale is different. Therefore, con-
verting between degrees Celsius and kelvins is easy:
T(K) = T(°C) + 273 (1.4)

where
T(K) = absolute temperature in kelvins
T(°C) = temperature in degrees Celsius
Also, the official terminology is “degrees Celsius” or °C but just “kelvins”
or K.
EXAMPLE 1.1 Y
A resistor at a temperature of 25 °C is connected across the input of an ampli-
fier with a bandwidth of 50 kHz. How much noise does the resistor supply to
the input of the amplifier?
SOLUTION
First we have to convert the temperature to kelvins.
From Equation (1.4),
T(K) = T(°C) + 273

= 25 + 273
= 298 K
Now substitute into Equation (1.3),
PN = kTB
= 1.38 × 10−23 × 298 × 50 × 103
= 2.06 × 10−16 W
= 0.206 fW
10 ! CHAPTER 1
This is not a lot of power to be sure, but received signal levels in radio
communication systems are also very small.
X
Signal-to-Noise Maintaining an adequate ratio of signal power to noise power is essential for
Ratio any communication system, though the exact definition of “adequate” var-
ies greatly. Obviously there are two basic ways to improve S/N: increase the
signal power or reduce the noise power. Increasing signal power beyond a
certain point can cause problems, particularly where portable, battery-
powered devices are concerned. Reducing noise power requires limiting
bandwidth and, if possible, reducing the noise temperature of a system. The
system bandwidth must be large enough to accommodate the signal band-
width, but should be no larger than that. Some modulation schemes are
more efficient than others at transmitting information with a given power
and bandwidth.
Noise Figure The noise temperature of a complex system is not necessarily equal to the
and Noise actual operating temperature, but may be higher or lower. The noise temper-
Temperature ature for electronic systems is often found by way of the noise figure, so let
us look briefly at that specification.
Noise figure describes the way in which a device adds noise to a signal
and thereby degrades the signal-to-noise ratio. It is defined as follows:
( S/N )i
NF = (1.5)
( S/N )o
where
(S/N) i = signal-to-noise ratio at the input
(S/N) o = signal-to-noise ratio at the output
All of the above are expressed as power ratios, not in decibels. When a
device has multiple stages, each stage contributes noise, but the first stage is
the most important because noise inserted there is amplified by all other
stages. The equation that expresses this is:
NF2 − 1 NF3 − 1
NFT = NF1 + + +⋅⋅⋅ (1.6)
A1 A1A2
where
NF T = total noise figure for the system
NF 1 = noise figure of the first stage
NF 2 = noise figure of the second stage

A 1 = gain of the first stage
A 2 = gain of the second stage
Again, all these are ratios, not in decibels. The noise figure for the system is
usually specified in dB in the usual way:
NF(dB) = 10 log NF (1.7)

Converting noise figure to noise temperature is quite easy:
T eq = 290(NF − 1) (1.8)
where
T eq = equivalent noise temperature in kelvins
NF = noise figure as a ratio (not in dB)
The noise temperature due to the equipment must be added to the noise
temperature contributed by the antenna and its transmission line to find the
total system noise temperature. We’ll see how that is done after we have
looked at receivers, antennas, and transmission lines separately.
EXAMPLE 1.2 Y
A three-stage amplifier has stages with the following specifications. Gain
and noise figure are given as ratios.
Stage Power Gain Noise Figure
1 10 2
2 25 4
3 30 5
Calculate the power gain in decibels, noise figure in decibels, and equivalent
noise temperature for the whole amplifier.
SOLUTION
The power gain is the product of the individual gains:
AT = A1A2A3 = 10 × 25 × 30 = 7500 = 38.8 dB

12 ! CHAPTER 1
The noise figure is found from
NF2 − 1 NF3 − 1
NFT = NF1 + + +⋅⋅⋅
A1 A1A2
4 −1 5−1
= 2+ +
10 10 × 25
= 2.316
= 3.65 dB
The equivalent noise temperature is:
Teq = 290(NF − 1)
= 290(2.316 − 1)
= 382 K
X
' 1.5 The Frequency Domain

The reader is probably familiar with the time-domain representation of sig-
nals. An ordinary oscilloscope display, showing amplitude on one scale and
time on the other, is a good example.
Signals can also be described in the frequency domain. In a frequency-
domain representation, amplitude or power is shown on one axis and
frequency is displayed on the other. A spectrum analyzer gives a frequency-
domain representation of signals.
Any signal can be represented either way. For example, a 1-kHz sine wave
is shown in both ways in Figure 1.5. The time-domain representation should
need no explanation. As for the frequency domain, a sine wave has energy
only at its fundamental frequency, so it can be shown as a straight line at
that frequency.
Notice that our way of representing the signal in the frequency domain
does not show the phase of the signal. The signal in Figure 1.5(b) could be a
cosine wave just as easily as a sine wave.
One example of a frequency-domain representation with which the
reader will already be familiar is the tuning dial of an ordinary broad-
cast radio receiver. Each station is assigned a different carrier frequency. Pro-
vided that these frequencies are far enough apart, the signals will not
interfere with each other. Dividing up the spectrum in this way is known as
frequency-division multiplexing (FDM), and can only be understood by
referring to the frequency domain.
FIGURE 1.5 Sine wave in time and frequency domains
When we move into the study of individual radio signals, frequency-

domain representations are equally useful. For instance, the bandwidth of a
modulated signal generally has some fairly simple relationship to that of
the baseband signal. This bandwidth can easily be found if the baseband sig-
nal can be represented in the frequency domain. As we proceed, we will see
many other examples in which the ability to work with signals in the fre-
quency domain will be required.
Fourier Series It should be obvious by now that we need a way to move freely between the
two domains. Any well-behaved periodic waveform can be represented as a
series of sine and/or cosine waves at multiples of its fundamental frequency
plus (sometimes) a dc offset. This is known as a Fourier series. This very use-
ful (and perhaps rather surprising) fact was discovered in 1822 by Joseph
Fourier, a French mathematician, in the course of research on heat conduc-
tion. Not all signals used in communication are strictly periodic, but they
are often close enough for practical purposes.
Fourier’s discovery, applied to a time-varying signal, can be expressed
mathematically as follows:
Ao
ƒ(t ) = + A1 cos ωt + B1 sin ωt + A2 cos 2ωt + B2 sin 2ωt (1.9)
2
+ A3 cos 3ωt + B3 sin 3ωt + ⋅ ⋅ ⋅
where
ƒ(t) = any well-behaved function of time. For our purposes, ƒ(t)

will generally be either a voltage v(t) or a current i(t).
A n and B n = real-number coefficients; that is, they can be positive,
negative, or zero.
ω = radian frequency of the fundamental.
14 ! CHAPTER 1
The radian frequency can be found from the time-domain representa-

tion of the signal by finding the period (that is, the time T after which the
whole signal repeats exactly) and using the equations:
1
ƒ =
Τ
and
ω = 2πƒ
The simplest ac signal is a sinusoid. The frequency-domain representa-

tion of a sine wave has already been described and is shown in Figure 1.5 for
a voltage sine wave with a period of 1 ms and a peak amplitude of 1 V. For
this signal, all the Fourier coefficients are zero except for B 1 , which has a
value of 1 V. The equation becomes:
v(t) = sin (2000πt) V
which is certainly no surprise. The examples below use the equations for the
various waveforms shown in Table 1.1 on page 15.
EXAMPLE 1.3 Y
Find and sketch the Fourier series corresponding to the square wave in Fig-
ure 1.6(a).
FIGURE 1.6
TABLE 1.1 Fourier Series for Common Repetitive Waveforms
1. Half-wave rectified sine wave
V V 2V  cos 2ωt cos 4ωt cos 6ωt 

v (t ) = + sin ωt −  + + + ⋅ ⋅ ⋅
π 2 π  1×3 3×5 5×7 
2. Full-wave rectified sine wave

(a) With time zero at voltage zero
2V 4V  cos 2ωt cos 4ωt cos 6ωt 

v (t ) = −  + + + ⋅ ⋅ ⋅
π π  1×3 3×5 5×7 
(b) With time zero at voltage peak
2V  2 cos 2ωt 2 cos 4ωt 2 cos 6ωt 

v (t ) = 1 + − + − ⋅ ⋅ ⋅
π  1×3 3×5 5×7 
3. Square wave
(a) Odd function
4V  1 1 
v (t ) =  sin ωt + sin 3ωt + sin 5ωt + ⋅ ⋅ ⋅ (continues)
π  3 5 
16 ! CHAPTER 1
TABLE 1.1 (continued)
(b) Even function
4V  1 1 
v (t ) =  cos ωt − cos 3ωt + cos 5ωt − ⋅ ⋅ ⋅
π  3 5 
4. Pulse train
Vτ 2Vτ  sin πτ/Τ sin 2 πτ/Τ sin 3πτ/Τ 

v (t ) = +  cos ωt + cos 2ωt + cos 3ωt + ⋅ ⋅ ⋅
T T  πτ/Τ 2 πτ/Τ 3πτ/Τ 
5. Triangle wave
8V  1 1 
v (t ) = 2 
cos ωt + 2 cos 3ωt + 2 cos 5ωt + ⋅ ⋅ ⋅
π  3 5 
6. Sawtooth wave
(a) With no dc offset
2V  1 1 
v (t ) =  sin ωt − sin 2ωt + sin 3ωt − ⋅ ⋅ ⋅ (continues)
π  2 3 
TABLE 1.1 (continued)
(b) Positive-going
V V  1 1 
v (t ) = −  sin ωt + sin 2ωt + sin 3ωt + ⋅ ⋅ ⋅
2 π  2 3 
SOLUTION
A square wave is another signal with a simple Fourier representation,
although not quite as simple as for a sine wave. For the signal shown in Fig-
ure 1.6(a), the frequency is 1 kHz, as before, and the peak voltage is 1 V.
According to Table 1.1, this signal has components at an infinite number
of frequencies: all odd multiples of the fundamental frequency of 1 kHz.
However, the amplitude decreases with frequency, so that the third har-
monic has an amplitude one-third that of the fundamental, the fifth
harmonic an amplitude of one-fifth that of the fundamental, and so on.
Mathematically, a square wave of voltage with a rising edge at t = 0 and no dc
offset can be expressed as follows (see Table 1.1):
4V  1 1 
v (t ) =  sin ωt + sin 3ωt + sin 5ωt + ⋅ ⋅ ⋅
π  3 5 
where
V = peak amplitude of the square wave
ω = radian frequency of the square wave
t = time in seconds
For this example, the above equation becomes:

4  1 1 
v (t ) =  sin( 2 π × 10 t ) + sin(6π × 10 t ) + sin(10π × 10 t ) + ⋅ ⋅ ⋅ V
3 3 3
π  3 5 
This equation shows that the signal has frequency components at odd
multiples of 1 kHz, that is, at 1 kHz, 3 kHz, 5 kHz, and so on. The 1-kHz com-
ponent has a peak amplitude of
4
V1 = = 127
. V
π
18 ! CHAPTER 1
The component at three times the fundamental frequency (3 kHz) has an

amplitude one-third that of the fundamental, that at 5 kHz has an amplitude
one-fifth that of the fundamental, and so on.
4
V3 = = 0424
. V
3π
4
V5 = = 0.255 V
5π
4
V7 = = 0182
. V
7π
The result for the first four components is sketched in Figure 1.6(b). The-
oretically, an infinite number of components would be required to describe
the square wave completely, but as the frequency increases, the amplitude of
the components decreases rapidly.
X
The representations in Figures 1.6(a) and 1.6(b) are not two different sig-
nals but merely two different ways of looking at the same signal. This can be
shown graphically by adding the instantaneous values of several of the sine
waves in the frequency-domain representation. If enough of these compo-
nents are included, the result begins to look like the square wave in the
time-domain representation. Figure 1.7 shows the results for two, four, and
ten components. It was created by taking the instantaneous values of all
the components at the same time and adding them algebraically. This was
done for a large number of time values. Doing these calculations by hand
would be simple but rather tedious, so a computer was used to perform the
calculations and plot the graphs. A perfectly accurate representation of
the square wave would require an infinite number of components, but we
can see from the figure that using ten terms gives a very good representation
because the amplitudes of the higher-frequency components of the signal
are very small.
It is possible to go back and forth at will between time and frequency do-
mains, but it should be apparent that information about the relative phases
of the frequency components in the Fourier representation of the signal is
required to reconstruct the time-domain representation. The Fourier equa-
tions do have this information, but the sketch in Figure 1.6(b) does not. If
the phase relationships between frequency components are changed in a
communication system, the signal will be distorted in the time domain.
Figure 1.8 illustrates this point. The same ten coefficients were used as
in Figure 1.7, but this time the waveforms alternated between sine and
FIGURE 1.7
Construction of a square
wave from Fourier
components
20 ! CHAPTER 1
FIGURE 1.8 Addition of square-wave Fourier components with wrong phase

angles
cosine: sine for the fundamental, cosine for the third harmonic, sine for
the fifth, and so on. The result is a waveform that looks the same on the
frequency-domain sketch of Figure 1.6(b) but very different in the time
domain.
EXAMPLE 1.4 Y
Find the Fourier series for the signal in Figure 1.9(a).
SOLUTION
The positive-going sawtooth wave of Figure 1.9(a) has a Fourier series with a
dc term and components at all multiples of the fundamental frequency.
From Table 1.1, the general equation for such a wave is
A  A  1 1 
v (t ) = −    sin ωt + sin 2ωt + sin 3ωt + ⋅ ⋅ ⋅
2  π 2 3 
The first (dc) term is simply the average value of the signal.
For the signal in Figure 1.9, which has a frequency of 1 kHz and a peak
amplitude of 5 V, the preceding equation becomes:
v(t) = 2.5 − 1.59 (sin(2π × 103t) + 0.5 sin (4π × 103t)

+ 0.33 sin (6π × 103t) + ⋅ ⋅ ⋅ ) V
FIGURE 1.9
The first four voltage components are:
dc component: V0 = 2.5 V
1-kHz component: V1 = −1.59 V (the minus sign represents a phase angle
of 180 degrees. A graph of peak values will not usually indicate signs,
and a spectrum analyzer will not show phase angles)
2-kHz component: V2 = −1.59/2 = −0.795 V
3-kHz component: V3 = −1.59/3 = −0.53 V
The spectrum is shown in Figure 1.9(b).
X
Effect of Filtering As we have seen, many signals have a bandwidth that is theoretically infi-
on Signals nite. Limiting the frequency response of a channel removes some of the fre-
quency components and causes the time-domain representation to be
distorted. An uneven frequency response will emphasize some components
at the expense of others, again causing distortion. Nonlinear phase shift will
also affect the time-domain representation. For instance, shifting the phase
angles of some of the frequency components in the square-wave representa-
tion of Figure 1.8 changed the signal to something other than a square wave.
However, Figure 1.7 shows that while an infinite bandwidth may theo-
retically be required, for practical purposes quite a good representation of a
22 ! CHAPTER 1
square wave can be obtained with a band-limited signal. In general, the

wider the bandwidth, the better, but acceptable results can be obtained with
a band-limited signal. This is welcome news, because practical systems
always have finite bandwidth.
Noise in the It was pointed out earlier, in Section 1.4, that noise power is proportional
Frequency to bandwidth. That implies that there is equal noise power in each hertz of
Domain bandwidth. Sometimes this kind of noise is called white noise, since it con-
tains all frequencies just as white light contains all colors. In fact, we can talk
about a noise power density in watts per hertz of bandwidth. The equation
for this is very simply derived. We start with Equation (1.3):
P N = kTB
This gives the total noise power in bandwidth, B. To find the power per
hertz, we just divide by the bandwidth to get an even simpler equation:
N 0 = kT (1.10)
where
N 0 = noise power density in watts per hertz

T = temperature in kelvins
EXAMPLE 1.5 Y
(a) A resistor has a noise temperature of 300 K. Find its noise power density
and sketch the noise spectrum.
(b) A system with a noise temperature of 300 K operates at a frequency of
100 MHz with a bandwidth of 1 MHz. Sketch the noise spectrum.
SOLUTION
(a) From Equation (1.10):
N0 = kT
= 1.38 × 10−23 J/K × 300 K
= 4.14 × 10−21 W/Hz
The spectrum is shown in Figure 1.10(a). Note that the spectrum is a

straight line, showing that the noise power density is the same at all frequen-
cies. The frequency scale shown runs from 0 to 1 GHz, but theoretically the
spectrum remains flat indefinitely. Real systems, of course, never have infi-
nite frequency response.
FIGURE 1.10
(b) Here the noise power density is the same as in part (a) but only over the
1-MHz bandwidth illustrated. Hence the band-limited spectrum of Fig-
ure 1.10(b). The exact shape of the pattern will depend on the type of
filter used. In the sketch an ideal filter, with complete attenuation out-
side the passband, is assumed.
X
' 1.6 The Radio-Frequency Spectrum

Radio waves are a form of electromagnetic radiation, as are infrared, visible
light, ultraviolet light, and gamma rays. The major difference is in the fre-
quency of the waves. The portion of the frequency spectrum that is useful
for radio communication at present extends from roughly 100 kHz to about
50 GHz. Table 1.2 shows the conventional designations of the various
24 ! CHAPTER 1
TABLE 1.2 The Radio-Frequency Spectrum
Frequency Frequency Wavelength Wavelength

Designation Range Range Designation
Extremely High 30–300 GHz 1 mm–1 cm Millimeter Waves

Frequency (EHF)
Super High Frequency 3–30 GHz 1–10 cm Microwaves

(SHF) (microwave
region
conventionally
starts at 1 GHz)
Ultra High Frequency 300 Mhz–3 GHz 10 cm–1 m

(UHF)
Very High Frequency 30–300 MHz 1–10 m

(VHF)
High Frequency (HF) 3–30 MHz 10–100 m Short Waves
Medium Frequency 300 kHz–3 MHz 100 m–1 km Medium Waves

(MF)
frequency ranges and their associated wavelength ranges. Note that micro-
waves and millimeter waves are wavelength designations and fit only ap-
proximately into the frequency designations. Wireless communication as
described in this book occupies mainly the VHF, UHF, and SHF portions of
the spectrum. Lower-frequency systems need inconveniently large antennas
and involve methods of signal propagation that are undesirable for the sys-
tems we look at. Extremely high frequencies are still difficult to generate and
amplify at reasonable cost, though that may well change in the future.
Conversion between frequency and wavelength is quite easy. The gen-
eral equation that relates frequency to wavelength for any wave is
v = ƒλ (1.11)
where
v = velocity of propagation of the wave in meters per second

ƒ = frequency of the wave in hertz
λ = wavelength in meters
For radio waves in free space (and air is generally a reasonable approxima-
tion to free space) the velocity is the same as that of light: 300 × 106 m/s. The
usual symbol for this quantity is c. Equation (1.11) then becomes:
c = ƒλ (1.12)
EXAMPLE 1.6 Y
Find the wavelength of a signal at each of the following frequencies:
(a) 850 MHz (cell phone range)
(b) 1.9 GHz (Personal Communication Systems range)
(c) 28 GHz (used for Local Multipoint Distribution Systems (LMDS) for local
delivery of television signals by microwave)
SOLUTION
For all of these the method is the same. The problem is repeated to give the
reader a feeling for some of the frequencies and wavelengths used in wireless
communication. Simply rewrite Equation (1.12) in the form
c
λ =
ƒ
300 × 106
(a) λ = = 0353
. m = 353 mm
850 × 106
300 × 106
(b) λ = = 0158
. m = 158 mm
. × 109
19
300 × 106
(c) λ = = 00107
. m = 107
. mm
28 × 109
X
Bandwidth The carrier wave is a sine wave for almost any communication system. A sine
Requirements wave, of course, exists at only one frequency and therefore occupies zero
bandwidth. As soon as the signal is modulated to transmit information,
however, the bandwidth increases. A detailed knowledge of the bandwidth
of various types of modulated signals is essential to the understanding of the
communication systems to be described in this book. Thorough study of sig-
nal bandwidths will have to wait until we know more about the modulation
schemes referred to above. However, at this time it would be well to look at
the concept of bandwidth in more general terms.
26 ! CHAPTER 1
First, bandwidth in radio systems is always a scarce resource. Not all fre-
quencies are useful for a given communication system, and there is often
competition among users for the same part of the spectrum. In addition, as
we have seen, the degrading effect of noise on signals increases with band-
width. Therefore, in most communication systems it is important to con-
serve bandwidth to the extent possible.
There is a general rule known as Hartley’s Law which relates bandwidth,
time, and information content. We will not yet be able to use it for actual cal-
culations, but it would be well to note it for future reference, as Hartley’s Law
applies to the operation of all communication systems. Here it is:
I = ktB (1.13)
where
I = amount of information to be transmitted in bits
k = a constant that depends on the modulation scheme and the
signal-to-noise ratio
t = time in seconds
B = bandwidth in hertz
Our problem thus far is that we do not have precise ways of quantifying
either the amount of information I or the constant k. However, the general
form of the equation is instructive. It tells us that the rate at which informa-
tion is transmitted is proportional to the bandwidth occupied by a commu-
nication system. To transmit more information in a given time requires
more bandwidth (or a more efficient modulation scheme).
EXAMPLE 1.7 Y
Telephone voice transmission requires transmission of frequencies up to
about 3.4 kHz. Broadcast video using the ordinary North American standard,
on the other hand, requires transmission of frequencies up to 4.2 MHz. If a
certain modulation scheme needs 10 kHz for the audio transmission, how
much bandwidth will be needed to transmit video using the same method?
SOLUTION
Hartley’s Law states that bandwidth is proportional to information rate,
which in this case is given by the baseband bandwidth. Assuming that audio
needs a bandwidth from dc to 3.4 kHz, while video needs dc to 4.2 MHz, the
bandwidth for video will be
BV
BTV = BTA ×
BA
where
BTV = transmission bandwidth for video
BTA = transmission bandwidth for audio
BV = baseband bandwidth for video
BA = baseband bandwidth for audio
Substituting the numbers from the problem we get
BV
BTV = BTA ×
BA
4.2 MHz
= 10 kHz ×
34
. kHz
= 12.3 MHz
Obviously the type of baseband signal to be transmitted makes a great

deal of difference to any spectrum management plan.
X
Hartley’s Law also shows that it is possible to reduce the bandwidth

required to transmit a given amount of information by using more time for
the process. This is an important possibility where data must be sent, but of
course it is not practical when real-time communication is required—in
a telephone call, for instance. The reader may have experienced this trade-
off of time for bandwidth in downloading an audio or video file from the
internet. If the bandwidth of the connection is low, such a file may take
much longer to download than it does to play.
Frequency Reuse Spectrum space in wireless systems is nearly always in short supply. Even
with the communication bandwidth restricted as much as possible, there is
often more traffic than can be accommodated. Of course the spectrum used
for a given purpose in one area can be reused for a different purpose in an-
other area that is physically far enough away that signals do not travel from
one area to the other with sufficient strength to cause unacceptable interfer-
ence levels. How far that is depends on many factors such as transmitter
power, antenna gain and height, and the type of modulation used. Many re-
cent systems, such as cellular telephony, automatically reduce transmitter
power to the minimum level consistent with reliable communication,
thereby allowing frequencies to be reused at quite small distances. Such
schemes can use spectrum very efficiently.
28 ! CHAPTER 1
' 1.7 “Convergence” and Wireless Communication

There has been much talk recently about “convergence,” the merger of all
kinds of previously separate electronic systems, for example, telephony
(both wireline and wireless), broadcast and cable television, and data com-
munication (most notably the internet). Convergence does seem to be
beginning to happen, though more slowly than many had anticipated. The
process is slowed both by technical problems involving the very different
types of signals and media used in fields that have evolved separately for
many years, and by more mundane but equally serious problems of regula-
tory jurisdiction and commercial interests. It is by no means clear exactly
how wireless communication will fit into the final picture, even if a field as
dynamic as this can be imagined to have a final state. Some people (many of
whom seem to work for wireless phone companies) have suggested that
eventually wireless phones will replace wireline equipment, and everyone
will have one phone (with one number) which they will carry with them
everywhere. Wired communication will then do what it does best: carry
high-bandwidth signals like television to fixed locations. On the other
hand, there is very serious development work underway involving
high-bandwidth wireless communication for world-wide web access from
portable devices, for instance. If it is unclear even to the experts what the
future holds, we must be careful in our predictions. This much is certain
though: wireless communication will be a large part of the total communica-
tion picture, and a knowledge of the technologies involved will certainly
help a technologist to understand future developments as they occur.
' Summary The main points to remember from this chapter are:
( Any wireless communication system requires a transmitter and a receiver
connected by a channel.
( Simplex communication systems allow communication in one direction
only. Half-duplex systems are bidirectional, but work in only one direc-
tion at a time. Full-duplex systems can communicate in both directions
simultaneously.
( Most wireless networks are variations of the star network configuration,
often with radio repeaters at the hub.
( Radio systems transmit information by modulating a sine-wave carrier
signal. Only three basic parameters can be modulated: amplitude, fre-
quency, and phase. Many variations are possible, however.
( The ratio of signal power to noise power is one of the most important
specifications for any communication system. Thermal noise is the most
important type of noise in most wireless systems.
( The frequency domain is useful for the observation and understanding of
both signals and noise in communication systems.
( Many signals can be analyzed in the frequency domain with the aid of
Fourier series.
( The bandwidth required by a system depends on the modulation scheme
employed and the information transmission rate required. Bandwidth
should be kept to the minimum necessary to reduce noise problems and
to conserve radio-frequency spectrum.
( Convergence is a term describing the possible merger of many different
kinds of communication and related technologies.
( Equation List
e(t) = Ec sin(ωct + θ) (1.1)
PN = kTB (1.3)
T(K) = T(°C) + 273 (1.4)
( S/N )i
NF = (1.5)
( S/N )o
NF2 − 1 NF3 − 1
NFT = NF1 + + +⋅⋅⋅ (1.6)
A1 A1A2
Teq = 290(NF − 1) (1.8)
Ao
ƒ(t ) = + A1 cos ωt + B1 sin ωt + A2 cos 2ωt + B2 sin 2ωt (1.9)
2
+ A3 cos 3ωt + B3 sin 3ωt + ⋅ ⋅ ⋅
N0 = kT (1.10)
v = ƒλ (1.11)
c = ƒλ (1.12)
I = ktB (1.13)
30 ! CHAPTER 1
( Key Terms
bandwidth portion of frequency spectrum occupied by a signal
baseband information signal
carrier high-frequency signal which is modulated by the baseband signal
in a communication system
Citizens’ Band (CB) radio short-distance unlicensed radio
communication system
demodulation recovery of a baseband signal from a modulated signal
Fourier series expression showing the structure of a signal in the
frequency domain
frequency domain method of analyzing signals by observing them on a
power-frequency plane
frequency-division multiplexing combining of several signals into one
communication channel by assigning each a different carrier
frequency
full-duplex communication two-way communication in which both
terminals can transmit simultaneously
half-duplex communication two-way communication system in which
only one station can transmit at a time
handoff transfer of a call in progress from one cell site to another
Improved Mobile Telephone Service (IMTS) a mobile telephone service,
now obsolescent, using trunked channels but not cellular in nature
intelligence information to be communicated
modulating signal the information signal that is used to modulate a
carrier for transmission
network an organized system for communicating among terminals
noise an unwanted random signal that extends over a considerable
frequency spectrum
noise power density the power in a one-hertz bandwidth due to a noise
source
personal communication system (PCS) a cellular telephone system
designed mainly for use with portable (hand-carried) telephones
public switched telephone network (PSTN) the ordinary public wireline
phone system
repeater a transmitter-receiver combination used to receive and
retransmit a signal
signal-to-noise ratio ratio between the signal power and noise power at
some point in a communication system
simplex a unidirectional communication system; for example,
broadcasting
spectrum analyzer test instrument that typically displays signal power as
a function of frequency
star network a computer network topology in which each terminal is
connected to a central mainframe or server
time domain representation of a signal as a function of time and some
other parameter, such as voltage
white noise noise containing all frequencies with equal power in every
hertz of bandwidth
( Questions
1. Why were the first radio communication systems used for telegraphy
only?
2. When were the first two-way mobile radio communication systems
installed, and for what purpose?
3. What characteristics of CB radio led to its great popularity?
4. Why are cellular radio systems more efficient in their use of spectrum
than earlier systems?
5. What types of modulation are used with cellular phones?
6. Explain the differences among simplex, half-duplex, and full-duplex
communication.
7. Identify each of the following communication systems as simplex,
half-duplex, or full-duplex.
(a) cordless telephone
(b) television broadcast
(c) intercom with push-to-talk bar
8. Why is it necessary to use a high-frequency carrier with a radio commu-
nication system?
9. Name the three basic modulation methods.
10. Suppose that a voice frequency of 400 Hz is transmitted using a trans-
mitter operating at 800 MHz. Which of these is:
(a) the information frequency?
(b) the carrier frequency?
32 ! CHAPTER 1
(c) the baseband frequency?

(d) the modulating frequency?
11. What effect does doubling the bandwidth of a system have on its noise
level?
12. What is the meaning of the signal-to-noise ratio for a system, and why is
it important?
13. What is the difference between the kelvin and Celsius temperature
scales?
14. State whether the time or frequency domain would be more appropriate
for each of the following:
(a) a display of all UHF television channels
(b) measuring the peak voltage of a waveform
(c) measuring the bandwidth of a waveform
(d) determining the rise time of a signal
15. What is meant by the term frequency-division multiplexing?
16. Why is thermal noise sometimes called white noise?
17. Give the frequency designation for each of the following systems:
(a) marine radio at 160 MHz
(b) cell phones at 800 MHz
(c) direct-to-home satellite television at 12 GHz
(d) CB radio at 27 Mhz
( Problems
1. Express the frequency of a 10-kHz signal in radians per second.
2. Find the noise power produced by a resistor at a temperature of 60 °C in
a bandwidth of 6 MHz in
(a) watts
(b) dBm
(c) dBf
3. If the signal power at a certain point in a system is 2 W and the noise
power is 50 mW, what is the signal-to-noise ratio, in dB?
4. Sketch the spectrum for the half-wave rectified signal in Figure 1.11, show-
ing harmonics up to the fifth. Show the voltage and frequency scales and
indicate whether your voltage scale shows peak or RMS voltage.
FIGURE 1.11
5. Sketch the frequency spectrum for the triangle wave shown in Fig-
ure 1.12 for harmonics up to the fifth. Show the voltage and frequency
scales.
FIGURE 1.12
6. A 1-kHz square wave passes through each of three communication

channels whose bandwidths are given below. Sketch the output in the
time domain for each case.
(a) 0 to 10 kHz
(b) 2 kHz to 4 kHz
(c) 0 to 4 kHz
7. Sketch the spectrum for the pulse train shown in Figure 1.13.
FIGURE 1.13
34 ! CHAPTER 1
8. Sketch the spectrum for the sawtooth waveform in Figure 1.14. Explain
why this waveform has no dc component, unlike the sawtooth wave-
form in Example 1.3.
FIGURE 1.14
9. Visible light consists of electromagnetic radiation with free-space wave-

lengths between 400 and 700 nanometers (nm). Express this range in
terms of frequency.
10. Equation (1.11) applies to any kind of wave. The velocity of sound
waves in air is about 344 m/s. Calculate the wavelength of a sound wave
with a frequency of 1 kHz.
Analog Modulation
Schemes
2
Objectives
( Explain the concept of modulation.
( Describe the differences among analog modulation schemes.
( Analyze amplitude-modulated signals in the time and frequency domains.
( Analyze frequency-modulated signals in the frequency domain.
( Describe phase modulation.
( Explain the need for pre-emphasis and de-emphasis with FM signals.
36 ! CHAPTER 2
' 2.1 Introduction

In Chapter 1 we saw that modulation is necessary in order to transmit intelli-
gence over a radio channel. A radio-frequency signal can be modulated by
either analog or digital information. In either case, the information sig-
nal must change one or more of three parameters: amplitude, frequency, and
phase.
With the exception of Morse code transmission, which is digital though
not binary, the earliest wireless communication systems used analog modu-
lation, and these schemes are still very popular in such diverse areas as
broadcasting and cellular telephony. Analog modulation schemes tend to
be more intuitive and hence easier to understand than their digital variants,
so they will be considered first. Of the analog schemes, amplitude modula-
tion (AM) is simplest and was first historically, therefore, it seems logical
to begin with it. Frequency modulation (FM) is more common in modern
systems, so it will be discussed next. Finally, phase modulation (PM) is
seen less often than the others in analog systems, but it is very common
in digital communication, so we will introduce it here but leave the details
for later.
' 2.2 Amplitude Modulation

An amplitude-modulated signal can be produced by using the instantaneous
amplitude of the information signal (the baseband or modulating sig-
nal) to vary the peak amplitude of a higher-frequency signal. Figure 2.1(a)
shows a baseband signal consisting of a 1-kHz sine wave, which can be com-
bined with the 10-kHz carrier signal shown in Figure 2.1(b) to produce the
amplitude-modulated signal of Figure 2.1(c). If the peaks of the individual
waveforms of the modulated signal are joined, an envelope results that re-
sembles the original modulating signal. It repeats at the modulating fre-
quency, and the shape of each “half” (positive or negative) is the same as
that of the modulating signal.
Figure 2.1(c) shows a case where there are only ten cycles of the carrier
for each cycle of the modulating signal. In practice, the ratio between car-
rier frequency and modulating frequency is usually much greater. For in-
stance, an AM citizens’ band (CB) station would have a carrier frequency of
about 27 MHz and a modulating frequency on the order of 1 kHz. A wave-
form like this is shown in Figure 2.2. Since there are thousands of cycles of
the carrier for each cycle of the envelope, the individual RF cycles are not
visible, and only the envelope can be seen.
ANALOG MODULATION SCHEMES ! 37
FIGURE 2.1 Amplitude modulation
The AM envelope allows for very simple demodulation. All that is neces-
sary is to rectify the signal to remove one-half of the envelope, then low-pass
filter the remainder to recover the modulation. A simple but quite practical
AM demodulator is shown in Figure 2.3.
Because AM relies on amplitude variations, it follows that any amplifier
used with an AM signal must be linear, that is, it must reproduce amplitude
variations exactly. This principle can be extended to any signal that has an
envelope. This point is important, because nonlinear amplifiers are typically
less expensive and more efficient than linear amplifiers.
38 ! CHAPTER 2
FIGURE 2.2
AM envelope
FIGURE 2.3
AM demodulator
Time-Domain Now that we understand the general idea of AM, it is time to examine the
Analysis system in greater detail. We shall look at the modulated signal in both
the time and frequency domains, as each method emphasizes some of the
important characteristics of AM. The time domain is probably more familiar,
so we begin there.
AM is created by using the instantaneous modulating signal voltage to

vary the amplitude of the modulated signal. The carrier is almost always a
sine wave. The modulating signal can be a sine wave, but is more often an ar-
bitrary waveform, such as an audio signal. However, an analysis of sine-wave
modulation is very useful, since Fourier analysis often allows complex sig-
nals to be expressed as a series of sinusoids.
We can express the above relationship by means of an equation:
v(t) = (E c + e m ) sin ω c t (2.1)

where
v(t) = instantaneous amplitude of the modulated signal in volts
E c = peak amplitude of the carrier in volts
e m = instantaneous amplitude of the modulating signal in volts
ω c = the frequency of the carrier in radians per second
t = time in seconds
If the modulating (baseband) signal is a sine wave, Equation (2.1) has the fol-
lowing form:
v(t) = (E c + E m sin ω m t) sin ω c t (2.2)

where
E m = peak amplitude of the modulating signal in volts
ω m = frequency of the modulating signal in radians per second
and the other variables are as defined for Equation (2.1).
EXAMPLE 2.1 Y
A carrier with an RMS voltage of 2 V and a frequency of 1.5 MHz is modu-
lated by a sine wave with a frequency of 500 Hz and amplitude of 1 V RMS.
Write the equation for the resulting signal.
SOLUTION
First, note that Equation (2.2) requires peak voltages and radian frequencies.
We can easily get these as follows:
Ec = 2 × 2 V
= 2.83 V
Em = 2 × 1 V
= 1.41 V
40 ! CHAPTER 2
ωc = 2π × 1.5 × 106
= 9.42 × 106 rad/s
ωm = 2π × 500
= 3.14 × 103 rad/s
So the equation is
v(t) = (Ec + Em sin ωmt) sin ωct

= [2.83 + 1.41 sin (3.14 × 103t)] sin (9.42 × 106t) V
X
Modulation The ratio between the amplitudes of the modulating signal and the carrier is
Index defined as the modulation index, m. Mathematically,
m = E m /E c (2.3)
Modulation can also be expressed as a percentage, by multiplying m by 100.
For example, m = 0.5 corresponds to 50% modulation.
Substituting m into Equation (2.2) gives:
v(t) = E c (1 + m sin ω m t) sin ω c t (2.4)
EXAMPLE 2.2 Y
Calculate m for the signal of Example 2.1 and write the equation for this sig-
nal in the form of Equation (2.4).
SOLUTION
To avoid an accumulation of round-off errors we should go back to the origi-
nal voltage values to find m.
m = Em /Ec
= 1/2
= 0.5
It is all right to use the RMS values for calculating this ratio, as the factors of
2, if used to find the peak voltages, will cancel.
Now we can rewrite the equation:
v(t) = Ec(1 + m sin ωmt) sin ωct

= 2.83 [1+ 0.5 sin (3.14 × 103t)] sin (9.42 × 106t) V
X
It is worthwhile to examine what happens to Equation (2.4) and to the

modulated waveform, as m varies. To start with, when m = 0, Em = 0 and we
have the original, unmodulated carrier. As m varies between 0 and 1, the
changes due to modulation become more pronounced. Resultant waveforms
for several values of m are shown in Figure 2.4. Note especially the result
for m = 1 or 100%. Under these conditions the peak signal voltage will vary
between zero and twice the unmodulated carrier amplitude.
FIGURE 2.4 Envelopes for various values of m

42 ! CHAPTER 2
Overmodulation When the modulation index is greater than one, the signal is said to be
overmodulated. There is nothing in Equation (2.4) that would seem to pre-
vent E m from being greater than E c , that is, m greater than one. There are
practical difficulties, however. Figure 2.5(a) shows the result of simply sub-
stituting m = 2 into Equation (2.4). As you can see, the envelope no longer re-
sembles the modulating signal. Thus the type of demodulator described
earlier no longer gives undistorted results, and the signal is no longer a
full-carrier AM signal.
Whenever we work with mathematical models, we must remember
to keep checking against physical reality. This situation is a good example.
It is possible to build a circuit that does produce an output that agrees with
Equation (2.4) for m greater than 1. However, most practical AM modulators
produce the signal shown in Figure 2.5(b) under these conditions. This
waveform is completely useless for communication. In fact, if this signal
were subjected to Fourier analysis, the sharp “corners” on the waveform as
the output goes to zero on negative modulation peaks would be found to
represent high-frequency components added to the original baseband
signal. This type of overmodulation creates spurious frequencies known as
splatter, which cause the modulated signal to have increased bandwidth.
This can cause interference with a signal on an adjacent channel.
From the foregoing, we can conclude that for full-carrier AM, m must be
in the range from 0 to 1. Overmodulation creates distortion in the demodu-
lated signal and may result in the signal occupying a larger bandwidth than
normal. Since spectrum space is tightly controlled by law, overmodulation
of an AM transmitter is actually illegal, and means must be provided to pre-
vent it.
FIGURE 2.5 Overmodulation

Modulation Index Practical AM systems are seldom used to transmit sine waves, of course.
for Multiple The information signal is more likely to be a voice signal, which contains
Modulating many frequencies. When there are two or more sine waves of different,
Frequencies uncorrelated frequencies (that is, frequencies that are not multiples of each
other) modulating a single carrier, m is calculated by using the equation
mT = m12 + m22 + ⋅ ⋅ ⋅ (2.5)
where
m T = total resultant modulation index
m 1 , m 2 , etc. = modulation indices due to the individual
modulating components.
EXAMPLE 2.3 Y
Find the modulation index if a 10-volt carrier is amplitude modulated by
three different frequencies, with amplitudes of 1, 2, and 3 volts respectively.
SOLUTION
The three separate modulation indices are:
m1 = 1/10 = 0.1
m2 = 2/10 = 0.2
m3 = 3/10 = 0.3
mT = m12 + m22 + m32

= 01 . 2 + 0.2 2 + 03
. 2
= 0374
.
X
Measurement of If we let Em and Ec be the peak modulation and carrier voltages respectively,
Modulation then we can see, from Equation (2.4), that the maximum envelope voltage is
Index simply
E max = E c (1 + m) (2.6)
and the minimum envelope voltage is
E min = E c (1 − m) (2.7)
44 ! CHAPTER 2
Note, by the way, that these results agree with the conclusions expressed
earlier: for m = 0, the peak voltage is Ec, and for m = 1, the envelope voltage
ranges from 2Ec to zero.
Applying a little algebra to the above expressions, it is easy to show that
E max − E min
m = (2.8)
E max + E min
Of course, doubling both Emax and Emin will have no effect on this equa-
tion, so it is quite easy to find m by displaying the envelope on an oscillo-
scope and measuring the maximum and minimum peak-to-peak values for
the envelope voltage.
EXAMPLE 2.4 Y
Calculate the modulation index for the waveform shown in Figure 2.2.
SOLUTION
It is easiest to use peak-to-peak values with an oscilloscope. From the figure
we see that:
E max = 150 mV p-p E min = 70 mV p-p
E max − E min
m =
E max + E min
150 − 70
=
150 + 70
= 0364
.
X
Frequency- So far we have looked at the AM signal exclusively in the time domain, that
Domain Analysis is, as it can be seen on an oscilloscope. In order to find out more about this
signal, however, it is necessary to consider its spectral makeup. We could use
Fourier methods to do this, but for a simple AM waveform it is easier, and
just as valid, to use trigonometry.
To start, we should observe that although both the carrier and the modu-
lating signal may be sine waves, the modulated AM waveform is not a sine
wave. This can be seen from a simple examination of the waveform of Fig-
ure 2.1(c). It is important to remember that the modulated waveform is
not a sine wave when, for instance, trying to find RMS from peak voltages.
The usual formulas, so laboriously learned in fundamentals courses, do not
apply here!
If an AM signal is not a sine wave, then what is it? We already have a

mathematical expression, given by Equation (2.4):
v(t) = E c (1 + m sin ω m t) sin ω c t

Expanding it and using a trigonometric identity will prove useful. Ex-
panding gives
v(t) = E c sin ω c t + mE c sin ω c t sin ω m t

The first term is just the carrier. The second can be expanded using the trigo-
nometric identity
1
sin A sin B =
2
[cos( A − B) − cos( A + B)]
to give
mE c
v (t ) = E c sin ω ct +
2
[cos(ω c − ω m )t − cos(ω c + ω m )t ]
which can be separated into three distinct terms:
mE c mE c
v (t ) = E c sin ω ct + cos(ω c − ω m )t − cos(ω c + ω m )t (2.9)
2 2
We now have, in addition to the original carrier, two other sinusoidal
waves, one above the carrier frequency and one below. When the complete
signal is sketched in the frequency domain as in Figure 2.6, we see the carrier
and two additional frequencies, one to each side. These are called, logically
enough, side frequencies. The separation of each side frequency from the
carrier is equal to the modulating frequency; and the relative amplitude
of the side frequency, compared with the carrier, is proportional to m, be-
coming half the carrier voltage for m = 1. In a real situation there is generally
more than one set of side frequencies, because there is more than one modu-
lating frequency. Each modulating frequency produces two side frequencies.
Those above the carrier can be grouped into a band of frequencies called the
upper sideband. There is also a lower sideband, which looks like a mirror
image of the upper, reflected in the carrier.
FIGURE 2.6
AM in the frequency
domain
46 ! CHAPTER 2
From now on we will generally use the term sideband, rather than side
frequency, even for the case of single-tone modulation, because it is more
general and more commonly used in practice.
Mathematically, we have:
ƒ usb = ƒ c + ƒ m (2.10)
ƒ lsb = ƒ c − ƒ m (2.11)
mE c
E lsb = E usb = (2.12)
2
where
ƒ usb = upper sideband frequency
ƒ lsb = lower sideband frequency
E usb = peak voltage of the upper-sideband component
E lsb = peak voltage of the lower-sideband component
E c = peak carrier voltage
EXAMPLE 2.5 Y
(a) A 1-MHz carrier with an amplitude of 1 volt peak is modulated by a
1-kHz signal with m = 0.5. Sketch the voltage spectrum.
(b) An additional 2-kHz signal modulates the carrier with m = 0.2. Sketch
the voltage spectrum.
SOLUTION
(a) The frequency scale is easy. There are three frequency components. The
carrier is at:
ƒc = 1 MHz
The upper sideband is at:
ƒusb = ƒc + ƒm
= 1 MHz + 1 kHz
= 1.001 MHz
The lower sideband is at:
ƒlsb = ƒc − ƒm
= 1 MHz − 1 kHz
= 0.999 MHz
Next we have to determine the amplitudes of the three components. The car-
rier is unchanged with modulation, so it remains at 1 V peak. The two side-
bands have the same peak voltage:
mE c
E lsb = E usb =
2
. ×1
05
=
2
= 0.25 V
Figure 2.7(a) shows the solution.

(b) The addition of another modulating signal at a different frequency sim-
ply adds another set of side frequencies. It does not change anything
that was done in part (a). The new frequency components are at 1.002
and 0.998 MHz, and their amplitude is 0.1 volt. The result is shown in
Figure 2.7(b).
FIGURE 2.7
X
Bandwidth Signal bandwidth is one of the most important characteristics of any modu-
lation scheme. In general, a narrow bandwidth is desirable. In any situation
where spectrum space is limited, a narrow bandwidth allows more signals
to be transmitted simultaneously than does a wider bandwidth. It also
allows a narrower bandwidth to be used in the receiver. The receiver must
have a wide enough bandwidth to pass the complete signal, including all
the sidebands, or distortion will result. Since thermal noise is evenly distrib-
uted over the frequency domain, a narrower receiver bandwidth includes
48 ! CHAPTER 2
less noise and this increases the signal-to-noise ratio, unless there are other
factors.
The bandwidth calculation is very easy for AM. The signal extends from
the lower side frequency, which is the difference between the carrier fre-
quency and the modulation frequency, to the upper side frequency, at the
sum of the carrier frequency and the modulation frequency. The difference
between these is simply twice the modulation frequency. If there is more
than one modulating frequency, the bandwidth is twice the highest modulat-
ing frequency. Mathematically, the relationship is:
B = 2F m (2.13)
where
F m = the highest modulating frequency in hertz
EXAMPLE 2.6 Y
Citizens’ band radio channels are 10 kHz wide. What is the maximum modu-
lation frequency that can be used if a signal is to remain entirely within its
assigned channel?
SOLUTION
From Equation (2.13) we have
B = 2 Fm
so
B
Fm =
2
10 kHz
=
2
= 5 kHz
X
Power Power is important in any communication scheme, because the crucial

Relationships signal-to-noise ratio at the receiver depends as much on the signal power
being large as on the noise power being small. The power that is most impor-
tant, however, is not the total signal power but only that portion that is used
to transmit information. Since the carrier in an AM signal remains un-

changed with modulation, it contains no information. Its only function is to
aid in demodulating the signal at the receiver. This makes AM inherently
wasteful of power, compared with some other modulation schemes to be
described later.
The easiest way to look at the power in an AM signal is to use the fre-
quency domain. We can find the power in each frequency component, then
add to get total power. We shall assume that the signal appears across a resis-
tance R, so that reactive volt-amperes can be ignored. We will also assume
that the power required is average power.
Suppose that the modulating signal is a sine wave. Then the AM signal
consists of three sinusoids, the carrier and two sidebands, as shown in Fig-
ure 2.6.
The power in the carrier is easy to calculate, since the carrier by itself is a
sine wave. The carrier is given by the equation
ec = E c sin ω c t
where
e c = instantaneous carrier voltage
E c = peak carrier voltage
ω c = carrier frequency in radians per second
Since Ec is the peak carrier voltage, the power developed when this signal
appears across a resistance R is simply
2
 Ec 
 
 
 2 
Pc =
R
2
E
= c
2R
The next step is to find the power in each sideband. The two frequency
components have the same amplitude, so they have equal power. Assuming
sine-wave modulation, each sideband is a cosine wave whose peak voltage is
given by Equation (2.12):
E lsb = E usb = mE c /2
Since the carrier and both sidebands are part of the same signal, the side-
bands appear across the same resistance, R, as the carrier. Looking at the
lower sideband,
50 ! CHAPTER 2
2
E lsb
Plsb =
2R
2
 mE c 
 
 2 
=
2R
m 2 E c2
=
4 × 2R
m2 E2
= × c
4 2R
m2
Plsb = Pusb = Pc (2.14)
4
Since the two sidebands have equal power, the total sideband power is
given by
m2
Psb = Pc (2.15)
2
The total power in the whole signal is just the sum of the power in the
carrier and the sidebands, so it is
 m2 
Pt = Pc +   Pc
 2 
or
 m2 
Pt = Pc  1 +  (2.16)
 2 
These latest equations tell us several useful things:

( The total power in an AM signal increases with modulation, reaching
a value 50% greater than that of the unmodulated carrier for 100%
modulation.
( The extra power with modulation goes into the sidebands: the carrier
power does not change with modulation.
( The useful power, that is, the power that carries information, is rather
small, being a maximum of one-third of the total signal power for
100% modulation and much less at lower modulation indices. For
this reason, AM transmission is more efficient when the modulation
index is as close to 1 as practicable.
EXAMPLE 2.7 Y
An AM transmitter has a carrier power output of 50 W. What would be the to-
tal power produced with 80% modulation?
SOLUTION
 m2 
Pt = Pc  1 + 
 2 
 0.82 
= 50 W  1 + 
 2 
= 66 W
X
Measuring Since the ratio between sideband and carrier power is a simple function of m,
Modulation it is quite possible to measure modulation index by observing the spectrum
Index in the of an AM signal. The only complication is that spectrum analyzers generally
Frequency display power ratios in decibels. The power ratio between sideband and car-
Domain rier power can easily be found from the relation:
Plsb  dB 
= antilog   (2.17)
Pc  10 
where
P c = carrier power
P lsb = power in one sideband
dB = difference between sideband and carrier signals, measured
in dB (this number will be negative)
Once the ratio between carrier and sideband power has been found, it is
easy to find the modulation index from Equation (2.14):
m2
Plsb = Pc
4
4 Plsb
m2 =
Pc
Plsb
m = 2 (2.18)
Pc
52 ! CHAPTER 2
Although the time-domain measurement described earlier is simpler

and uses less-expensive equipment, frequency-domain measurement en-
ables much smaller values of m to be found. A modulation level of 5%, for
instance, would be almost invisible on an oscilloscope, but it is quite obvi-
ous, and easy to measure, on a spectrum analyzer. The spectrum analyzer
also allows the contribution from different modulating frequencies to be
observed and calculated separately.
EXAMPLE 2.8 Y
Calculate the modulation frequency and modulation index for the spectrum
analyzer display shown in Figure 2.8.
FIGURE 2.8
SOLUTION
First let us find ƒm. The difference between the carrier and either sideband is
2 divisions at 5 kHz/division, or 10 kHz. So ƒm = 10 kHz.
Next, we need to find the modulation index. The two sidebands have
the same power, so we can use either. The spectrum analyzer is set for 10
dB/division, and each sideband is 1.5 divisions, or 15 dB, below the carrier.
This corresponds to a power ratio of
Plsb  −15 
= antilog  
Pc  10 
= 0.0316
Plsb
m = 2
Pc
= 2 00316
.
= 0356
.
X
' 2.3 Suppressed-Carrier AM Systems

It is possible to improve the efficiency and reduce the bandwidth of an AM
signal by removing the carrier and/or one of its sidebands. Recall from the
previous section that the carrier has at least two-thirds of the power in an
AM signal, but none of the information. This can be understood by noting
that the presence of modulation has no effect on the carrier. Removing the
carrier to create a double-sideband suppressed-carrier (DSBSC) AM signal
should therefore result in a power gain for the information-carrying part of
the signal of at least three (or about 4.8 dB), assuming that the power re-
moved from the carrier could be put into the sidebands. Note also that the
upper and lower sidebands are mirror images of each other, containing ex-
actly the same information. Removing one of these sidebands would reduce
the signal bandwidth by half. Assuming that the receiver bandwidth is also
reduced by half, this should result in a reduction of the noise power by a fac-
tor of two (3 dB). Therefore, removing the carrier and one sideband should
cause the resulting single-sideband suppressed-carrier AM (SSBSC or just SSB)
signal to have a signal-to-noise improvement of 7.8 dB or more, compared
with full-carrier double-sideband AM.
It is quite practical to remove the carrier from an AM signal, provided it
is re-inserted at the receiver. Removing one sideband is also effective, and
there is no need to replace it. Single-sideband AM is quite popular for voice
communication systems operating in the high-frequency range (3–30 MHz)
and has also been used for terrestrial point-to-point microwave links carry-
ing telephone and television signals.
Figure 2.9 shows the idea. Figure 2.9(a) shows the baseband spectrum of
a typical voice signal. In Figure 2.9(b) we have double-sideband suppressed-
carrier AM (DSBSC). The carrier frequency of 1 MHz is indicated but there is
no carrier, just the upper and lower sidebands. In Figure 2.9(c), the lower
sideband has been removed and only the upper sideband is transmitted.
54 ! CHAPTER 2
FIGURE 2.9
DSB and SSB transmission
Since single-sideband is a variant of AM, an SSB signal does have an enve-

lope and must be used with linear amplifiers. The envelope is different from
that for a full-carrier AM signal, however. Figure 2.10 shows a signal with two
modulation frequencies, called a two-tone test signal. Note that the envelope is
caused by the algebraic addition of the two sideband components. Its fre-
quency is that of the difference between the two modulating signal frequen-
cies, in this case 2 kHz.
FIGURE 2.10 Two-tone modulation
' 2.4 Frequency and Phase Modulation

Frequency modulation (FM) is probably the most commonly used analog
modulation technique, seeing application in everything from broadcast-
ing to cordless phones. Phase modulation (PM) is rarely used in analog sys-
tems but is very common in digital communication. Obviously, frequency
and phase are very closely related, so it makes sense to discuss the two
schemes together. In fact, they are often grouped under the heading of angle
modulation.
56 ! CHAPTER 2
In our discussion of amplitude modulation, we found that the ampli-

tude of the modulated signal varied in accordance with the instantaneous
amplitude of the modulating signal. In FM it is the frequency, and in PM
the phase of the modulated signal that varies with the amplitude of the mod-
ulating signal. This is important to remember: in all types of modulation it
is the amplitude, not the frequency, of the baseband signal that does the
modulating.
The amplitude and power of an angle-modulation signal do not change
with modulation. Thus, an FM signal has no envelope. This is actually an
advantage; an FM receiver does not have to respond to amplitude varia-
tions, and this lets it ignore noise to some extent. Similarly, FM equipment
can use nonlinear amplifiers throughout, since amplitude linearity is not
important.
Frequency Figure 2.11 demonstrates the concept of frequency modulation. Although a

Modulation sine wave is mathematically simpler, a square-wave modulating signal is
used in the figure to make the process easier to follow by eye. Figure 2.11(a)
shows the unmodulated carrier and the modulating signal. Figure 2.11(b)
shows the modulated signal in the time domain, as it would appear on an
oscilloscope. The amount of frequency change has been exaggerated for
clarity. The amplitude remains as before, and the frequency changes can
be seen in the changing times between zero crossings for the waveforms. Fig-
ure 2.11(c) of the figure shows how the signal frequency varies with time
in accordance with the amplitude of the modulating signal. Finally, in Fig-
ure 2.11(d) we see how the phase angle varies with time. When the fre-
quency is greater than ƒ c , the phase angle of the signal gradually increases
until it leads that of the carrier, and when the frequency is lower than ƒc, the
signal phase gradually lags that of the carrier.
The maximum amount by which the transmitted frequency shifts in one
direction from the carrier frequency is defined as the deviation. The total
frequency swing is thus twice the deviation. A frequency modulation
index, mƒ, is also defined:
δ
mƒ = (2.19)
ƒm
where
m ƒ = frequency modulation index
δ = peak deviation in hertz
ƒ m = modulating frequency in hertz
The FM modulation index varies with the modulating frequency, unlike

the case for AM. This choice of a definition for mƒ causes the modulation in-
dex to be equal to the peak phase deviation in radians, which is inversely
proportional to the modulating frequency. The modulation index for phase
modulation is also defined as the peak phase deviation.
FIGURE 2.11 Frequency modulation

58 ! CHAPTER 2
EXAMPLE 2.9 Y
A cell phone transmitter has a maximum frequency deviation of 12 kHz.
Calculate the modulation index if it operates at maximum deviation with a
voice frequency of
(a) 300 Hz
(b) 2500 Hz
SOLUTION
(a) δ
mƒ =
ƒm
12 kHz
=
300 Hz
= 40
(b) δ
mƒ =
ƒm
12 kHz
=
2500 Hz
= 4.8
X
Note that there is no requirement for the FM (or PM) modulation index
to be less than 1. When FM modulation is expressed as a percentage, it is the
deviation as a percentage of the maximum allowed deviation that is being
stated.
The Angle Frequency modulation produces an infinite number of sidebands, even for
Modulation single-tone modulation. These sidebands are separated from the carrier by
Spectrum multiples of ƒm, but their amplitude tends to decrease as their distance from
the carrier frequency increases. Sidebands with amplitude less than about
1% of the total signal voltage can usually be ignored; for practical purposes
an angle-modulated signal can be considered to be band-limited. In most
cases, though, its bandwidth is much larger than that of an AM signal.
Bessel Functions The equation for modulation of a carrier with amplitude A and radian fre-
quency ωc by a single-frequency sinusoid is of the form
v(t) = A sin (ω c t + m ƒ sin ω m t) (2.20)

This equation cannot be simplified by ordinary trigonometry, as is the

case for amplitude modulation. About the only useful information that can
be gained by inspection is the fact that the signal amplitude remains con-
stant regardless of the modulation index. This observation is significant,
since it demonstrates one of the major differences between AM and FM or
PM, but it provides no information about the sidebands.
This signal can be expressed as a series of sinusoids by using Bessel func-
tions of the first kind. Proving this is beyond the scope of this text, but it
can be done. The Bessel functions themselves are rather tedious to evalu-
ate numerically, but that, too, has been done. Some results are presented in
Figure 2.12 and Table 2.1 shown on page 60. Bessel functions are equally
valid for FM and PM systems, since the modulation index is equal to the
peak phase deviation, in radians, for both techniques.
FIGURE 2.12
Bessel Functions
The table and graph of Bessel functions represent normalized voltages

for the various frequency components of an FM signal. That is, the numbers
in the tables will represent actual voltages if the unmodulated carrier has an
amplitude of one volt. J0 represents the component at the carrier frequency.
J1 represents each of the first pair of sidebands, at frequencies of ƒc + ƒm and
ƒc − ƒm. J2 represents the amplitude of each of the second pair of sidebands,
which are separated from the carrier frequency by twice the modulating
frequency, and so on. Figure 2.13 shows this on a frequency-domain plot.
All of the Bessel terms should be multiplied by the voltage of the un-
modulated carrier to find the actual sideband amplitudes. Of course, the
60
TABLE 2.1 Bessel Functions
m J0 J1 J2 J3 J4 J5 J6 J7 J8 J9 J10 J11 J12 J13 J14 J15 J16 J17 J18 J19 J20
0 1.00
0.25 0.98 0.12
0.5 0.94 0.24 0.03
0.75 0.86 0.35 0.07 0.01
1 0.77 0.44 0.11 0.02
1.25 0.65 0.51 0.17 0.04 0.01
1.5 0.51 0.56 0.23 0.06 0.01
1.75 0.37 0.58 0.29 0.09 0.02
2 0.22 0.58 0.35 0.13 0.03 0.01
2.25 0.08 0.55 0.40 0.17 0.05 0.01
2.4 0.00 0.52 0.43 0.20 0.06 0.02
2.5 −0.05 0.50 0.45 0.22 0.07 0.02
2.75 −0.16 0.43 0.47 0.26 0.10 0.03 0.01
3 −0.26 0.34 0.49 0.31 0.13 0.04 0.01
3.5 −0.38 0.14 0.46 0.39 0.20 0.08 0.03 0.01
4 −0.40 −0.07 0.36 0.43 0.28 0.13 0.05 0.01
4.5 −0.32 −0.23 0.22 0.42 0.35 0.20 0.08 0.03 0.01
5 −0.18 −0.33 0.05 0.36 0.39 0.26 0.13 0.05 0.02 0.01
5.5 0.00 −0.34 −0.12 0.26 0.40 0.32 0.19 0.09 0.03 0.01
6 0.15 −0.28 −0.24 0.11 0.36 0.36 0.25 0.13 0.06 0.02 0.01
6.5 0.26 −0.15 −0.31 −0.03 0.28 0.37 0.30 0.18 0.09 0.04 0.01
7 0.30 −0.01 −0.30 −0.17 0.16 0.35 0.34 0.23 0.13 0.06 0.02 0.01
7.5 0.27 0.14 −0.23 −0.26 0.02 0.28 0.35 0.28 0.17 0.09 0.04 0.01 0.01
8 0.17 0.24 −0.11 −0.29 −0.11 0.19 0.34 0.32 0.22 0.13 0.06 0.03 0.01
8.5 0.04 0.27 0.02 −0.26 −0.21 0.07 0.29 0.34 0.27 0.17 0.09 0.04 0.02 0.01
8.65 0.00 0.27 0.06 −0.24 −0.23 0.03 0.27 0.34 0.28 0.18 0.10 0.05 0.02 0.01
9 −0.09 0.25 0.14 −0.18 −0.27 −0.06 0.20 0.33 0.30 0.21 0.13 0.06 0.03 0.01
10 −0.25 0.04 0.26 0.06 −0.22 −0.23 −0.01 0.22 0.32 0.29 0.21 0.12 0.06 0.03 0.01
11 −0.17 −0.18 0.14 0.23 −0.01 −0.24 −0.20 0.02 0.23 0.31 0.28 0.20 0.12 0.06 0.03 0.01 0.01
12 0.05 −0.22 −0.08 0.20 0.18 −0.07 −0.24 −0.17 0.04 0.23 0.30 0.27 0.20 0.12 0.07 0.03 0.01 0.10
13 0.21 −0.07 −0.22 0.00 0.22 0.13 −0.12 −0.24 −0.14 0.07 0.23 0.29 0.26 0.19 0.12 0.07 0.03 0.01 0.01
14 0.17 0.13 −0.15 −0.18 0.08 −0.15 −0.23 −0.11 0.08 0.24 0.29 0.25 0.19 0.12 0.07 0.03 0.02 0.01
15 −0.01 0.20 0.04 −0.19 −0.12 0.13 0.21 0.03 −0.17 −0.22 −0.09 0.10 0.24 0.28 0.25 0.18 0.12 0.07 0.03 0.02 0.01
16 −0.17 0.09 0.19 −0.04 −0.20 −0.06 0.17 0.18 −0.01 −0.19 −0.21 −0.07 0.11 0.24 0.27 0.24 0.18 0.11 0.07 0.03 0.02
17 −0.17 −0.10 0.16 0.14 −0.11 −0.19 0.00 0.19 0.15 −0.04 −0.20 −0.19 −0.05 0.12 0.24 0.27 0.23 0.17 0.11 0.07 0.04
FIGURE 2.13 FM in the frequency domain
Bessel coefficients are equally valid for peak or RMS voltages, but the user
should be careful to keep track of which type of measurement is being used.
When Bessel functions are used, the signal of Equation (2.20) becomes
v(t) = A sin (ω c t + m ƒ sin ω m t)

= A {J 0 (m ƒ ) sin ω c t − J 1 (m ƒ )[sin (ω c − ω m )t − sin (ω c + ω m )t]
+ J 2 (m ƒ )[sin (ω c − 2ω m )t + sin (ω c + 2ω m )t]
− J 3 (m ƒ )[sin (ω c − 3ω m )t + sin (ω c + 3 ω m )t]
+ ⋅ ⋅ ⋅} (2.21)
With angle modulation, the total signal voltage and power do not
change with modulation. Therefore, the appearance of power in the side-
bands indicates that the power at the carrier frequency must be reduced be-
low its unmodulated value in the presence of modulation. In fact, the
carrier-frequency component disappears for certain values of mƒ (for exam-
ple, 2.4 and 5.5).
This constant-power aspect of angle modulation can be demonstrated
using the table of Bessel functions. For simplicity, normalized values can be
used. Let the unmodulated signal have a voltage of one volt RMS across a re-
sistance of one ohm. Its power is, of course, one watt. When modulation is
applied, the carrier voltage will be reduced and sidebands will appear. J0 from
62 ! CHAPTER 2
the table will represent the RMS voltage at the carrier frequency and the
power at the carrier frequency will be
Vc2
Pc =
R
J 02
=
1
= J 02
Similarly, the power in each of the first pair of sidebands will be

PSB1 = J12
The combined power in the first set of sidebands will be twice as much as
this, of course. The power in the whole signal will then be
PT = J 02 + 2 J12 + 2 J 22 + ⋅ ⋅ ⋅
If the series is carried on far enough, the result will be equal to one watt,
regardless of the modulation index.
The bandwidth of an FM or PM signal is to some extent a matter of defi-
nition. The Bessel series is infinite, but as can be seen from the table or the
graph, the amplitude of the components will gradually diminish until at
some point they can be ignored. The process is slower for large values of m,
so the number of sets of sidebands that has to be considered is greater for
larger modulation indices. A practical rule of thumb is to ignore sidebands
with a Bessel coefficient of less than 0.01. The bandwidth, for practical pur-
poses, is equal to twice the number of the highest significant Bessel coeffi-
cient, multiplied by the modulating frequency.
EXAMPLE 2.10 Y
An FM signal has a deviation of 3 kHz and a modulating frequency of 1 kHz.
Its total power is 5 W, developed across a 50 Ω resistive load. The carrier fre-
quency is 160 MHz.
(a) Calculate the RMS signal voltage.
(b) Calculate the RMS voltage at the carrier frequency and each of the first
three sets of sidebands.
(c) Calculate the frequency of each sideband for the first three sideband
pairs.
(d) Calculate the power at the carrier frequency, and in each sideband, for
the first three pairs.
(e) Determine what percentage of the total signal power is unaccounted for
by the components described above.
(f) Sketch the signal in the frequency domain, as it would appear on a spec-
trum analyzer. The vertical scale should be power in dBm, and the hori-
zontal scale should be frequency.
SOLUTION
(a) The signal power does not change with modulation, and neither does
the voltage, which can easily be found from the power equation.
VT2
PT =
RL
VT = PT RL
= 5 W × 50Ω
= 15.8 V(RMS)
(b) The modulation index must be found in order to use Bessel functions to
find the carrier and sideband voltages.
δ
mƒ =
ƒm
3 kHz
=
1 kHz
= 3
From the Bessel function table, the coefficients for the carrier and the
first three sideband pairs are:
J0 = −0.26 J1 = 0.34 J2 = 0.49 J3 = 0.31
These are normalized voltages, so they will have to be multiplied by the to-
tal RMS signal voltage to get the RMS sideband and carrier-frequency voltages.
For the carrier,
Vc = J0VT
J0 has a negative sign. This simply indicates a phase relationship between
the components of the signal. It would be required if we wanted to add
together all the components to get the resultant signal. For our present
purpose, however, it can be ignored, and we can use
Vc = J 0 VT
= 0.26 × 15.8 V
= 4.11 V
64 ! CHAPTER 2
Similarly we can find the voltage for each of the three sideband pairs.
Note that these are voltages for individual components. There will be a lower
and an upper sideband with each of these calculated voltages.
V1 = J1VT
= 0.34 × 15.8 V
= 5.37 V
V2 = J2VT
= 0.49 × 15.8 V
= 7.74 V
V3 = J3VT
= 0.31 × 15.8 V
= 4.9 V
(c) The sidebands are separated from the carrier frequency by multiples of
the modulating frequency. Here, ƒc = 160 MHz and ƒm = 1 kHz, so there
are sidebands at each of the following frequencies.
ƒUSB1 = 160 MHz + 1 kHz = 160.001 MHz
ƒUSB2 = 160 MHz + 2 kHz = 160.002 MHz
ƒUSB3 = 160 MHz + 3 kHz = 160.003 MHz
ƒLSB1 = 160 MHz − 1 kHz = 159.999 MHz
ƒLSB2 = 160 MHz − 2 kHz = 159.998 MHz
ƒLSB3 = 160 MHz − 3 kHz = 159.997 MHz
(d) Since each of the components of the signal is a sinusoid, the usual equa-
tion can be used to calculate power. All the components appear across
the same 50 Ω load.
Vc2
Pc =
RL
4.112
=
50
= 0.338 W
Similarly, it can be shown that
P1 = 0.576 W P2 = 1.2 W P3 = 0.48 W
(e) To find the total power in the carrier and the first three sets of sidebands, it
is only necessary to add the powers calculated above, counting each of the
sideband powers twice, because each of the calculated powers represents
one of a pair of sidebands. We only count the carrier once, of course.
PT = Pc + 2(P1 + P2 + P3)
= 0.338 + 2(0.576 + 1.2 + 0.48) W
= 4.85 W
This is not quite the total signal power, which was given as 5 W. The re-
mainder is in the additional sidebands. To find how much is unaccounted
for by the carrier and the first three sets of sidebands, we can subtract. Call
the difference Px.
Px = 5 − 4.85 = 0.15 W
As a percentage of the total power this is

015
.
Px (%) = × 100
5
= 3%
All the information we need for the sketch is on hand, except that the
power values have to be converted to dBm using the equation
P
P(dBm) = 10 log
1 mW
This gives
Pc(dBm) = 10 log 338 = 25.3 dBm
P1(dBm) = 10 log 576 = 27.6 dBm
P2(dBm) = 10 log 1200 = 30.8 dBm
P3(dBm) = 10 log 480 = 26.8 dBm
The sketch is shown in Figure 2.14.
FIGURE 2.14
X
66 ! CHAPTER 2
Bandwidth For PM, the bandwidth varies directly with the modulating frequency, since
doubling the frequency doubles the distance between sidebands. It is also
roughly proportional to the maximum phase deviation, since increasing mp
increases the number of sidebands. For FM, however, the situation is compli-
cated by the fact that
δ
mƒ =
ƒm
For a given amount of frequency deviation, the modulation index is in-

versely proportional to the modulating frequency. Recall that the frequency
deviation is proportional to the amplitude of the modulating signal. Then, if
the amplitude of the modulating signal remains constant, increasing its fre-
quency reduces the modulation index. Reducing mƒ reduces the number of
sidebands with significant amplitude. On the other hand, the increase in ƒm
means that the sidebands will be further apart in frequency, since they are
separated from each other by ƒm. These two effects are in opposite directions.
The result is that the bandwidth does increase somewhat with increasing
modulating-signal frequency, but the bandwidth is not directly propor-
tional to the frequency. Sometimes FM is called a “constant-bandwidth”
communication mode for this reason, though the bandwidth is not really
constant. Figure 2.15 provides a few examples that show the relationship be-
tween modulating frequency and bandwidth. For this example the deviation
remains constant at 10 kHz as the modulating frequency varies from 2 kHz
to 10 kHz.
One other point must be made about the sidebands. With AM, restrict-
ing the bandwidth of the receiver has a very simple effect on the signal.
Since the side frequencies farthest from the carrier contain the high-
frequency baseband information, restricting the receiver bandwidth reduces
its response to high-frequency baseband signals, leaving all else unchanged.
When reception conditions are poor, bandwidth can be restricted to the
minimum necessary for intelligibility. For FM the situation is more compli-
cated, since even low-frequency modulating signals can generate sidebands
that are far removed from the carrier frequency. FM receivers must be de-
signed to include all the significant sidebands that are transmitted; other-
wise severe distortion, not just limited frequency response, will result.
Carson’s Rule The calculation of the bandwidth of an FM signal from Bessel functions is
easy enough, since the functions are available in a table, but it can be a bit te-
dious. There is an approximation, known as Carson’s rule, that can be used
to find the bandwidth of an FM signal. It is not as accurate as using Bessel
functions, but can be applied almost instantly, without using tables or even
a calculator.
FIGURE 2.15 Variation of FM bandwidth with modulating frequency

68 ! CHAPTER 2
Here is Carson’s rule:
B ≅ 2[δ (max) + ƒ m(max) ] (2.22)

Equation (2.22) assumes that the bandwidth is proportional to the sum
of the deviation and the modulating frequency. This is not strictly true. Car-
son’s rule also makes the assumption that maximum deviation occurs with
the maximum modulating frequency. Sometimes this leads to errors in prac-
tical situations, where often the highest baseband frequencies have much
less amplitude than lower frequencies, and therefore do not produce as
much deviation.
EXAMPLE 2.11 Y
Use Carson’s rule to calculate the bandwidth of the signal used in Example
2.10.
SOLUTION
Here there is only one modulating frequency, so
B ≅ 2(δ + fm)
= 2(3 kHz + 1 kHz)
= 8 kHz
In the previous example we found that 97% of the power was contained
in a bandwidth of 6 kHz. An 8-kHz bandwidth would contain more of the
signal power. Carson’s rule gives quite reasonable results in this case, with
very little work.
X
Narrowband and We mentioned earlier that there are no theoretical limits to the modulation
Wideband FM index or the frequency deviation of an FM signal. The limits are practical
and result from a compromise between signal-to-noise ratio and bandwidth.
In general, larger values of deviation result in an increased signal-to-noise
ratio, while also resulting in greater bandwidth. The former is desirable, but
the latter is not, especially in regions of the spectrum where frequency space
is in short supply. It is also necessary to have some agreement about devia-
tion, since receivers must be designed for a particular signal bandwidth.
For these reasons, the bandwidth of FM transmissions is generally lim-
ited by government regulations that specify the maximum frequency devia-
tion and the maximum modulating frequency, since both of these affect
bandwidth. In general, relatively narrow bandwidth (on the order of 10 to
30 kHz) is used for voice communication, with wider bandwidths for such
services as FM broadcasting (about 200 kHz) and satellite television (36 MHz
for one system).
FM and Noise The original reason for developing FM was to give improved performance in
the presence of noise, and that is still one of its main advantages over AM.
This improved noise performance can actually result in a better signal-
to-noise ratio at the output of a receiver than is found at its input.
One way to approach the problem of FM and noise is to think of the
noise voltage as a phasor having random amplitude and phase angle. The
noise adds to the signal, causing random variations in both the amplitude
and phase angle of the signal as seen by the receiver. Figure 2.16 shows this
vector addition.
FIGURE 2.16
Effect of noise on an
FM signal
The amplitude component of noise is easily dealt with in a well-designed

FM system. Since FM signals do not depend on an envelope for detection,
the receiver can employ limiting to remove any amplitude variations from
the signal. That is, it can use amplifiers whose output amplitude is the same
for a wide variety of input signal levels. The effect of this on the amplitude
of a noisy signal is shown in Figure 2.17. As long as the signal amplitude is
FIGURE 2.17 Limiting

70 ! CHAPTER 2
considerably larger than the noise to begin with, the amplitude component
of the noise will not be a problem.
It is not possible for the receiver to ignore phase shifts, however. A PM re-
ceiver obviously must respond to phase changes, but so will an FM receiver
because, as we have seen, phase shifts and frequency shifts always occur to-
gether. Therefore, phase shifts due to noise are associated with frequency
shifts that will be interpreted by the receiver as part of the modulation.
Figure 2.18 shows the situation at the input to the receiver. The circle
represents the fact that the noise phasor has a constantly changing angle
with respect to the signal. Its greatest effect, and thus the peak phase shift to
the signal, will occur when the noise phasor is perpendicular to the resul-
tant. At that time, the phase shift due to noise is
E 
φ N = sin −1 N  (2.23)
 ES 
FIGURE 2.18
Phase shift due to
noise
(EN/ES) is the reciprocal of the voltage signal-to-noise ratio at the input. A

little care is needed here, as S/N is usually given as a power ratio, in decibels,
and will have to be converted to a voltage ratio before being used in Equa-
tion (2.23).
Equation (2.23) can be simplified as long as the signal is much larger
than the noise. This will cause the phase deviation to be small, and for small
angles the sine of the angle is approximately equal to the angle itself, in radi-
ans. Thus in a practical situation we can use
EN
φN ≈ (2.24)
ES
The phase shift due to noise can be reduced by making the signal volt-
age, relative to the noise voltage, as large as possible. This requires increased
transmission power, a better receiver noise figure, or both. Perhaps less obvi-
ous is the fact that the relative importance of phase shifts due to noise can be
reduced by having the phase shifts in the signal as large as possible. This is
accomplished by keeping the value of mƒ high, since mƒ represents the peak
phase shift in radians. It would seem that the ratio of signal voltage to noise
voltage at the output would be proportional to mƒ, and this is approximately
true under strong-signal conditions.
EXAMPLE 2.12 Y
An FM signal has a frequency deviation of 5 kHz and a modulating frequency of
1 kHz. The signal-to-noise ratio at the input to the receiver detector is 20 dB.
Calculate the approximate signal-to-noise ratio at the detector output.
SOLUTION
First, notice the word “approximate.” Our analysis is obviously a little sim-
plistic, since noise exists at more than one frequency. We are also going to as-
sume that the detector is completely unresponsive to amplitude variations
and that it adds no noise of its own. Our results will not be precise but they
will show the process that is involved.
First, let us convert 20 dB to a voltage ratio.
ES ( S/N )(dB )
= log −1
EN 20
20
= log −1
20
= 10
EN 1
=
ES 10
= 01
.
Since ES >> EN, we can use Equation (2.24).
EN
φN ≈
ES
= 01
. rad
Remembering that the receiver will interpret the noise as an FM signal
with a modulation index equal to φN, we find
mƒ N = 0.1
This can be converted into an equivalent frequency deviation δN due to

the noise.
δN = mƒƒm
= 0.1 × 1 kHz
= 100 Hz
72 ! CHAPTER 2
The frequency deviation due to the signal is given as 5 kHz, and the re-
ceiver output voltage is proportional to the deviation. Therefore, the output
S/N as a voltage ratio will be equal to the ratio between the deviation due to
the signal and that due to the noise.
 ES  δS
  =
 EN  δN
5 kHz
=
100 Hz
= 50
Since S/N is nearly always expressed in decibels, change this to dB.
(S/N)o (dB) = 20 log 50

= 34 dB
This is an improvement of 14 dB over the S/N at the input.

X
Threshold Effect An FM signal can produce a better signal-to-noise ratio at the output of a
and Capture receiver than an AM signal with a similar input S/N, but this is not always
Effect the case. The superior noise performance of FM depends on there being a
sufficient input S/N ratio. There exists a threshold S/N below which the per-
formance is no better than AM. In fact, it is worse, because the greater band-
width of the FM signal requires a wider receiver noise bandwidth. When the
signal strength is above the threshold, the improvement in noise perfor-
mance for FM can be more than 20 dB compared with AM.
The noise-rejection characteristic of FM applies equally well to interfer-
ence. As long as the desired signal is considerably stronger than the interfer-
ence, the ratio of desired to interfering signal strength will be greater at the
output of the detector than at the input. We could say that the stronger sig-
nal “captures” the receiver, and in fact this property of FM is usually called
the capture effect. It is very easy to demonstrate with any FM system. For ex-
ample, it is the reason that there is less interference between cordless tele-
phones, which share a few channels in the 46- and 49-MHz bands, than one
might expect.
Pre-emphasis and An FM receiver interprets the phase shifts due to noise as frequency modula-
De-emphasis tion. Phase and frequency deviation are related by Equation (2.19):
δ
mƒ =
ƒm
which can be restated
δ = mƒ ƒm
The modulation index mƒ is simply the peak phase deviation in radians.

The frequency deviation is proportional to the modulating frequency. This
tells us, if the phase deviation due to thermal noise is randomly distributed
over the baseband spectrum, the amplitude of the demodulated noise will
be proportional to frequency. This relationship between noise voltage
and frequency is shown in Figure 2.19. Since power is proportional to the
square of voltage, the noise power will have the parabolic spectrum shown
in Figure 2.19. An improvement in S/N can be made by boosting (pre-
emphasizing) the higher frequencies in the baseband signal before modula-
tion, with a corresponding cut in the receiver after demodulation. Obviously
it is necessary to use similar filter characteristics for pre- and de-emphasis.
Usually simple first-order filters are used.
FIGURE 2.19
Spectrum of demodulated
noise
Note that pre-emphasis and de-emphasis are unnecessary with phase

modulation. Since the phase deviation due to noise is converted directly
into baseband noise output without the intermediate step of conversion
into an equivalent frequency deviation, the output noise will have a flat
spectrum, assuming thermal noise at the input to the demodulator.
74 ! CHAPTER 2
( In the time domain, the process of amplitude modulation creates a signal
with an envelope that closely resembles the original information signal.
( In the frequency domain, an amplitude-modulated signal consists of the
carrier, which is unchanged from its unmodulated state, and two side-
bands. The total bandwidth of the signal is twice the maximum modulat-
ing frequency.
( An amplitude-modulated signal can be demodulated by an envelope de-
tector, which consists of a diode followed by a lowpass filter.
( The peak voltage of an amplitude-modulated signal varies with the modu-
lation index, becoming twice that of the unmodulated carrier for the
maximum modulation index of 1.
( The power in an amplitude-modulated signal increases with modulation.
The extra power goes into the sidebands. At maximum modulation, the
total power is 50% greater than the power in the unmodulated carrier.
( Angle modulation includes frequency and phase modulation, which are
closely related.
( Frequency modulation is widely used for analog communication, while
phase modulation sees greatest application in data communication.
( The power of an angle-modulation signal does not change with modula-
tion, but the bandwidth increases due to the generation of multiple sets
of sidebands.
( The voltage and power of each sideband can be calculated using Bessel
functions. An approximate bandwidth is given by Carson’s rule.
( Frequency modulation has a significant advantage compared with AM in
the presence of noise or interference, provided the deviation is relatively
large and the signal is reasonably strong.
( The signal-to-noise ratio for FM can be improved considerably by using
pre-emphasis and de-emphasis. This involves greater gain for the higher
baseband frequencies before modulation, with a corresponding reduc-
tion after demodulation.
( Equation List
v(t) = (Ec + Em sin ωmt) sin ωct (2.2)
m = Em/Ec (2.3)
v(t) = Ec(1 + m sin ωmt) sin ωct (2.4)

mT = m12 + m22 + ⋅ ⋅ ⋅ (2.5)
Emax = Ec(1 + m) (2.6)
Emin = Ec(1 − m) (2.7)
E max − E min
m = (2.8)
E max + E min
mE c mE c
v (t ) = E c sin ω ct + cos(ω c − ω m )t − cos(ω c + ω m )t (2.9)
2 2
ƒusb = ƒc + ƒm (2.10)
ƒlsb = ƒc − ƒm (2.11)
mE c
E lsb = E usb = (2.12)
2
B = 2fm (2.13)
m2
Plsb = Pusb = Pc (2.14)
4
m2
PSB = Pc (2.15)
2
 m2 
Pt = Pc  1 +  (2.16)
 2 
Plsb
m = 2 (2.18)
Pc
δ
mƒ = (2.19)
ƒm
v(t) = A sin (ωct + mƒ sin ωmt)

= A {J0(mƒ) sin ωct − J1(mƒ)[sin (ωc − ωm)t − sin (ωc + ωm)t]
+ J2(mƒ) [sin (ωc − 2ωm)t + sin (ωc + 2ωm)t]
− J3(mƒ)[sin (ωc − 3ωm)t + sin (ωc + 3 ωm)t]
+ ⋅ ⋅ ⋅} (2.21)
76 ! CHAPTER 2
B ≅ 2[δ(max) + ƒm(max)] (2.22)
EN
φN ≈ (2.24)
ES
( Key Terms
angle modulation term that applies to both frequency modulation (FM)
and phase modulation (PM) of a transmitted signal
capture effect tendency of an FM receiver to receive the strongest signal
and reject others
deviation in FM, the peak amount by which the instantaneous signal
frequency differs from the carrier frequency in each deviation
envelope imaginary pattern formed by connecting the peaks of
individual RF waveforms in an amplitude-modulated signal
frequency modulation modulation scheme in which the transmitted
frequency varies in accordance with the instantaneous amplitude of
the information signal
frequency modulation index peak phase shift in a frequency-modulated
signal, in radians
modulation index number indicating the degree to which a signal is
modulated
overmodulation modulation to an extent greater than that allowed for
either technical or regulatory reasons
phase modulation communication system in which the phase of a high-
frequency carrier is varied according to the amplitude of the baseband
(information) signal
side frequencies frequency components produced above and below the
carrier frequency by the process of modulation
sideband a group of side frequencies above or below the carrier
frequency
splatter frequency components produced by a transmitter that fall
outside its assigned channel
( Questions
1. What is meant by the “envelope” of an AM waveform, and what is its
significance?
2. Although amplitude modulation certainly involves changing the am-

plitude of the signal; it is not true to say that the amplitude of the carrier
is modulated. Explain this statement.
3. Why is it desirable to have the modulation index of an AM signal as
large as possible, without overmodulating?
4. Describe what happens when a typical AM modulator is over-
modulated, and explain why overmodulation is undesirable.
5. How does the bandwidth of an AM signal relate to the information
signal?
6. Describe two ways in which the modulation index of an AM signal can
be measured.
7. By how much does the power in an AM signal increase with modula-
tion, compared to the power of the unmodulated carrier?
8. What two types of modulation are included in the term “angle modu-
lation”?
9. Compare, in general terms, the bandwidth and signal-to-noise ratio of
FM and AM.
10. Describe and compare two ways to determine the practical bandwidth
of an FM signal.
11. What is pre-emphasis and how is it used to improve the signal-to-noise
ratio of FM transmissions?
12. For FM, what characteristic of the modulating signal determines the in-
stantaneous frequency deviation?
13. What is the capture effect?
14. Where is phase modulation used?
15. Explain why the signal-to-noise ratio of FM can increase with the band-
width. Is this always true for FM? Compare with the situation for AM.
16. Compare the effects of modulation on the carrier power and the total
signal power in FM and AM.
17. What is the threshold effect?
18. Explain how limiting reduces the effect of noise on FM signals.
19. Explain how noise affects FM signals even after limiting.
20. Explain the fact that there is no simple relationship between modulat-
ing frequency and bandwidth for an FM signal.
21. Why does limiting the receiver bandwidth to less than the signal band-
width cause more problems with FM than with AM?
78 ! CHAPTER 2
( Problems
1. An AM signal has the equation:
v(t) = (15 + 4 sin 44 × 103t) sin 46.5 × 106t volts

(a) Find the carrier frequency.
(b) Find the frequency of the modulating signal.
(c) Find the value of m.
(d) Find the peak voltage of the unmodulated carrier.
(e) Sketch the signal in the time domain showing voltage and time
scales.
2. An AM signal has a carrier frequency of 3 MHz and an amplitude of 5 V
peak. It is modulated by a sine wave with a frequency of 500 Hz and a
peak voltage of 2 V. Write the equation for this signal and calculate the
modulation index.
3. An AM signal consists of a 10-MHz carrier modulated by a 5-kHz sine
wave. It has a maximum positive envelope voltage of 12 V and a mini-
mum of 4 V.
(a) Find the peak voltage of the unmodulated carrier.
(b) Find the modulation index and percent.
(c) Sketch the envelope.
(d) Write the equation for the signal voltage as a function of time.
4. An AM transmitter is modulated by two sine waves, at 1 kHz and
2.5 kHz, with a modulation due to each of 25% and 50% respectively.
What is the effective modulation index?
5. For the AM signal sketched in Figure 2.20, calculate:
(a) the modulation index
(b) the peak carrier voltage
(c) the peak modulating-signal voltage
6. For the signal of Figure 2.21, calculate:
(a) the index of modulation
(b) the RMS voltage of the carrier without modulation
7. An audio system requires a frequency response from 50 Hz to15 kHz for
high fidelity. If this signal were transmitted using AM, what bandwidth
would it require?
8. A transmitter operates with a carrier frequency of 7.2 MHz. It is ampli-
tude modulated by two tones with frequencies of 1500 Hz and 3000 Hz.
What frequencies are produced at its output?
FIGURE 2.20 FIGURE 2.21
9. Sketch the signal whose equation is given in Problem 1 in the frequency

domain, showing frequency and voltage scales.
10. An AM signal has the following characteristics:
ƒc = 150 MHz ƒm = 3 kHz Ec = 50 V Em = 40 V

For this signal, find:
(a) the modulation index
(b) the bandwidth
(c) the peak voltage of the upper sideband
11. An AM signal observed on a spectrum analyzer shows a carrier at +12
dBm, with each of the sidebands 8 dB below the carrier. Calculate:
(a) the carrier power in milliwatts
(b) the modulation index
12. An AM transmitter with a carrier power of 10 W at a frequency of 25 MHz
operates into a 50-Ω load. It is modulated at 60% by a 2-kHz sine wave.
(a) Sketch the signal in the frequency domain. Show power and fre-
quency scales. The power scale should be in dBm.
(b) What is the total signal power?
(c) What is the RMS voltage of the signal?
80 ! CHAPTER 2
13. A 5-MHz carrier is modulated by a 5-kHz sine wave. Sketch the result in
both frequency and time domains for each of the following types of
modulation. Time and frequency scales are required, but amplitude
scales are not.
(a) DSB full-carrier AM
(b) DSBSC AM
(c) SSBSC AM (USB)
14. If a transmitter power of 100 W is sufficient for reliable communication

over a certain path using SSB, approximately what power level would be
required using:
(a) DSBSC
(b) full-carrier AM
15. An AM transmitter has a carrier power of 50 W at a carrier frequency of

12 MHz. It is modulated at 80% by a 1-kHz sine wave.
(a) How much power is contained in the sidebands?
(b) Suppose the transmitter in part (a) can also be used to transmit a
USB signal with an average power level of 50 watts. By how much
(in dB) will the signal-to-noise ratio be improved when the trans-
mitter is used in this way, compared with the situation in part (a)?
16. An FM signal has a deviation of 10 kHz and a modulating frequency of 2

kHz. Calculate the modulation index.
17. Calculate the frequency deviation for an FM signal with a modulating

frequency of 5 kHz and a modulation index of 2.
18. A sine-wave carrier at 100 MHz is modulated by a 1-kHz sine wave. The
deviation is 100 kHz. Draw a graph showing the variation of instanta-
neous modulated signal frequency with time.
19. A phase-modulated signal has a modulation index of 2 with a modulat-

ing signal having an amplitude of 100 mV and a frequency of 4 kHz.
What would be the effect on the modulation index of:
(a) changing the frequency to 5 kHz
(b) changing the voltage to 200 mV
20. An FM signal has a deviation of 10 kHz and is modulated by a sine wave

a frequency of 5 kHz. The carrier frequency is 150 MHz, and the signal
has a total power of 12.5 W, operating into an impedance of 50 Ω.
(a) What is the modulation index?
(b) How much power is present at the carrier frequency?
(c) What is the voltage level of the second sideband below the carrier
frequency?
(d) What is the bandwidth of the signal, ignoring all components
which have less than 1% of the total signal voltage?
21. An FM transmitter operates with a total power of 10 watts, a deviation of
5 kHz, and a modulation index of 2.
(a) What is the modulating frequency?
(b) How much power is transmitted at the carrier frequency?
(c) If a receiver has a bandwidth sufficient to include the carrier and the
first two sets of sidebands, what percentage of the total signal power
will it receive?
22. An FM transmitter has a carrier frequency of 220 MHz. Its modulation
index is 3 with a modulating frequency of 5 kHz. The total power output
is 100 watts into a 50 Ω load.
(a) What is the deviation?
(b) Sketch the spectrum of this signal, including all sidebands with
more than 1% of the signal voltage.
(c) What is the bandwidth of this signal according to the criterion used
in part (b)?
(d) Use Carson’s rule to calculate the bandwidth of this signal, and
compare with the result found in part (c).
23. An FM transmitter has a carrier frequency of 160 MHz. The deviation is
10 kHz and the modulation frequency is 2 kHz. A spectrum analyzer
shows that the carrier-frequency component of the signal has a power
of 5 W. What is the total signal power?
24. Use Carson’s rule to compare the bandwidth that would be required to
transmit a baseband signal with a frequency range from 300 Hz to 3 kHz
using:
(a) narrowband FM with maximum deviation of 5 kHz
(b) wideband FM with maximum deviation of 75 kHz
25. An FM receiver operates with a signal-to-noise ratio of 30 dB at its detec-
tor input and is operating with mƒ = 10.
(a) If the received signal has a voltage of 10 mV, what is the amplitude
of the noise voltage?
(b) Find the maximum phase shift that could be given to the signal by
the noise voltage.
(c) Calculate the signal-to-noise ratio at the detector output, assuming
the detector is completely insensitive to amplitude variations.
82 ! CHAPTER 2
26. A certain full-carrier DSB AM signal has a bandwidth of 20 kHz. What

would be the approximate bandwidth required if the same information
signal were to be transmitted using:
(a) DSBSC AM
(b) SSB
(c) FM with 10 kHz deviation
(d) FM with 50 kHz deviation
27. Using the table of Bessel functions, demonstrate that the total power in
an FM signal is equal to the power in the unmodulated carrier for m = 2.
Compare with the situation for full-carrier AM and for SSBSC AM.
28. Suppose you were called upon to recommend a modulation technique
for a new communication system for voice frequencies. State which of
the techniques studied so far you would recommend, and why, in each
of the following situations:
(a) simple, cheap receiver design is of greatest importance
(b) narrow signal bandwidth is of greatest importance
(c) immunity to noise and interference is of greatest importance
Digital Communication 3
Objectives
( Compare analog and digital communication techniques and discuss the
advantages of each.
( Calculate the minimum sampling rate for a signal and explain the necessity
for sampling at that rate or above.
( Find the spurious frequencies produced by aliasing when the sample rate is
too low.
( Describe the common types of analog pulse modulation.
( Describe pulse-code modulation and calculate the number of quantizing
levels, the bit rate, and the dynamic range for PCM systems.
( Explain companding, show how it is accomplished, and explain its effects.
( Describe the coding and decoding of a PCM signal.
( Describe differential PCM and explain its operation and advantages.
( Describe delta modulation and explain the advantages of adaptive delta
modulation.
( Distinguish between lossless and lossy compression and provide examples
of each.
( Describe the operation of common types of vocoders.
84 ! CHAPTER 3
' 3.1 Introduction

In the previous chapter we looked at analog modulation techniques. These
were historically the earliest ways to transmit information by radio, and
they are still very popular. However, recent developments have made digital
methods more important, to the point where it is expected that most new
wireless communication systems will be digital.
In this chapter we will look at the digital encoding of analog signals.
Voice is a good example of such a signal; it begins and ends as analog infor-
mation. Once coded digitally, the analog signal is indistinguishable from
any other data stream, such as a word-processing document. In Chapter 4 we
will look at the transmission of this data by radio.
While the advantage of digital transmission for data is self-evident, it is
not immediately obvious why it is desirable to go to the trouble and expense
of converting analog signals to digital form and back again. There are in fact
several very good reasons for doing so. Here are some of the most important
for wireless communication systems.
( Digital signals can be manipulated more easily than analog signals.
They are easier to multiplex, for instance.
( Digital signals can easily be encrypted to ensure privacy.
( When an analog signal goes through a chain of signal processors,
such as transmitters, receivers, and amplifiers, noise and distortion
accumulate. This process can be made much less severe in digital sys-
tems by regenerating the signal and by using various types of error
control. These concepts will be explained in more detail later in this
chapter.
( Data compression can be used with a digital signal to reduce its band-
width to less than that required to transmit the original analog signal.
Until recently the use of digital techniques was restricted because dig-
ital communication required more bandwidth than analog. That situ-
ation has changed quite radically in recent years.
You might have noticed that “higher fidelity” and “improved frequency
response” do not appear in the above list. It is certainly true that some digi-
tal schemes, such as compact disc audio, have better fidelity than some ana-
log schemes, like FM broadcasts; but this is not always the case. The
distortion and frequency response of a digital audio system are fixed by the
sampling process, which will be described in the next section. The quality of
the reproduction can be predicted quite accurately, and the use of regenera-
tion and error correction can prevent it from deteriorating very much, but
not all digital transmission is of high quality. For instance, whether you
know it or not you have been communicating digitally for years: practically
DIGITAL COMMUNICATION ! 85
all modern telephone switches are digital. Digital telephony sounds pretty
good, for a telephone call, but does not compare with compact disc audio.
We will find out why in the next sections.
' 3.2 Sampling

An analog signal varies continuously with time. If we want to transmit such
a signal digitally, that is, as a series of numbers, we must first sample the sig-
nal. This involves finding its amplitude at discrete time intervals. Only in
this way can we arrive at a finite series of numbers to transmit.
Sampling Rate In 1928, Harry Nyquist showed mathematically that it is possible to recon-
struct a band-limited analog signal from periodic samples, as long as the
sampling rate is at least twice the frequency of the highest-frequency com-
ponent of the signal. This assumes that an ideal low-pass filter prevents
higher frequencies from entering the sampler. Since real filters are not ideal,
in practice the sampling rate must be considerably more than twice the max-
imum frequency to be transmitted.
If the sampling rate is too low, a form of distortion called aliasing
or foldover distortion is produced. In this form of distortion, frequencies
in the sampled signal are translated downward. Figure 3.1 shows what
happens.
In Figure 3.1(a) the sampling rate is adequate and the signal can be re-
constructed. In Figure 3.1(b), however, the rate is too low and the attempt to
reconstruct the original signal results in a lower-frequency output signal.
Once aliasing is present, it cannot be removed.
The frequency of the interference generated by aliasing is easier to see by
looking at the frequency domain. For simplicity, assume that the signal to be
sampled, which we will call the baseband signal, is a sine wave:
e b = E b sin ω b t (3.1)
where
e b = instantaneous baseband signal voltage
E b = peak baseband signal voltage
ω b = the radian frequency of the baseband signal
The sampling process is equivalent to multiplying the baseband signal

by a pulse train at the sampling frequency. See Figure 3.2. For simplicity, as-
sume that the pulses have an amplitude of one volt. Then, when a pulse is
86 ! CHAPTER 3
FIGURE 3.1 Aliasing
present, the output of the sampler will be the same as the baseband signal
amplitude, and when there is no pulse, the sampler output will be zero.
The spectrum for the pulse train in Figure 3.2 is given in Chapter 1 as:
1 1  sin πτ/ T sin 2πτ/ T sin 3πτ/ T 
es = +2  cos ω st + cos 2ω st + cos 3ω st + ⋅ ⋅ ⋅ (3.2)
T T  πτ/ T 2πτ/ T 3πτ/ T 
1 2 sin πτ/ T sin 2πτ/ T 2 sin 3πτ/ T
= + cos ω st + cos 2ω st + cos 3ω st + ⋅ ⋅ ⋅
T πτ πτ 3πτ
where
es = instantaneous voltage of the sampling pulse
τ = pulse duration
T = pulse period
ωs = radian frequency of the pulse train
FIGURE 3.2
Sampling pulses
Multiplying the two signals given in Equations (3.1) and (3.2) together
gives the following output:
Eb 2 E b sin πτ / T
v (t ) = sin ω bt + sin ω bt cos ω s t (3.3)
T πτ
E b sin 2 πτ / T
+ sin ω bt cos 2ω s t
πτ
2 E b sin 3πτ / T
+ sin ω bt cos 3ω s t
3πτ
As is often the case with equations of this type, we do not have to “solve”
anything to understand what is happening. The first term is simply the origi-
nal baseband signal multiplied by a constant. If only this term were present,
the original signal could be recovered from the sampled signal. However, we
need to look at the other terms to see whether they will interfere with the
signal recovery.
The second term contains the product of sin ωbt and cos ωst. Recall the
trigonometric identities:
1
sin A cos B =
2
[sin( A − B) + sin( A + B)] (3.4)
and
sin( − A) = − sin A (3.5)
Equation (3.5) can be used to express Equation (3.4) in a more convenient

form for the problem at hand. Substituting Equation (3.5) into Equation
(3.4) gives:
1
sin A cos B =
2
[sin( B + A) − sin( B − A)] (3.6)
88 ! CHAPTER 3
This new identity can be used to expand the second term of Equation
(3.3) as follows:
2 E b sin πτ / T
sin ω bt cos ω s t (3.7)
πτ
E b sin πτ / T
=
πτ
[sin(ω s + ω b )t − sin(ω s − ω b )t ]
E b sin πτ / T E sin πτ / T
= sin(ω s + ω b )t − b sin(ω s − ω b )t
πτ πτ
Now we can see that this term consists of components at the sum and
difference of the baseband and sampling frequencies. The sum term can
easily be eliminated by a low-pass filter, since its frequency is obviously
much higher than the baseband. The difference term is more interesting. If
ws > 2ωb, then ωs − ωb > ωb and the difference part of this term can also be re-
moved by a low-pass filter, at least in theory. However, if ωs < 2ωb, the differ-
ence will be less than ωb. An aliased component will appear as
ƒa = ƒs − ƒb (3.8)
and low-pass filtering will not be effective in removing it.
The other terms in Equation (3.3) are not interesting here because they
all represent frequencies greater than ωs − ωb. Therefore, if we make sure that
(ωs − ωb) > ωb, these other terms will not be a problem. Let us, then, rewrite
Equation (3.3), including only the first term and the expanded second term:
Eb E sin πτ / T
v (t ) = sin ω bt + b sin(ω s + ω b )t (3.9)
T πτ
E b sin πτ / T
− sin(ω s − ω b )t
πτ
An example will further clarify the problem.
EXAMPLE 3.1 Y
A digital communication system uses sampling at 10 kilosamples per second
(kSa/s). The receiver filters out all frequencies above 5 kHz. What frequen-
cies appear at the receiver for each of the following signal frequencies at the
input to the transmitter?
(a) 1 kHz
(b) 5 kHz
(c) 6 kHz
SOLUTION
(a) The first term in Equation (3.9) is simply the input frequency, which is,
of course, the only one we want to see in the output. The second term
is the sum of the input and the sampling frequencies, and the third is the
difference. In this case, the frequencies generated are:
ƒb = 1 kHz ƒs + ƒb = 11 kHz ƒs − ƒb = 9 kHz
However, only the 1-kHz component passes through the filter and the sys-
tem operates correctly.
(b) ƒb = 5 kHz ƒs + ƒb = 15 kHz ƒs − ƒb = 5 kHz
Again, the system works properly and the 15-kHz component is removed
by the filter, and only the input frequency of 5 kHz appears at the output.
(c) ƒb = 6 kHz ƒs + ƒb = 16 kHz ƒs − ƒb = 4 kHz
Here we have a serious problem. The 16-kHz component is removed by

the filter but the 4-kHz component is not. With 6 kHz at the input we pro-
duce 6 kHz and 4 kHz at the output.
X
Natural and The equations in the previous section assumed that a sample consisted of
Flat-Topped the baseband signal multiplied by a rectangular pulse. To simplify the math-
Sampling ematics, we assumed that the pulse had an amplitude of 1 V. These assump-
tions yield a sample pulse whose shape follows that of the original signal, as
shown in Figure 3.3(a). This technique is called natural sampling.
FIGURE 3.3 Natural and flat-topped sampling

90 ! CHAPTER 3
Practical systems generally sample by using a sample-and-hold circuit,

which maintains the signal level at the start of the sample pulse. The results
of such a method are shown in Figure 3.3(b). This technique is known as
flat-topped sampling. If the samples were to be transmitted directly as ana-
log pulses of different amplitudes, there would be some small difference in
frequency response between the two techniques. However, for digital trans-
mission there is no practical difference between the two sampling methods.
Analog Pulse As previously mentioned, it would be possible to transmit the samples di-
Modulation rectly as analog pulses. This technique, called pulse-amplitude modulation
(PAM), does not offer any great advantage over conventional analog trans-
mission. In current systems, PAM is used as an intermediate step; before be-
ing transmitted, the PAM signal is digitized. Similarly, at the receiver, the
digital signal is converted back to PAM as part of the demodulation process.
The original signal can then be recovered using a low-pass filter.
Some improvement in noise performance can be made by transmitting
pulses of equal amplitude but variable length (with the duration of the
pulses corresponding to the amplitude of the samples). This technique is
called pulse-duration modulation (PDM) or pulse-width modulation
(PWM). PDM has uses in communication, in some telemetry systems for in-
stance; but it is not likely to be seen in modern wireless systems. Similarly, it
is possible to transmit the information signal by using pulses of equal ampli-
tude and duration but changing their timing in accordance with the sample
amplitude. This system, called pulse-position modulation (PPM), is men-
tioned only for completeness, as it is rarely seen. Figure 3.4 shows the basic
nature of all these systems.
' 3.3 Pulse-Code Modulation

Pulse-code modulation (PCM) is the most commonly used digital modula-
tion scheme. In PCM the available range of signal voltages is divided into
levels, and each is assigned a binary number. Each sample is then repre-
sented by the binary number representing the level closest to its amplitude,
and this number is transmitted in serial form. In linear PCM, levels are sepa-
rated by equal voltage gradations.
Quantization and The number of levels available depends on the number of bits used to ex-
Quantizing Noise press the sample value. The number of levels is given by
N = 2m (3.10)
FIGURE 3.4 Analog pulse modulation
where
N = number of levels
m = number of bits per sample
EXAMPLE 3.2 Y
Calculate the number of levels if the number of bits per sample is:
(a) 8 (as used in telephony)
(b) 16 (as used in the compact disc audio system)
SOLUTION
(a) The number of levels with 8 bits per sample is, from Equation (3.10),
N = 2m
= 28
= 256
92 ! CHAPTER 3
(b) The number of levels with 16 bits per sample is, from the same equation,
N = 2m
= 216
= 65536
X
This process is called quantizing. Since the original analog signal can
have an infinite number of signal levels, the quantizing process will produce
errors called quantizing errors or often quantizing noise.
Figure 3.5 shows how quantizing errors arise. The largest possible error is
one-half the difference between levels. Thus the error is proportionately
greater for small signals. This means that the signal-to-noise ratio varies
with the signal level and is greatest for large signals. The level of quantizing
noise can be decreased by increasing the number of levels, which also in-
creases the number of bits that must be used per sample.
The dynamic range of a system is the ratio between the strongest possi-
ble signal that can be transmitted and the weakest discernible signal. For
a linear PCM system, the maximum dynamic range in decibels is given
approximately by
DR = 1.76 + 6.02m dB (3.11)

where
DR = dynamic range in decibels
This equation ignores any noise contributed by the analog portion of
the system.
FIGURE 3.5 Quantizing error

EXAMPLE 3.3 Y
Find the maximum dynamic range for a linear PCM system using 16-bit
quantizing.
SOLUTION
From Equation (3.11)
DR = 1.76 + 6.02m Db
= 1.76 + 6.02 × 16
= 98.08 dB
X
Bit Rate Increasing the number of bits per sample increases the bit rate, which is
given very approximately by
D = ƒsm (3.12)
where
D = data rate in bits per second
ƒ s = sample rate in samples per second
Extra bits are often included to detect and correct errors. A few bits,
called framing bits, are also needed to ensure that the transmitter and re-
ceiver agree on which bits constitute one sample. The actual bit rate will
therefore be somewhat higher than calculated above.
EXAMPLE 3.4 Y
Calculate the minimum data rate needed to transmit audio with a sampling
rate of 40 kHz and 14 bits per sample.
SOLUTION
D = ƒsm
= 40 × 103 × 14
= 560 × 103 b/s
= 560 kb/s
X
94 ! CHAPTER 3
Companding The transmission bandwidth varies directly with the bit rate. In order to
keep the bit rate and thus the required bandwidth low, companding is often
used. This system involves using a compressor amplifier at the transmitter,
with greater gain for low-level than for high-level signals. The compressor
will reduce the quantizing error for small signals. The effect of compression
on the signal can be reversed by using an expander at the receiver, with a
gain characteristic that is the inverse of that at the transmitter.
It is necessary to follow the same standards at both ends of the circuit
so that the dynamics of the output signal are the same as at the input. The sys-
tem used in the North American telephone system uses a characteristic known
as the µ (mu) law, which has the following equation for the compressor:
Vo ln(1 + µv i /Vi )
vo = (3.13)
ln(1 + µ)
where
vo = actual output voltage from the compressor
Vo = maximum output voltage
Vi = the maximum input voltage
vi = the actual input voltage
µ = a parameter that defines the amount of compression
(contemporary systems use µ = 255)
European telephone systems use a similar but not identical scheme
called A-law compression.
Figure 3.6 on page 95 shows the µ-255 curve. The curve is a transfer func-
tion for the compressor, relating input and output levels. It has been normal-
ized, that is, vi/Vi and vo/Vo are plotted, rather than vi and vo.
EXAMPLE 3.5 Y
A signal at the input to a mu-law compressor is positive, with its voltage
one-half the maximum input voltage. What proportion of the maximum
output voltage is produced?
SOLUTION
vo =
ln(1 + µ)
Vo ln(1 + 255 × 05
. )
=
ln(1 + 255)
= 0876
. Vo
This problem can also be solved graphically, using Figure 3.6.

X
FIGURE 3.6
Mu-Law
compression
Digital companding is also possible. The method is to quantize a signal

using a greater number of bits than will be transmitted, and then perform
arithmetic on the samples to reduce the number of bits. This is the way
companding is done in most modern telephone equipment. This type of
companding is part of the coding and decoding process, which is the topic
of the next section.
Coding and The process of converting an analog signal into a PCM signal is called cod-
Decoding ing, and the inverse operation, converting back from digital to analog, is
known as decoding. Both procedures are often accomplished in a single in-
tegrated-circuit device called a codec.
Figure 3.7 is a block diagram showing the steps for converting an analog
signal into a PCM code. The first block is a low-pass filter, required to pre-
vent aliasing. As shown in section 3.2, the filter must block all frequency
components above one-half the sampling rate. This requires a high-order
filter.
96 ! CHAPTER 3
FIGURE 3.7 PCM coding
The next step is to sample the incoming waveform using a sample-

and-hold circuit. There are many such circuits; a simple one is shown in
Figure 3.8. The field-effect transistor Q turns on during the sample time, al-
lowing the capacitor to charge to the amplitude of the incoming signal. The
transistor then turns off, and the capacitor stores the signal value until the
analog-to-digital converter has had time to convert the sample to digital
form. The two operational amplifiers, connected as voltage followers, isolate
the circuit from the other stages. The low output impedance of the first stage
ensures that the capacitor quickly charges or discharges to the value of the
incoming signal when the transistor conducts.
The samples must now be coded as binary numbers. If we are using linear
PCM, all that is required is a standard analog to digital (A/D) converter.
Compression, if required, can be applied to the analog signal, but it is more
common to incorporate the compression into the coding process.
The codecs used in telephony generally accomplish compression by us-
ing a piecewise-linear approximation to the mu-law curve shown earlier in
Figure 3.6. The positive- and negative-going parts of the curve are each di-
vided into seven segments, with an additional segment centered around
zero, resulting in a total of fifteen segments. Figure 3.9 shows the segmented
curve. Segments 0 and 1 have the same slope and do not compress the seg-
ment. For each higher-numbered segment, the step size is double that of
the previous segment. Each segment has sixteen steps. The result is a close
approximation to the actual curve.
The binary number produced by the codec in a telephone system has
eight bits. The first is a sign bit, which is a one for a positive voltage and a
zero for negative. Bits 2, 3, and 4 represent the segment number, from zero to
seven. The last four bits determine the step within the segment. If we nor-
malize the signal, that is, set the maximum input level equal to one volt, the
FIGURE 3.8
Sample-and-
hold circuit
FIGURE 3.9
Segmented mu-law
curve
step sizes can easily be calculated as follows: let the step size for the 0 and 1
segments be x mV. Then segment 2 has a step size of 2x, segment 3 a step size
of 3x, and so on. Since each segment has 16 steps, the value of x can be found
as follows.
16(x + x + 2x + 4x + 8x + 16x + 32x + 64x) = 1000 mV
x = 0.488 mV
The relationship between input voltage and segment is shown in Ta-
ble 3.1.
TABLE 3.1 Mu-Law Compressed PCM Coding
Segment Voltage Range (mV) Step Size (mV)
0 0–7.8 0.488
1 7.8–15.6 0.488
2 15.6–31.25 0.9772
3 31.25–62.5 1.953
4 62.5–125 3.906
5 125–250 7.813
6 250–500 15.625
7 500–1000 31.25
EXAMPLE 3.6 Y
Code a positive-going signal with amplitude 30% of the maximum allowed
as a PCM sample.
98 ! CHAPTER 3
SOLUTION
The signal is positive, so the first bit is a one. On the normalized voltage
scale, the amplitude is 300 mV. A glance at Table 3.1 shows that the signal is
in segment 6. That means the next three bits are 110 (6 in binary). This seg-
ment starts at 250 mV and increases 15.625 mV per step. The signal voltage is
50 mV above the lower limit, which translates into 50/15.625 = 3.2 steps.
This is less than halfway from step 3 to step 4, so it will be quantized as
step 3, making the last four bits 0011 (3 in binary). Therefore the code repre-
senting this sample is 11100011.
X
In operation, many modern codecs achieve compression by first encod-

ing the signal using a 12-bit linear PCM code, then converting the 12-bit
linear code into an 8-bit compressed code by discarding some of the bits.
This is a simple example of digital signal processing (DSP). Once an analog
signal has been digitized, it can be manipulated in a great many ways simply
by performing arithmetic with the bits that make up each sample. In the case
of the 12-to-8 bit conversion described here, some precision will be lost for
large-amplitude samples, but the data rate needed to transmit the informa-
tion will be much less than for 12-bit PCM. Since most of the samples in an
audio signal have amplitudes much less than the maximum, there is a gain
in accuracy compared with 8-bit linear PCM.
Briefly, the conversion works as follows. The 12-bit PCM sample begins
with a sign bit, which is retained. The other eleven bits describe the ampli-
tude of the sample, with the most significant bit first. For low-level samples,
the last few bits and the sign bit may be the only non-zero bits. The segment
number for the 8-bit code can be determined by subtracting the number of
leading zeros (not counting the sign bit) in the 12-bit code from 7. The next
four bits after the first 1 are counted as the level number within the segment.
Any remaining bits are discarded.
EXAMPLE 3.7 Y
Convert the 12-bit sample 100110100100 into an 8-bit compressed code.
SOLUTION
Copy the sign bit to the 8-bit code. Next count the leading zeros (2) and sub-
tract from 7 to get 5 (101 in binary). The first four bits of the 8-bit code are
thus 1101. Now copy the next four bits after the first 1 (not counting the sign
bit) to the 8-bit code. Thus the next four bits are 1010. Discard the rest. The
corresponding 8-bit code is 11011010.
X
The decoding process is the reverse of coding. It is illustrated in the block
diagram in Figure 3.10. The expansion process follows an algorithm analo-
gous to that used in the compressor. The low-pass filter at the output re-
moves the high-frequency components in the PAM signal that exits from the
digital-to-analog converter.
FIGURE 3.10 PCM decoding
Differential PCM Instead of coding the entire sample amplitude for each sample, it is possible
to code and transmit only the difference between the amplitude of the cur-
rent sample and that of the previous sample. Since successive samples often
have similar amplitudes, it should be possible to use fewer bits to encode the
changes. The most common (and most extreme) example of this process is
delta modulation, which is discussed in the next section.
' 3.4 Delta Modulation

In delta modulation, only one bit is transmitted per sample. That bit is a one
if the current sample is more positive than the previous sample, zero if the
current sample is more negative. Since only a small amount of information
about each sample is transmitted, delta modulation requires a much higher
sampling rate than PCM for equal quality of reproduction. Nyquist did not
say that transmitting samples at twice the maximum signal frequency would
always give undistorted results, only that it could, provided the samples
were transmitted accurately.
Figure 3.11 shows how delta modulation generates errors. In region (i),
the signal is not varying at all; the transmitter can only send ones and zeros,
however, so the output waveform has a triangular shape, producing a noise
signal called granular noise. On the other hand, the signal in region (iii)
changes more rapidly than the system can follow, creating an error in the
output called slope overload.
100 ! CHAPTER 3
FIGURE 3.11 Delta modulation
Adaptive Delta Adaptive delta modulation, in which the step size varies according to previ-
Modulation ous values, is more efficient. Figure 3.12 shows how it works. After a number
of steps in the same direction, the step size increases. A well-designed adap-
tive delta modulation scheme can transmit voice at about half the bit rate of
a PCM system, with equivalent quality.
' 3.5 Data Compression

In the previous section we looked at companding, which we noticed con-
sisted of compression at the transmitter and expansion at the receiver.
Essentially this is an analog technique for improving dynamic range by
increasing the signal-to-noise ratio for low-level signals although, as we
FIGURE 3.12 Adaptive delta modulation
saw, it can be implemented using digital signal processing. We now turn

our attention to the bits that result from the analog-to-digital conver-
sion just discussed and consider whether there is any way to reduce the num-
ber of bits that have to be transmitted per second. This reduction is also
called compression, but it is really a completely different process from
the one just described. We shall call it data compression to emphasize the
difference.
Generally, without data compression more bandwidth is required to
transmit an analog signal in digital form. For instance, analog telephony re-
quires less than 4 kHz per channel with single-sideband AM transmission.
Digital telephony conventionally operates at 64 kb/s. The exact bandwidth
requirement for this depends on the modulation scheme but is likely to be
much more than 4 kHz unless the channel has a very high signal-to-noise
radio and an elaborate modulation scheme is used. In order to use digital
102 ! CHAPTER 3
techniques in wireless communication, it is very desirable to reduce the

bandwidth to no more, and preferably less, than that needed for analog
transmission.
Lossy and There are two main categories of data compression. Lossless compression in-
Lossless volves transmitting all of the data in the original signal but using fewer bits.
Compression Lossy compression, on the other hand, allows for some reduction in the quality
of the transmitted signal. Obviously there has to be some limit on the loss in
quality, depending on the application. For instance, up until now the expecta-
tion of voice quality has been less for a mobile telephone than for a wireline
telephone. This expectation is now changing as wireless telephones become
more common. People are no longer impressed with the fact that wireless tele-
phony works at all; they want it to work as well as a fixed telephone.
Lossless compression schemes generally look for redundancies in the
data. For instance, a string of zeros can be replaced with a code that tells the
receiver the length of the string. This technique is called run-length encod-
ing. It is very useful in some applications: facsimile (fax) transmission, for
instance, where it is unnecessary to transmit as much data for white space on
the paper as for the message.
In voice transmission it is possible to greatly reduce the bit rate, or even
stop transmitting altogether, during time periods in which there is no
speech. For example, during a typical conversation each person generally
talks for less than half the time. Taking advantage of this to increase the
bandwidth for transmission in real time requires there to be more than one
signal multiplexed. When the technique is applied to a radio system, it also
allows battery-powered transmitters to conserve power by shutting off or
reducing power during pauses in speech.
Lossy compression can involve reducing the number of bits per sample
or reducing the sampling rate. As we have seen, the first reduces the
signal-to-noise ratio and the second limits the high-frequency response of
the signal, so there are limits to both methods. Other lossy compression
methods rely on knowledge of the type of signal, and often, on knowledge of
human perception. This means that voice, music, and video signals would
have to be treated differently. These more advanced methods often involve
the need for quite extensive digital signal processing. Because of this, they
have only recently become practical for real-time use with portable equip-
ment. A couple of brief examples will show the sort of thing that is possible.
Vocoders A vocoder (voice coder) is an example of lossy compression applied to human

speech. A typical vocoder tries to reduce the amount of data that needs to be
transmitted by constructing a model for the human vocal system. Human
sounds are produced by emitting air from the lungs at an adjustable rate. For
voiced sounds this air causes the vocal cords to vibrate at an adjustable fre-
quency; for unvoiced sounds the air passes the vocal cords without vibrating
them. In either case, the sound passes through the larynx and mouth, which
act as filters, changing the frequency response of the system at frequent in-
tervals. Typically there are from three to six resonant peaks in the frequency
response of the vocal tract.
Vocoders can imitate the human voice with an electronic system. Mod-
ern vocoders start with the vocal-tract model above. There is an excitation
function, followed by a multi-pole bandpass filter. Parameters for the excita-
tion and the filter response must be transmitted at intervals of about 20 ms,
depending on the system. Vocoders of this type are known as linear predictive
coders because of the mathematical process used to generate the filter param-
eters from an analysis of the voice signal.
The first step in transmitting a signal using a vocoder is to digitize it in
the usual way, using PCM, generally at 64 kb/s. Then the signal is analyzed
and the necessary excitation and filter parameters extracted. Only these pa-
rameters need to be sent to the receiver where the signal is reconstructed.
The transmitted data rate is typically in the range of about 2.4 to 9.6 kb/s,
allowing a much smaller transmission bandwidth than would be required
for the original 64 kb/s rate.
There are two main ways of generating the excitation signal in a linear
predictive vocoder. In pulse excited linear predictive (PELP or sometimes
RPELP, for regular pulse excited linear predictive) vocoders, a white noise gener-
ator is used for unvoiced sounds, and a variable-frequency pulse generator
produces the voiced sounds. The pulse generator creates a tone rich in har-
monics, as is the sound produced by human vocal cords. Both sources have
variable amplitudes. Figure 3.13 illustrates the process at the receiver.
Residual excited linear predictive (RELP) vocoders, on the other hand,
apply the inverse of the filter that will be used at the receiver to the voice
signal. The output of this filter is a signal that, when applied to the receiver
filter, will reproduce the original signal exactly. Figure 3.14 shows how this
FIGURE 3.13 PELP vocoder

104 ! CHAPTER 3
FIGURE 3.14 Generation of excitation signal using codebook
process works at the transmitter. The residual signal is too complex to trans-
mit exactly with the available bit rate, so it must be represented in a more
economical way. One method is to compare it with values in a table, called a
codebook, and transmit the number of the closest codebook entry. The re-
ceiver looks up the codebook entry, generates the corresponding signal, and
uses it instead of the pulse and noise generators shown in Figure 3.13. Many
other vocoder variations are possible as well.
Reasonable quality can be achieved with vocoders using data rates much
lower than those required for PCM. So far, the quality is not quite as good as
for straightforward PCM, however.
It should be obvious that vocoders are intended for use with voice only;
whereas, the PCM system described above can be used to send any 64 kb/s
data stream, including music, fax, or computer files. None of these will work
properly with a vocoder. Vocoders even tend to give a somewhat unnatural
quality to human speech. Still, the gain in bit rate and hence bandwidth,
compared to PCM, is so great that vocoders are very common in digital wire-
less voice communication.
( Modern communication systems are often a mixture of analog and digital
sources and transmission techniques. The trend is toward digital systems.
( Modern digital systems have better performance and use less bandwidth
than equivalent analog systems.
( An analog signal that is to be transmitted digitally must be sampled at
least twice per cycle of its highest-frequency component. Failure to do so
creates undesirable aliasing.
( PCM requires that the amplitude of each sample of a signal be converted
to a binary number. The more bits used for the number, the greater the ac-
curacy, but the greater the bit rate required.
( Delta modulation transmits only one bit per sample, indicating whether
the signal level is increasing or decreasing, but it needs a higher sampling
rate than PCM for equivalent results.
( The signal-to-noise ratio for either PCM or delta modulation signals can
often be improved by using companding.
( Lossless compression eliminates redundant data bits, thereby reducing
the bit rate with no effect on signal quality.
( Lossy compression compromises signal quality in order to reduce the bit
rate. For voice transmissions, vocoders are often used to achieve great re-
ductions in bit rate.
( Equation List
fa = fs − fb (3.8)
Eb E sin πτ / T
v (t ) = sin ω bt + b sin(ω s + ω b )t (3.9)
T πτ
E b sin πτ / T
− sin(ω s − ω b )t
πτ
N = 2m (3.10)
DR = 1.76 + 6.02m dB (3.11)
D = ƒsm (3.12)
vo = (3.13)
ln(1 + µ)
( Key Terms
aliasing distortion created by using too low a sampling rate when coding
an analog signal for digital transmission
codec device that converts sampled analog signal to and from its PCM or
delta modulation equivalent
coding conversion of a sampled analog signal into a PCM or delta
modulation bitstream
companding combination of compression at the transmitter and
expansion at the receiver of a communication system
106 ! CHAPTER 3
decoding conversion of a PCM or delta modulation bitstream to analog

samples
delta modulation a coding scheme that records the change in signal
level since the previous sample
digital signal processing (DSP) filtering of signals by converting them to
digital form, performing arithmetic operations on the data bits, then
converting back to analog form
flat-topped sampling sampling of an analog signal using a sample-and-
hold circuit, such that the sample has the same amplitude for its
whole duration
foldover distortion see aliasing
framing bits bits added to a digital signal to help the receiver to detect
the beginning and end of data frames
natural sampling sampling of an analog signal so that the sample
amplitude follows that of the original signal for the duration of the
sample
pulse-amplitude modulation (PAM) a series of pulses in which the
amplitude of each pulse represents the amplitude of the information
signal at a given time
pulse-code modulation (PCM) a series of pulses in which the amplitude
of the information signal at a given time is coded as a binary
number
pulse-duration modulation (PDM) a series of pulses in which the
duration of each pulse represents the amplitude of the information
signal at a given time
pulse-position modulation (PPM) a series of pulses in which the timing
of each pulse represents the amplitude of the information signal at a
given time
pulse-width modulation (PWM) see pulse-duration modulation (PDM)
quantizing representation of a continuously varying quantity as one of a
number of discrete values
quantizing errors inaccuracies caused by the representation of a
continuously varying quantity as one of a number of discrete values
quantizing noise see quantizing errors
run-length encoding method of data compression by encoding the
length of a string of ones or zeros instead of transmitting all the one
or zero bits individually
slope overload in delta modulation, an error condition that occurs when

the analog signal to be digitized varies too quickly for the system to
follow
vocoder circuit for digitizing voice at a low data rate by using knowledge
of the way in which voice sounds are produced
( Questions
1. Give four advantages and one disadvantage of using digital (rather than
analog) techniques for the transmission of voice signals.
2. Explain the necessity for sampling an analog signal before transmitting
it digitally.
3. What is the Nyquist rate? What happens when a signal is sampled at less
than the Nyquist rate?
4. Explain the difference between natural and flat-topped sampling.
5. (a) List three types of analog pulse modulation.
(b) Which pulse modulation scheme is used as an intermediate step in
the creation of PCM?
(c) Which pulse modulation scheme also finds use in audio amplifiers
and motor speed-control systems?
6. What is meant by the term quantizing noise?
7. For a PCM signal, describe the effects of:
(a) increasing the sampling rate
(b) increasing the number of bits per sample
8. (a) Briefly explain what is meant by companding.
(b) What advantage does companded PCM have over linear PCM for
voice communication?
9. How does differential PCM differ from standard PCM?
10. Explain why the sampling rate must be greater for delta modulation
than for PCM.
11. What is meant by slope overload in a delta modulation system? How
can this problem be reduced?
12. What are the two functions of a codec? Where in a telephone system is
it usually located?
13. Explain briefly how µ-law compression is implemented in a typical
codec.
108 ! CHAPTER 3
14. Explain the difference between lossless and lossy data compression.
Give an example of each.
15. How do vocoders model the human vocal cords? How do they model
the mouth and larynx?
16. What gives vocoders their somewhat artificial voice quality?
17. Does digital audio always have higher quality than analog audio?
Explain.
( Problems
1. It is necessary to transmit the human voice using a frequency range
from 300 Hz to 3.5 kHz using a digital system.
(a) What is the minimum required sampling rate, according to theory?
(b) Why would a practical system need a higher rate than the one you
calculated in part (a)?
2. The human voice actually has a spectrum that extends to much higher
frequencies than are necessary for communication. Suppose a fre-
quency of 5 kHz was present in a sampler that sampled at 8 kHz.
(a) What would happen?
(b) How can the problem described in part (a) be prevented?
3. A 1-kHz sine wave with a peak value of 1 volt and no dc offset is sampled
every 250 microseconds. Assume the first sample is taken as the voltage
crosses zero in the upward direction. Sketch the results over 1 ms using:
(a) PAM with all pulses in the positive direction
(b) PDM
(c) PPM
4. The compact disc system of digital audio uses two channels with TDM.
Each channel is sampled at 44.1 kHz and coded using linear PCM with
sixteen bits per sample. Find:
(a) the maximum audio frequency that can be recorded (assuming
ideal filters)
(b) the maximum dynamic range in decibels
(c) the bit rate, ignoring error correction and framing bits
(d) the number of quantizing levels
5. Suppose an input signal to a µ-law compressor has a positive voltage
and an amplitude 25% of the maximum possible. Calculate the output
voltage as a percentage of the maximum output.
6. Suppose a composite video signal with a baseband frequency range

from dc to 4 MHz is transmitted by linear PCM, using eight bits per sam-
ple and a sampling rate of 10 MHz.
(a) How many quantization levels are there?
(b) Calculate the bit rate, ignoring overhead.
(c) What would be the maximum signal-to-noise ratio, in decibels?
(d) What type of noise determines the answer to part (c)?
7. How would a signal with 50% of the maximum input voltage be coded
in 8-bit PCM, using digital compression?
8. Convert a sample coded (using mu-law compression) as 11001100 to a
voltage with the maximum sample voltage normalized as 1 V.
9. Convert the 12-bit PCM sample 110011001100 to an 8-bit compressed
sample.
10. A typical PCS system using a vocoder operates at 9600 b/s. By what fac-
tor has the amount of data required been reduced, compared with stan-
dard digital telephony?
Digital Modulation 4
Objectives
( Describe the basic types of digital modulation.
( Calculate the maximum data rate for a channel with a given modulation
scheme and signal-to-noise ratio.
( Explain the use of eye diagrams and constellation diagrams.
( Explain the difference between bit rate and baud rate and calculate both
for typical digital modulation systems.
( Describe and compare FSK, PSK, and QAM and perform simple calculations
with each.
( Explain the concepts of multiplexing and multiple access using frequency
and time division.
( Describe the principles of spread-spectrum communication and distinguish
between frequency-hopping and direct-sequence systems.
( Calculate spreading gain and signal-to-noise ratio for spread-spectrum
systems.
( Describe code-division multiple access and compare with FDMA and TDMA.
112 ! CHAPTER 4
' 4.1 Introduction

Digital signals have become very important in wireless communication.
Some of the information to be transmitted is already in digital form. In
Chapter 3, we examined the ways in which analog signals such as voice can
be converted to digital form for transmission.
In order to send data by radio, it is necessary to use a higher frequency car-
rier wave, just as for analog communication. Since the high-frequency carrier is
a sine wave for digital as well as analog signals, the same three basic parameters
are available for modulation: amplitude, frequency, and phase. All three, singly
and in combination, are used in digital systems.
Often the modulator and demodulator are collectively described as a
modem. You are probably familiar with telephone modems. These modulate
digital data onto an audio-frequency carrier. Radio-frequency modems are
similar in principle though quite a bit different in construction due to the
much higher carrier frequencies used.
In a holdover from the days of Morse code (a digital, though not a bi-
nary, communication scheme), the word keying is still often used to denote
digital modulation schemes. Thus we have frequency-shift keying (FSK)
and phase-shift keying (PSK). Straightforward amplitude-shift keying
(ASK) is rare in digital communication unless we count Morse code, but
quadrature AM (QAM) is very common. In this chapter we shall look at all
of these schemes as they are used in wireless transmission.
In Chapter 1 we noted that information capacity is proportional to
bandwidth. This fundamental relationship is given by Hartley’s law, which
is repeated here for convenience:
I = ktB (4.1)
where
I = amount of information to be transmitted in bits
k = a constant that depends on the modulation scheme and the
t = time in seconds
We noted in Chapter 1 that the constant k is important. In Chapter 2,

when we studied analog signals, we saw that the bandwidth required for the
same information rate varied greatly with the modulation scheme. In that
chapter we used baseband bandwidth as a rough measure of information
content. We are now in a position to discuss the value of k for different digi-
tal modulation schemes. The information rate for digital communication
will be expressed in bits per second.
DIGITAL MODULATION ! 113
Digital signals result in discrete, rather than continuous, changes in the

modulated signal. The receiver examines the signal at specified times, and
the state of the signal at each such time is called a symbol. The eventual out-
put will be binary, but it is certainly possible to use more than two states for
the transmitted symbol. Complex schemes using several levels can send
more data in a given bandwidth.
There is a theoretical limit to the maximum data rate that can be trans-
mitted in a channel with a given bandwidth. The Shannon-Hartley theorem
states:
C = 2B log 2 M (4.2)
where
C = information capacity in bits per second
B = channel bandwidth in hertz
M = number of possible states per symbol
For instance, if an FSK modulator can transmit either of two frequencies,

M = 2. A more elaborate modulator, using four different phase angles, each of
which can have two different amplitudes, has M = 4 × 2 = 8.
The limiting effect of bandwidth on data rate is understood most easily
by looking at a low-pass, rather than a bandpass, channel. Suppose that the
channel can pass all frequencies from zero to some maximum frequency B.
Then, of course, the highest frequency that can be transmitted is B. Suppose
that a simple binary signal consisting of alternate ones and zeros is transmit-
ted through the channel. This time let a logic 1 be 1 V and a logic 0 be −1 V.
The input signal will look like Figure 4.1(a): it will be a square wave with a
frequency one-half the bit rate (since there are two bits, a one and a zero,
for each cycle). This signal, which is a square wave, has harmonics at all odd
FIGURE 4.1 Digital transmission through a low-pass channel

114 ! CHAPTER 4
multiples of its fundamental frequency, with declining amplitude as the

frequency increases. This was shown in Chapter 1, using Fourier analysis.
At very low bit rates the output signal after passage through the channel will
be similar to the input, but as the bit rate increases, the frequency of the
square wave also increases and more of its harmonics are filtered out. There-
fore, the output will become more and more distorted. Finally, for a bit rate
of 2B the frequency of the input signal will be B, and only the fundamental
of the square wave will pass through the channel, as shown in Figure 4.1(b).
The receiver will still be able to distinguish a one from a zero, and the infor-
mation will be transmitted. Thus, with binary information, the channel
capacity will be 2B, as predicted by Equation (4.2).
Often the ratio of bit rate to channel bandwidth, expressed in bits per
second per hertz, is used as a figure of merit in digital radio systems. In the
preceding paragraph, we can see that for a two-level code the theoretical
maximum bandwidth efficiency is 2 b/s/Hz.
If a random pattern of ones and zeros were transmitted using the code of
Figure 4.1 and the signal were applied to an oscilloscope sweeping at the bit
rate, we would see the patterns shown in Figure 4.2. These are called “eye
diagrams.” The reason should be obvious from Figure 4.2(b). Because some
sweeps of the scope take place during a high- and some during a low-signal
level, both levels appear on the scope. A diagram like that in Figure 4.2(a)
indicates a data rate much lower than can be carried by the channel. Fig-
ure 4.2(b) indicates an optimal system. If an attempt were made to transmit
data too quickly, the eye would begin to close due to intersymbol interfer-
ence, as shown in Figure 4.2(c).
Now suppose that four voltage levels, each corresponding to a two-bit se-
quence, are used, as in Figure 4.3. The bandwidth required for the funda-
mental of this signal is the same as before. We have, it seems, managed to
transmit twice as much information in the same bandwidth. Again, this
agrees with Equation (4.2).
Unfortunately, the information capacity of a channel cannot be in-
creased without limit by increasing the number of states because noise
makes it difficult to distinguish between signal states. The ultimate limit is
called the Shannon limit:
C = B log 2 (1 + S/N) (4.3)
where
C = information capacity in bits per second
S/N = signal-to-noise ratio (as a power ratio, not in decibels)
FIGURE 4.2 Eye diagrams
FIGURE 4.3
Four-level code
116 ! CHAPTER 4
By the way, it is quite easy to find logs to the base 2, even if your calcula-
tor lacks this function. Simply find the log to the base 10 (the common log)
of the given number and divide by the log of 2, that is:
log 10 N
log 2 N = (4.4)
log 10 2
The effect of noise on a signal can be seen in the eye diagram of Fig-
ure 4.4. The noise causes successive oscilloscope traces to be at different
amplitudes. If the noise is severe enough, the eye closes and data recovery is
unreliable.
FIGURE 4.4
Eye diagram showing
inter-symbol interference
and noise
EXAMPLE 4.1 Y
A radio channel has a bandwith of 10 kHz and a signal-to-noise ratio of
15 dB. What is the maximum data rate than can be transmitted:
(a) Using any system?
(b) Using a code with four possible states?
SOLUTION
(a) We can find the theoretical maximum data rate for this channel from
Equation (4.3). First, though, we need the signal-to-noise ratio as a
power ratio. We can convert the given decibel value as follows:
S  15 
= log −1  
N  10 
= 316
.
Now we can use Equation (4.3):
C = B log2(1 + S/N)
= 10 × 103 log2(1 + 31.6)
= 10 × 103 × 5.03
= 50.3 kb/s
(b) We can use Equation (4.2) to find the maximum possible bit rate given
the specified code and bandwidth. We will then have to compare this
answer with that of part (a). From Equation (4.2),
C = 2B log2 M
= 2 × 10 × 103 × log24
= 2 × 10 × 103 × 2
= 40 kb/s
Since this is less than the maximum possible for this channel, it should
be possible to transmit over this channel, with a four-level scheme, at
40 kb/s. A more elaborate modulation scheme would be required to attain
the maximum data rate of 50.3 kb/s for the channel.
X
At this point we should distinguish between bit rate and baud rate. The
bit rate is simply the number of bits transmitted per second (C in the preced-
ing two equations), while the baud rate is the number of symbols per second.
Therefore, if we let the baud rate be S (for symbols, since B is already being
used for bandwidth), then:
C = S log 2 M (4.5)
where
C = capacity in bits per second
S = baud rate in symbols per second
M = number of possible states per symbol
EXAMPLE 4.2 Y
A modulator transmits symbols, each of which has 64 different possible
states, 10,000 times per second. Calculate the baud rate and bit rate.
SOLUTION
The baud rate is simply the symbol rate, or 10 kbaud. The bit rate is given by
Equation (4.5):
118 ! CHAPTER 4
C = S log2 M
= 10 × 103 × log264
= 60 kb/s
' 4.2 Frequency-Shift Keying

Probably the simplest digital modulation scheme in current use is fre-
quency-shift keying (FSK). In its simplest form two frequencies are transmit-
ted, one corresponding to binary one, the other to zero. In digital
communication systems, a one is often denoted by the term mark, and a
zero is called a space. This is another holdover from telegraphy in which
Morse code messages were often recorded as marks on a paper tape. FSK is a
robust scheme; that is, like analog FM, it tends to be reliable in the presence
of noise. Its disadvantage is: since each symbol has only two possible states,
it is not very efficient in terms of bandwidth. It tends to be used for
low-data-rate applications, such as pagers, and for transmitting bursts of
data over systems that are mainly analog.
FSK is also used extensively in high-frequency radio systems for radio-
teletype transmission. High-frequency radio channels tend to be very noisy,
and phase shifts induced into the signal by travel through the ionosphere
make the use of any scheme requiring accurate phase information impracti-
cal. Data rates for HF communication are very low (on the order of 100 b/s)
and frequency shifts between mark and space vary from 170 Hz to 850 Hz.
Usually HF FSK is actually transmitted and received using SSBSC AM
equipment. Two different audio tones, one for mark and one for space, with
the required frequency difference are chosen. These are applied one at a time
to the microphone input of the transmitter. This results in the transmission
of RF frequencies that differ by the same amount as the audio frequencies. At
the receiver two audio tones differing by the required frequency shift will be
produced. Figure 4.5 shows the spectrum of a typical HF FSK transmission, as
well as a typical pair of audio tones used to modulate the transmitter.
A variant of FSK is AFSK (audio frequency-shift keying), in which two
different audio frequencies, corresponding to mark and space, are used to
modulate a carrier by any of the analog methods discussed in Chapter 2.
AFSK is often combined with FM modulation for the audio tones. Figure 4.6
shows how this can work using a VHF amateur radio link as an example.
There is an extensive amateur packet-radio network, using AFSK transmis-
sion with a frequency shift of 1 kHz between mark and space, operating at a
data rate of 1200 bits per second.
FIGURE 4.5 Spectrum of HF FSK transmission
FIGURE 4.6 AFSK using FM equipment

120 ! CHAPTER 4
It is possible to build an FSK system with more than two different fre-
quencies in order to increase the number of bits per symbol, but it is usually
more efficient to move to a system using phase shifts, or a combination of
amplitude and phase shifts, when this is required.
Gaussian A special case of FSK called Gaussian minimum-shift keying (GMSK) is

Minimum-Shift used in the GSM cellular radio and PCS systems to be described later. In a
Keying minimum-shift system, the mark and space frequencies are separated by half
the bit rate, that is:
ƒ m − ƒ s = 0.5 ƒ b (4.6)
where
ƒ m = frequency transmitted for mark (binary 1)
ƒ s = frequency transmitted for space (binary 0)
ƒ b = bit rate
If we use the conventional FM terminology from Chapter 2, we see that
GMSK has a deviation each way from the center (carrier) frequency, of
δ = 0.25 ƒ b
which corresponds to a modulation index of
δ
mƒ =
ƒm
0.25ƒ b
=
ƒb
= 0.25
The word Gaussian refers to the shape of a filter that is used before the
modulator to reduce the transmitted bandwidth of the signal. GMSK uses
less bandwidth than conventional FSK, because the filter causes the trans-
mitted frequency to move gradually between the mark and space frequen-
cies. With conventional FSK the frequency transition is theoretically
instantaneous, and in practice as rapid as the hardware allows, producing
sidebands far from the carrier frequency.
EXAMPLE 4.3 Y
The GSM cellular radio system uses GMSK in a 200-kHz channel, with a
channel data rate of 270.833 kb/s. Calculate:
(a) the frequency shift between mark and space
(b) the transmitted frequencies if the carrier (center) frequency is exactly

880 MHz
(c) the bandwidth efficiency of the scheme in b/s/Hz
SOLUTION
(a) The frequency shift is
ƒm − ƒs = 0.5 ƒb = 0.5 × 270.833 kHz = 135.4165 kHz
(b) The shift each way from the carrier frequency is half that found in (a) so
the maximum frequency is
ƒmax = ƒc + 0.25 ƒb = 880 MHz + 0.25 × 270.833 kHz = 880.0677 MHz
and the minimum frequency is
ƒmin = ƒc − 0.25 ƒb = 880 MHz − 0.25 × 270.833 kHz = 879.93229 MHz
(c) The GSM system has a bandwidth efficiency of 270.833/200 = 1.35

b/s/Hz, comfortably under the theoretical maximum of 2 b/s/Hz for a
two-level code.
X
' 4.3 Phase-Shift Keying

When somewhat higher data rates are required in a band-limited channel
than can be achieved with FSK, phase-shift keying (PSK) is often used. Mea-
suring phase requires a reference phase, which would be hard to maintain
accurately. Usually, the phase of each symbol is compared with that of the
previous symbol, rather than with a constant reference. This type of PSK is
more completely described as delta phase-shift keying (DPSK). Most DPSK
modems use a four-phase system called quadrature phase-shift keying
(QPSK or DQPSK). In QPSK, each symbol represents two bits and the bit rate
is twice the baud rate. This is called a dibit system. Such a system can carry
twice as much data in the same bandwidth as can a single-bit system like
FSK, provided the signal-to-noise ratio is high enough.
Figure 4.7 is a vector diagram that represents a typical DQPSK system.
Phase shifts are given with respect to the phase of the previous symbol. Each
of the four possible phase shifts (including no shift at all) is associated with a
two-bit sequence, as shown in Table 4.1.
122 ! CHAPTER 4
FIGURE 4.7
Delta quadrature
phase-shift keying
TABLE 4.1 DQPSK Coding
Phase Shift (degrees) Symbol

0 00
+90 01
−90 10
180 11
π/4 Delta The system shown in Figure 4.7 and Table 4.1 requires a 180 degree transi-
Phase-Shift tion for the symbol 11. The transmitted signal has to go to zero amplitude
Keying momentarily as it makes this transition. Accurate transmission of this signal
therefore requires a linear amplifier, unlike the case for FSK. In fact, the am-
plifier should be linear all the way down to zero output. This is quite possi-
ble, of course, but linear amplifiers are markedly less efficient than
nonlinear amplifiers. The need for linearity can be reduced, though not
eliminated, by changing to a system called π/4 DQPSK. Here the allowable
transitions from the previous phase angle are ±45° and ±135°. Neither of
these requires the signal amplitude to go through zero, relaxing the linearity
requirements somewhat. A typical π/4 DQPSK system has the state table
shown in Table 4.2, and a vector diagram showing the possible transitions
TABLE 4.2 π/4 DQPSK Coding
Phase Shift (degrees) Symbol

45 00
135 01
−45 10
−135 11
can be found in Figure 4.8. This system is used for the North American
TDMA cell phone and PCS systems.
FIGURE 4.8
π/4 DQPSK
EXAMPLE 4.4 Y
The North American TDMA digital cell phone standard transmits at
24.3 kilobaud using DQPSK. What is the channel data rate?
SOLUTION
Since this is a dibit system, the symbol rate, also known as the baud rate, is
half the bit rate. Therefore the data rate is 48.6 kb/s.
X
' 4.4 Quadrature Amplitude Modulation

The only way to achieve high data rates with a narrowband channel is
to increase the number of bits per symbol, and the most reliable way to do
that is to use a combination of amplitude and phase modulation known as
quadrature amplitude modulation (QAM). For a given system, there is a finite
number of allowable amplitude-phase combinations. Figure 4.9(a) is a con-
stellation diagram that shows the possibilities for a hypothetical system
with sixteen amplitude/phase combinations. Thus each transmitted symbol
represents four bits. This diagram is similar to the previous figure except that
the vectors are not drawn. Each dot represents a possible amplitude/phase
combination or state. With a noiseless channel, the number of combina-
tions could be increased indefinitely, but a practical limit is reached when
the difference between adjacent states becomes too small to be detected
reliably in the presence of noise and distortion. If a QAM signal is applied to
124 ! CHAPTER 4
FIGURE 4.9 Quadrature AM
the oscilloscope, the noise can be seen as a blurring of the points in the con-
stellation, as shown in Figure 4.9(b).
In fixed terrestrial microwave systems QAM is used with quite a large
number of states—up to 1024 in some cases. This requires a very high sig-
nal-to-noise ratio however, and portable and mobile systems are much more
limited.
QAM is more efficient in terms of bandwidth than either FSK or QPSK,
but it is also more susceptible to noise. Another disadvantage compared to
FSK is that QAM signals, like analog AM signals, vary in amplitude. This
means that transmitter amplifiers must be linear.
EXAMPLE 4.5 Y
A modem uses 16 different phase angles and 4 different amplitudes. How
many bits does it transmit for each symbol?
SOLUTION
The number of possible states per symbol is 16 × 4 = 64
The number of bits per symbol is log264 = 6
X
' 4.5 Multiplexing and Multiple Access

Multiplexing allows several signals to be carried on a single transmission
channel or medium. When the signals to be multiplexed originate at differ-
ent locations, the system is often described by using the term multiple
access. Both are used in wireless communication.
Frequency-Division The simplest multiple access scheme is the one used by radio and television
Multiplexing and broadcasting stations. Each signal is assigned a portion of the available fre-
Multiple Access quency spectrum on a full-time basis. This is called frequency-division
multiplexing (FDM) or frequency-division multiple access (FDMA) de-
pending on the situation. For instance, over-the-air broadcasts are FDMA
while a cable-television system, where all the signals are assigned slots on
the same cable by the headend equipment, is an example of FDM. Frequency
division can be and is used with both analog and digital signals.
Time-Division Time-division multiplexing (TDM) is used mainly for digital communica-

Multiplexing and tion. In TDM, each information signal is allowed to use all the available
Multiple Access bandwidth, but only for part of the time. From Hartley’s Law (Equation 4.1),
it can be seen that the amount of information transmitted is proportional to
both bandwidth and time. Therefore, at least in theory, it is equally possible
to divide the bandwidth or the time among the users of a channel. Continu-
ously varying signals, such as analog audio, are not well adapted to TDM,
because the signal is present at all times. On the other hand, sampled au-
dio is very suitable for TDM, as it is possible to transmit one sample or bit
from each of several sources sequentially, then send the next sample or
bit from each source, and so on. As already mentioned, sampling itself does
not imply digital transmission, but in practice sampling and digitizing usu-
ally go together.
Many signals can be sent on one channel by sending either a sample
from each signal, or a fixed number of bits from each signal, in rotation.
Time-division multiplexing requires that the total bit rate be multiplied
by the number of channels multiplexed. This means that the bandwidth
requirement is also multiplied by the number of signals.
TDM in TDM is used extensively in digital telephony. The simplest North American
Telephony standard is known as the DS-1 signal, which consists of 24 PCM voice chan-
nels, multiplexed using TDM. Each channel is sampled at 8 kHz, with 8 bits
per sample, as previously described. This gives a bit rate of 8 kb/s × 8 = 64 kb/s
for each voice channel.
The DS-1 signal consists of frames, each of which contains the bits repre-
senting one sample from each of the 24 channels. One extra bit, called the
framing bit, is added to each frame to help synchronize the transmitter and
receiver. Each frame contains 24 × 8 + 1 = 193 bits.
The samples must be transmitted at the same rate as they were obtained
in order for the signal to be reconstructed at the receiver without delay. This
requires the multiplexed signal to be sent at a rate of 8000 frames per second.
126 ! CHAPTER 4
Thus the bit rate is 193 × 8000 b/s = 1.544 Mb/s. See Figure 4.10 for an illus-
tration of a frame of a DS-1 signal.
FIGURE 4.10
DS-1 signal
Time-Division Time-division multiple access (TDMA) is like TDM except that it involves
Multiple Access signals originating at different points. The telephone system observed in the
previous section uses TDM because all the signals are combined at one point.
An example of TDMA is a digital cellular radio system where several signals
from mobile units are combined on one channel by assigning each a time
slot. TDMA systems are very similar in principle to TDM, but they tend to be
more complex to design. One complicating feature in TDMA radio systems
is: the propagation time for the signal from a mobile unit to a base station
varies with its distance to the base. We will look at the details later when we
discuss specific systems.
' 4.6 Spread-Spectrum Systems

As radio communication systems proliferate and traffic increases, interfer-
ence problems become more severe. Interference is nothing new, of course,
but it has been managed reasonably successfully in the past by careful regu-
latory control of transmitter locations, frequencies, and power levels. There
are some exceptions to this, however. Two examples are CB radio and cord-
less telephones. In fact, wherever government regulation of frequency use is
informal or nonexistent, interference is likely to become a serious problem.
Widespread use of such systems as cordless phones, wireless local-area net-
works, and wireless modems by millions of people obviously precludes the
tight regulation associated with services such as broadcasting, making both-
ersome interference almost inevitable. One approach to the problem, used
by cellular radio systems, is to employ a complex system of frequency reuse,
with computers choosing the best channel at any given time. However, this
system too implies strong central control, if not by government then by one
or more service providers, each having exclusive rights to certain radio chan-
nels. That is, in fact, the current situation with respect to cellular telephony,
but it can cause problems where several widely different services use the
same frequency range. The 49-MHz band, for instance, is currently used by
cordless phones, baby monitors, remote controlled models, and various
other users in an almost completely unregulated way. Similarly, the 2.4-GHz
band is shared by wireless LANs, wireless modems, cordless phones—and
even microwave ovens!
Another problem with channelized communication, even when tightly
controlled, is that the number of channels is strictly limited. If all available
channels are in use in a given cell of a cellular phone system, the next at-
tempt to complete a call will be blocked, that is, the call will not go through.
Service does not degrade gracefully as traffic increases; rather, it continues as
normal until the traffic density reaches the limits of the system and then
ceases altogether for new calls.
There is a way to reduce interference that does not require strong central
control. That technique, known as spread-spectrum communication, has
been used for some time in military applications where interference often
consists of deliberate jamming of signals. This interference, of course, is
not under the control of the communicator, nor is it subject to government
regulation.
Military communication systems need to avoid unauthorized eavesdrop-
ping on confidential transmissions, a problem alleviated by the use of spread-
spectrum techniques. Privacy is also a concern for personal communication
systems, but many current analog systems, such as cordless and cellular tele-
phone systems, have nonexistent or very poor protection of privacy.
For these reasons, and because the availability of large-scale integrated
circuits has reduced the costs involved, there has recently been a great deal
of interest in the use of spread-spectrum technology in personal communi-
cation systems for both voice and data.
The basic idea in spread-spectrum systems is, as the name implies, to
spread the signal over a much wider portion of the spectrum than usual. A
simple audio signal that would normally occupy only a few kilohertz of spec-
trum can be expanded to cover many megahertz. Thus only a small portion
of the signal is likely to be masked by any interfering signal. Of course, the
average power density, expressed in watts per hertz of bandwidth, is also re-
duced, and this often results in a signal-to-noise ratio of less than one (that
is, the signal power in any given frequency range is less than the noise power
in the same bandwidth).
It may seem at first glance that this would make the signal almost impos-
sible to detect, which is true unless special techniques are used to
“de-spread” the signal while at the same time spreading the energy from in-
terfering signals. In fact, the low average power density of spread-spectrum
signals is responsible for their relative immunity from both interference and
eavesdropping.
128 ! CHAPTER 4
EXAMPLE 4.6 Y
A voice transmission occupies a channel 30 kHz wide. Suppose a
spread-spectrum system is used to increase its bandwidth to 10 MHz. If the
signal has a total signal power of −110 dBm at the receiver input and the sys-
tem noise temperature referred to the same point is 300 K, calculate the
signal-to-noise ratio for both systems.
SOLUTION
Recall from Chapter 1 that thermal noise power is given by
PN = kTB
where
PN = noise power in watts
k = Boltzmann’s constant: 1.38 × 10−23 joules/kelvin (J/K)
T = absolute temperature in kelvins
In general, the noise power bandwidth for a system will be approxi-
mately equal to the receiver bandwidth. For the signal with a bandwidth of
30 kHz and a noise temperature of 300 K,
PN (30 kHz) = 1.38 × 10−23 J/K × 300 K × 30 × 103 Hz

= 124 × 10−18 W
= −129 dBm
When the signal bandwidth increases to 10 MHz, the signal is spread

over a much wider region of the spectrum, and a receiver designed to receive
the whole signal bandwidth would need a bandwidth of 10 MHz as well. It
would receive a noise power equal to
PN(10 MHz) = 1.38 × 10−23 J/K × 300 K × 10 × 106 Hz

= 41.4 × 10−15 W
= −104 dBm
With both signal and noise in dBm, we can subtract to get the signal-to-noise
ratio.
For the 30 kHz bandwidth,
S/N = −110 dBm − (−129 dBm) = 19 dB

For the 10 MHz bandwidth,
S/N = −110 dBm − (−104 dBm) = −6 dB

X
Spread-spectrum communication is especially effective in a portable or

mobile environment. Cancellation due to the reflection of signals often
causes very deep fades, called Rayleigh fading, over a narrow frequency
range. When one of these fades happens to coincide with the frequency in
use, the signal can be lost completely. With spread spectrum, only a small
part of the communication will be lost, and this can usually be made up, at
least in digital schemes, by the use of error-correcting codes.
There are two important types of spread-spectrum systems. They are
known as frequency hopping and direct sequence.
Frequency- Frequency hopping is the simpler of the two spread-spectrum techniques.

Hopping Systems A frequency synthesizer is used to generate a carrier in the ordinary way.
There is one difference, however: instead of operating at a fixed frequency,
the synthesizer changes frequency many times per second according to
a preprogrammed sequence of channels. This sequence is known as a
pseudo-random noise (PN) sequence because, to an outside observer who
has not been given the sequence, the transmitted frequency appears to hop
about in a completely random and unpredictable fashion. In reality, the
sequence is not random at all, and a receiver which has been programmed
with the same sequence can easily follow the transmitter as it hops and the
message can be decoded normally.
Since the frequency-hopping signal typically spends only a few millisec-
onds or less on each channel, any interference to it from a signal on that fre-
quency will be of short duration. If an analog modulation scheme is used
for voice, the interference will appear as a click and may pass unnoticed. If
the spread-spectrum signal is modulated using digital techniques, an error-
correcting code can be employed that will allow these brief interruptions in
the received signal to be ignored, and the user will probably not experience
any signal degradation at all. Thus reliable communication can be achieved
in spite of interference.
EXAMPLE 4.7 Y
A frequency-hopping spread-spectrum system hops to each of 100 frequen-
cies every ten seconds. How long does it spend on each frequency?
SOLUTION
The amount of time spent on each frequency is
t = 10 seconds/100 hops
= 0.1 second per hop
X
130 ! CHAPTER 4
If the frequency band used by the spread-spectrum system contains

known sources of interference, such as carriers from other types of service,
the frequency-hopping scheme can be designed to avoid these frequencies
entirely. Otherwise, the communication system will degrade gracefully as
the number of interfering signals increases, since each new signal will sim-
ply increase the noise level slightly.
Direct-Sequence The direct-sequence form of spread-spectrum communication is commonly

Systems used with digital modulation schemes. The idea is to modulate the transmit-
ter with a bit stream consisting of pseudo-random noise (PN) that has a
much higher rate than the actual data to be communicated. The term
pseudo-random means that the bit stream appears at first glance to be a ran-
dom sequence of zeros and ones but is actually generated in such a way as to
repeat exactly from time to time. The data to be transmitted is combined
with the PN. One common technique is to invert all the bits of the PN stream
during the time the “real” data is represented by a one and to leave the PN bit
stream unchanged when a data zero is to be transmitted. The extra bits trans-
mitted in this way are called chips, and the resulting bit rate is known as the
chipping rate. Most direct-sequence spread-spectrum systems use a chip-
ping rate at least ten times as great as the bit rate of the actual information to
be transmitted.
The use of the high-speed PN sequence results in an increase in the band-
width of the signal, regardless of what modulation scheme is used to encode
the bits into the signal. Recall from Hartley’s Law (Equation 4.1) that for any
given modulation scheme, the bandwidth is proportional to the bit rate. It
follows from Hartley’s Law that a direct-sequence system transmitting a to-
tal of ten bits for each information bit will use ten times as much bandwidth
as a narrowband signal with the same type of modulation and the same in-
formation rate. That is, the sidebands will extend ten times as far from the
carrier, as illustrated in Figure 4.11. Direct-sequence spread-spectrum
schemes typically use some form of phase-shift keying (PSK).
FIGURE 4.11 Direct-sequence spread-spectrum signal

EXAMPLE 4.8 Y
A digital communication scheme uses DQPSK. It is to transmit a compressed
PCM audio signal which has a bit rate of 16 kb/s. The chipping rate is 10
to 1. Calculate the number of signal changes (symbols) which must be trans-
mitted each second.
SOLUTION
The total bit rate, including the chips, is 10 times the data rate, or 160 kb/s.
Since there are four signal states, each state represents two bits. Therefore
the symbol rate is
160/2 = 80 kilobaud
X
Expanding the bandwidth by a factor of ten while keeping the transmit-
ted power constant will decrease the received signal-to-noise ratio by
the same factor. As before, the pseudo-random sequence is known to the re-
ceiver, which has to separate the information signal from the chips. A pro-
cessing gain, also called spreading gain, can be defined equal to the
bandwidth expansion (which, for direct-sequence spread spectrum, is also
equal to the ratio of chips to information bits):
BRF
Gp = (4.7)
BBB
where
G p = processing gain
B RF = RF (transmitted) bandwidth
B BB = baseband (before spreading) bandwidth
The processing gain also describes the amount by which the signal-
to-noise ratio of the signal is reduced by the spreading process during trans-
mission. Of course, this reduction is reversed at the receiver. Since
signal-to-noise ratio is generally given in decibels, it would make sense to ex-
press the processing gain that way too, that is:
G p (dB) = 10 log G p
G p (dB) = (S/N) i (dB) − (S/N) o (dB) (4.8)
where
G p (dB) = processing gain in decibels
(S/N) i (dB) = signal-to-noise ratio in dB before spreading
(S/N) o (dB) = signal-to-noise ratio in dB after spreading
132 ! CHAPTER 4
EXAMPLE 4.9 Y
A signal would have a bandwidth of 200 kHz and a signal-to-noise ratio of
20 dB if transmitted without spreading. It is spread using a chipping rate
of 50:1. Calculate its bandwidth and signal-to-noise ratio after spreading.
SOLUTION
The bandwidth after spreading can be found from Equation (4.7):
BRF
Gp =
BBB
BRF = G p BBB
= 50 × 200 kHz
= 10 MHz
The processing gain in decibels is
Gp (dB) = 10 log Gp
= 10 log 50
= 17 dB
The signal-to-noise ratio after spreading is given by Equation (4.8):

Gp (dB) = (S/N)i (dB) − (S/N)o (dB)
(S/N)o (dB) = (S/N)i (dB) − Gp (dB)
= 20 dB − 17 dB
= 3 dB
X
Reception of The type of receiver required for spread-spectrum reception depends on how
Spread-Spectrum the signal is generated. For frequency-hopped transmissions, what is needed is
Signals a relatively conventional narrowband receiver that hops in the same way as
and is synchronized with the transmitter. This requires that the receiver be
given the frequency-hopping sequence, and there be some form of synchroniz-
ing signal (such as the signal usually sent at the start of a data frame in digital
communication) to keep the transmitter and receiver synchronized. Some
means must also be provided to allow the receiver to detect the start of a trans-
mission, since, if this is left to chance, the transmitter and receiver will most
likely be on different frequencies when a transmission begins.
One way to synchronize the transmitter and receiver is to have the trans-
mitter send a tone on a prearranged channel at the start of each transmis-
sion, before it begins hopping. The receiver can synchronize by detecting
the end of the tone and then begin hopping according to the prearranged PN
sequence. Of course, this method fails if there happens to be an interfering
signal on the designated synchronizing channel at the time synchronization
is attempted.
A more reliable method of synchronizing frequency-hopping systems is
for the transmitter to visit several channels in a prearranged order before be-
ginning a normal transmission. The receiver can monitor all of these chan-
nels sequentially, and once it detects the transmission, it can sample the
next channel in the sequence for verification and synchronization.
Direct-sequence spread-spectrum transmissions require different recep-
tion techniques. Narrowband receivers will not work with these signals,
which occupy a wide bandwidth on a continuous basis. A wideband receiver
is required, but a conventional wideband receiver would output only noise.
In order to distinguish the desired signal from noise and interfering signals,
which over the bandwidth of the receiver are much stronger than the desired
signal, a technique called autocorrelation is used. Essentially this involves
multiplying the received signal by a signal generated at the receiver from the
PN code. When the input signal corresponds to the PN code, the output from
the autocorrelator will be large; at other times this output will be very small.
Of course, once again the transmitter and receiver will probably not be syn-
chronized at the start of a transmission, so the transmitter sends a preamble
signal, which is a prearranged sequence of ones and zeros, to let the receiver
synchronize with the transmitter.
Code-Division Traditionally, most analog communication systems have used frequency-

Multiple Access division multiplexing, and digital systems have employed time-division
(CDMA) multiplexing, in order to combine many information signals into a single
transmission channel. When the signals originate from different sources,
these two methods (which have already been studied) become frequency-
division multiple access (FDMA) and time-division multiple access (TDMA),
respectively.
Spread-spectrum communication allows a third method for multi-
plexing signals from different sources—code-division multiple access
(CDMA). All that is required is for each transmitter to be assigned a different
pseudo-noise (PN) sequence. If possible, orthogonal sequences should
be chosen; that is, the transmitters should never be in the same place at
the same time. The PN sequence for the transmitter is given only to the re-
ceiver that is to operate with that transmitter. This receiver will then receive
only the correct transmissions, and all other receivers will ignore these sig-
nals. This technique, which is applicable to both frequency-hopping and
direct-sequence transmissions, allows many transmissions to share the same
spread-spectrum channel.
134 ! CHAPTER 4
It is not necessary for all transmitters to have orthogonal PN sequences

in order to use CDMA. If some of the sequences are not orthogonal, there
will be some mutual interference between users. However, the system can
still operate until there are so many users that the signal-to-noise ratio
becomes unacceptably low.
In order that all transmissions other than the desired one blend together
at the receiver as noise, it is highly desirable in a CDMA system that all re-
ceived signals have about the same strength at the receiver. Obviously this is
not always possible, such as in an unlicensed environment like that of cord-
less phones. But where it is possible, the benefits of close control over trans-
mitter power levels are great. We will see this idea applied later when we look
at CDMA PCS systems.
( Digital wireless transmission uses frequency, phase, and amplitude varia-

tions, just as does analog transmission.
( The maximum data rate for a channel is a function of bandwidth, modu-
lation scheme, and signal-to-noise ratio.
( In general, more complex modulation schemes can achieve higher data
rates, but only when the signal-to-noise ratio is high.
( Frequency-shift keying (FSK) uses two (and occasionally more than two)
transmitted frequencies to achieve modest data rates with good perfor-
mance in noisy channels.
( Gaussian minimum-shift keying is a special case of FSK that achieves the
minimum bandwidth possible for a two-frequency FSK system at a given
data rate.
( Most phase-shift keying (PSK) systems use four phase angles for some-
what higher data rates than are achievable with FSK.
( Quadrature amplitude modulation (QAM) achieves higher data rates than
FSK or PSK by using a combination of amplitude and phase modulation.
QAM requires a relatively noise-free channel to realize its advantages.
( Frequency-division multiplexing and multiple-access schemes divide
available bandwidth among channels, with each operating full-time.
( Time-division multiplexing and multiple-access schemes divide the avail-
able time among channels, with each using the full bandwidth.
( Spread-spectrum systems reduce interference by spreading each signal
over a large bandwidth.
( Code-division multiple-access (CDMA) schemes assign all channels

full-time to the full available bandwidth. Channels are separated by
spreading and despreading codes.
( Equation List
I = ktB (4.1)
C = 2B log2 M (4.2)
C = B log2(1 + S/N) (4.3)
C = S log2 M (4.5)
ƒm − ƒs = 0.5 ƒb (4.6)
BRF
Gp = (4.7)
BBB
Gp (dB) = (S/N)i (dB) − (S/N)o (dB) (4.8)
( Key Terms
amplitude-shift keying (ASK) data transmission by varying the
amplitude of the transmitted signal
audio frequency-shift keying (AFSK) use of an audio tone of two or more
different frequencies to modulate a conventional analog transmitter
for data transmission
baud rate speed at which symbols are transmitted in a digital
communication system
bit rate speed at which data is transmitted in a digital communication
system
chips extra bits used to spread the signal in a direct-sequence
spread-spectrum system
code-division multiple access (CDMA) system to allow multiple users to
use the same frequency using separate PN codes and a
spread-spectrum modulation scheme
136 ! CHAPTER 4
constellation diagram in digital communication, a pattern showing all

the possible combinations of amplitude and phase for a signal
delta phase-shift keying (DPSK) digital modulation scheme that
represents a bit pattern by a change in phase from the previous state
dibit system any digital modulation scheme that codes two bits of
information per transmitted symbol
direct-sequence spread spectrum technique for increasing the
bandwidth of a transmitted signal by combining it with a
pseudo-random noise signal with a higher bit rate
frequency hopping form of spread-spectrum communication in which
the RF carrier continually moves from one frequency to another
according to a prearranged pseudo-random pattern
frequency-division multiple access (FDMA) sharing of a communication
channel among multiple users by assigning each a different carrier
frequency
frequency-division multiplexing (FDM) combining of several signals into
one communication channel by assigning each a different carrier
frequency
frequency-shift keying (FSK) digital modulation scheme using two or
more different output frequencies
Gaussian minimum-shift keying (GMSK) variant of FSK which uses the
minimum possible frequency shift for a given bit rate
mark in digital communication, a logic one
modem acronym for modulator-demodulator; device to enable data to
be transmitted via an analog channel
multiple access use of a single channel by more than one transmitter
multiplexing use of a single channel by more than one signal
phase-shift keying (PSK) digital modulation scheme in which the phase
of the transmitted signal is varied in accordance with the baseband
data signal
processing gain improvement in interference rejection due to spreading
in a spread-spectrum system
pseudo-random noise (PN) sequence a transmitted series of ones and
zeros that repeats after a set time, and which appears random if the
sequence is not known to the receiver
quadrature AM (QAM) modulation scheme in which both the amplitude
and phase of the transmitted signal are varied by the baseband signal
quadrature phase-shift keying (QPSK or DQPSK) digital modulation

scheme using four different transmitted phase angles
Rayleigh fading variation in received signal strength due to multipath
propagation
space binary zero
spreading gain improvement in interference rejection due to spreading
in a spread-spectrum system
symbol in digital communication, the state of the signal at a sampling
time
time-division multiple access (TDMA) system to allow several
transmissions to use a single channel by assigning time slots to each
time-division multiplexing (TDM) system to combine several data
streams onto a single channel by assigning time slots to each
( Questions
1. What is the meaning of the term modem?
2. What parameters of a sine-wave carrier can be modulated?
3. Name the three most common basic types of digital modulation.
4. Which type of modulation is likely to be used for:
(a) low data rates
(b) moderate data rates
(c) high data rates
5. What signal parameters are varied with QAM?
6. What factors limit the maximum data rate for a channel?
7. What is an eye diagram?
8. Explain the difference between the terms bit rate and baud rate.
9. Explain the origin and meaning of the terms mark and space.
10. What is the difference between FSK, AFSK, and GMSK?
11. Why is delta phase-shift keying the most common form of PSK?
12. What is the advantage of π/4 DQPSK?
13. What is represented by the dots in a constellation diagram for a QAM
system?
14. Compare the modulation schemes studied in this chapter, listing as
many advantages and disadvantages for each as you can.
138 ! CHAPTER 4
15. List and describe the three multiple-access systems in common use.
16. What is a DS-1 signal?
17. Compare frequency-hopping and direct-sequence spread-spectrum
systems.
18. What happens when a call is blocked?
19. How do spread-spectrum systems reduce the effect of fading?
20. Briefly describe what is meant by orthogonal spread-spectrum signals.
( Problems
1. The North American analog cellular radio system uses FM with channels
30 kHz wide. Suppose such a channel were used for digital communica-
tion. If the available signal-to-noise ratio is 20 dB, calculate the maxi-
mum theoretical bit rate and the corresponding baud rate using:
(a) a two-level code
(b) a four-level code
2. How much bandwidth would be required to transmit a DS-1 signal
(1.544 Mb/s) using a four-level code:
(a) assuming a noiseless channel?
(b) with a signal-to-noise ratio of 15 dB?
3. The AFSK system described in the text operates at 1200 bits per second
using an FM signal modulated by tones at 1200 and 2200 Hz, with a fre-
quency deviation of 5 kHz. Calculate the efficiency of this system in
bits per second per hertz of bandwidth by using Carson’s rule (see Chap-
ter 2) to calculate the approximate radio frequency bandwidth for this
system. Is this system bandwidth-efficient?
4. A typical HF radioteletype system uses 170 Hz shift between mark and
space frequencies and a bit rate of 45 bits per second. What would be the
bit rate if GMSK were used for this system?
5. Consider a QPSK system that will transmit three bits of information per
symbol.
(a) How many phase angles are needed?
(b) Draw a vector diagram for such a system.
(c) Would this system have any advantages compared with the dibit
systems described in the text? Any disadvantages?
6. A constellation diagram for a modem is shown in Figure 4.12.
FIGURE 4.12 (
(
(
(
(((( ((((
(
(
(
(
(a) What type of modulation is this?

(b) If the transmitted bit rate is 9600 b/s, what will be the baud rate us-
ing this modem?
7. A microwave radio system uses 256-QAM, that is, there are 256 possible
amplitude and phase combinations.
(a) How many bits per symbol does it use?
(b) If it has a channel with 40-MHz bandwidth, what is its maximum
data rate, ignoring noise?
8. Draw a constellation pattern for a modem that uses eight equally-
spaced phase angles and four equally-spaced amplitude levels. If the
modem operates at 4800 baud, what is its bit rate?
9. An ordinary broadcast television channel has a bandwidth of 6 MHz.
How many FM radio stations (with a bandwidth of 200 kHz each) could
be accommodated in one television channel using FDM?
10. Suppose the TV channel described in the previous problem is used for
voice-quality transmission using SSBSC AM with a bandwidth of 4 kHz
per channel. How many voice channels can be carried?
11. Suppose the TV channel described in the previous problem is used for
digital audio communication. Assume the RF system can support
2 b/s/Hz of RF bandwidth.
(a) What is the total available bit rate?
(b) How many telephone voice signals, at a bit rate of 64 kb/s each,
could be accommodated on this channel using TDM?
(c) How many high-quality stereo audio signals could be accommo-
dated? Assume 16-bit samples, with a sampling rate of 44.1 kSa/s,
and ignore error correction and data compression.
140 ! CHAPTER 4
12. Suppose that a voice signal normally occupies 30 kHz of bandwidth and
has a signal-to-noise ratio of 20 dB. Spread-spectrum techniques are
used to increase its bandwidth to 2 MHz.
(a) What is the signal-to-noise ratio of the spread signal?
(b) What is the processing gain, in decibels?
13. Suppose a frequency-hopping system hops among 500 channels. How
many orthogonal PN sequences are possible?
14. Suppose there is a narrowband analog signal on one of the channels vis-
ited by a frequency-hopping system.
(a) What is the effect on the narrowband signal of the spread-spectrum
signals?
(b) What is the effect on the spread-spectrum signals of the narrow-
band signal?
15. A direct-sequence spread-spectrum system uses FSK with a chipping rate
of 20 to 1. The signal-to-noise ratio for the spread signal is −5 dB (that is,
the signal is 5 dB weaker than the noise in the same bandwidth). If the
data is transmitted at 50 kb/s, calculate:
(a) the chipping rate
(b) the bandwidth occupied by the spread signal if the modulation
scheme used allows 1.5 bits/s/Hz
(c) the signal-to-noise ratio for the despread signal at the receiver
16. A signal has a bit rate of 20 kb/s. Find the baud rate if the signal is trans-
mitted using:
(a) FSK with two frequencies
(b) QPSK with four phase angles
(c) QAM with four phase angles and four amplitudes
17. Ten voice signals are to be multiplexed and transmitted. The analog sig-
nal occupies 4 kHz of bandwidth and can be digitized using a vocoder at
12 kb/s. Calculate the required bandwidth for each of the following pos-
sibilities.
(a) FDMA using analog FM with 12 kHz deviation (use Carson’s rule to
find the bandwidth for one signal). Ignore guard bands between
channels.
(b) FDMA using SSBSC AM. Ignore guard bands.
(c) TDM using GMSK. Assume a noiseless channel.
(d) TDMA using QPSK. Assume a noiseless channel.
(e) CDMA using frequency-hopping with 10 available channels. Use
GMSK and assume a noiseless channel.
(f) CDMA using direct-sequence, QPSK with a chipping rate of 10:1.
Basic Telephony 5
Objectives
( Describe the topology of the switched telephone network.
( Describe the various signals present on a local-loop telephone line and
explain the function of each.
( Describe and compare in-band and out-of-band signaling systems for
telephony.
( Explain the advantages of common-channel signaling.
( Describe Signaling System Seven and explain its use in keeping track
of calls.
( Explain the use of time-division multiplexing in telephony and perform bit
rate calculations with TDM signals.
144 ! CHAPTER 5
' 5.1 Introduction

The public switched telephone system (PSTN) is undoubtedly the larg-
est, and probably the most important communication system in the world.
The reasons for this are contained in those first two words. It is public in the
sense that anyone can connect to it. Because it is switched, it is possible, in
theory at least, for anyone to communicate with anyone else. This makes the
telephone system very different from broadcasting systems and from private
communication networks.
The addition of personal wireless communication to the PSTN has made
it even more ubiquitous. Radio systems that were separate entities are in-
creasingly becoming extensions of the telephone system. Consider, for
instance, the difference between citizens’ band radio and cellular radio. The
former is a separate system, useful only when those who wish to communi-
cate have compatible radios and are within radio range. The latter can con-
nect its user to any telephone in the world, regardless of distance.
The telephone network employs many of the most interesting develop-
ments in communication practice, such as cellular radio, fiber optics, and
digital signal transmission, but it remains in many ways consistent with its
origins in the nineteenth century. This will become obvious when we look
at telephone signaling systems and the voltage and current levels found
on subscriber lines. Compatibility has been maintained in most areas of
the system, so simple dial-type telephones can coexist with modern data-
communication equipment. Though originally intended only for voice
communication, the switched telephone network has been adapted to serve
many other needs, including data communication, facsimile, and even
video.
This chapter introduces the telephone system and describes the ways in
which it can connect with wireless systems. A basic knowledge of ordinary
voice telephony (plain old telephone service or POTS in telephone jargon) will
be very useful as we consider more advanced uses.
' 5.2 Network Topology

Switched networks can be categorized as circuit-switched or packet-
switched. In a circuit-switched network there is a dedicated physical path
from transmitter to receiver for the duration of the communication. The
PSTN is a circuit-switched network. Packet-switched networks route short
bursts of data, called packets, from point to point as needed. A virtual connec-
tion may exist, but it is merely a record of the addresses on the network
between which communication takes place. Successive packets may take
BASIC TELEPHONY ! 145
different paths through the network. The internet is a packet-switched net-

work.
Both types of networks have their advantages. Circuit switching is often
more reliable, especially when it is important that messages arrive quickly
and in the same order in which they were sent, as is the case with telephony.
On the other hand, packet switching can make more efficient use of network
resources. In a circuit-switched network the circuit is often idle. For in-
stance, during a phone call circuits are active in both directions, but most of
the time only one person is talking. It is possible to use packet-switched net-
works for telephony (internet phone is an example), and it looks as if this is
the direction networks are moving.
Wireless networks can be either circuit-switched (cellular phones) or
packet-switched (wireless local-area networks), and we shall examine both
types later in this book.
PSTN Structure Figure 5.1 shows the basic structure, or topology, of a local calling area
(known as a Local Access and Transport Area or LATA) in a typical switched
telephone system.
Each subscriber is normally connected via a separate twisted-pair line,
called a local loop, to a central office, also called an end office, where cir-
cuit switching is done. Actually the term office can be deceiving: in urban ar-
eas, it is quite possible for there to be more than one central office in the
same building.
The central office represents one exchange: that is, in a typical seven-
digit telephone number, all the lines connected to a single central office
begin with the same three digits. Thus there can be ten thousand telephones
FIGURE 5.1
Local access and
transport area
146 ! CHAPTER 5
connected to a central office. Subscribers connected to the same central

office can communicate with each other by means of the central office
switch, which can connect any line to any other line. Modern switches are
digital, so the analog local-loop signals are digitized as they enter the central
office.
The central offices themselves are connected together by trunk lines;
any subscriber can contact any other subscriber within a local calling area.
There are not enough trunks or enough switching facilities for every sub-
scriber to use the system at once, so there is the possibility of overload. This
can make it impossible for a subscriber to place a call, an occurrence that is
known as call blocking. The likely number of simultaneous conversations
is predicted by statistical methods, and should usually be exceeded only
during emergencies.
When there are no available trunks between two central offices, some-
times a connection can be made through a tandem office, which connects
central offices without having any direct connections to individual tele-
phones.
Long-distance calls are completed using a mesh of long-distance switch-
ing centers. The network usually lets the system find a direct route from one
area of the country to the other, and there is never a need for more than
one intermediate switch. Figure 5.2 shows this type of system.
The telephone system in the United States and Canada was formerly a
monopoly, and it still is in much of the world. This is changing. Competi-
tion for long-distance calls is well established, and competition for the local
subscriber loop is just beginning. Each competing long-distance carrier has
its own connection to the local access and transport area. To keep the dia-
gram simple, only two long-distance carriers are shown in Figure 5.2. Each
carrier has its own connection, called a point of presence (POP), to the local
telephone system. Wireless telephone systems are connected to the PSTN in
a similar fashion.
FIGURE 5.2
Long-distance
network
Until now, local-loop competitors have leased their physical connec-

tions from the original local monopoly. As local-loop competitors using
their own cable (for example, cable television companies) enter the market,
they too will require points of presence.
New switching equipment and trunk lines are digital, with time-division
multiplexing used to combine many signals on one line. Fiber optics tech-
nology is increasingly used for trunk lines; other common media include ter-
restrial microwave links, geostationary satellites, and coaxial cable. Short
distances between central offices may be covered using multipair cable,
which has many twisted pairs in one protective sheath. So far, most local
loops are still analog using twisted-pair copper wire, but that seems likely
to change in the future. Eventually the system will be digital from one end to
the other, and most of it will employ fiber-optic cable. However, telephone
equipment is built for maximum reliability and lasts a long time. It is also
expensive, and telephone companies quite naturally expect to use it until
it wears out. It will be a few more years before we have a complete digital
network.
' 5.3 The Local Loop and Its Signals

Normally each individual subscriber telephone is connected to the central
office by a single twisted pair of wires. The wires are twisted to help cancel
their magnetic fields and reduce interference, called crosstalk, between cir-
cuits in the same cable. It is common practice to run a four-conductor cable
to each residence, but only two of these wires (usually red and green) are
used for a single line. The others (black and yellow) allow for the installation
of a second line without running more cable.
Recently there has been a trend toward running multiplexed digital
signals to junction boxes in neighborhoods, switching at that point to ana-
log signals on individual pairs, in an effort to reduce the total amount of
copper cable. There has been some installation of optical fiber, though the
last section of loop to the customer normally remains copper. In the future,
fiber-in-the-loop (FITL) may be used. It will cost more because of the neces-
sity of converting back and forth between electrical and optical signals at
each subscriber location, but the bandwidth will be vastly increased. This
will allow a great number of additional services, such as cable television and
high-speed internet access, to be carried on the same fiber.
The local loop performs several functions. It carries voice signals both
ways. It must also carry signaling information both ways: dialing pulses or
tones to the central office from the customer and dial tones, ringing, busy
148 ! CHAPTER 5
signals, and prerecorded messages from the network to the subscriber. Wire-
less telephones must perform all of these functions, although usually not in
exactly the same way.
In addition to the previously mentioned functions, the twisted pair
must transmit power from the central office to operate the telephone and
ring the bell.
When the phone is on hook (not in use), the central office maintains a
voltage of about 48 V dc across the line. Of the two wires in the twisted pair,
one, normally the green, is designated tip and the other (red), ring. The ring
is connected to the negative side of the supply. Most of the time in electronic
equipment a red wire is positive but not here! The “tip” and “ring” terminol-
ogy dates from the days of manual switchboards; it describes the connec-
tions to the plugs used in these boards. The positive (tip) side of the supply is
grounded.
The central office supply is called the battery. The voltage does, in fact,
derive from a storage battery that is constantly under charge. This allows the
telephone system to function during electrical power outages, whether they
occur at the central office or at the customer’s premises, and has resulted in a
well-deserved reputation for reliability. If wireless systems are to strive for
similar performance, they too need emergency power. This is especially im-
portant for cellular phone and PCS services, because many people subscribe
to these services in order to have emergency communication when needed.
When the phone is on hook, it represents an open circuit to the dc bat-
tery voltage. The subscriber signals the central office that he or she wishes to
make a call by lifting the receiver, placing the instrument off hook. The tele-
phone has a relatively low resistance (about 200 ohms) when off hook,
which allows a dc current to flow in the loop. The presence of this current
signals the central office to make a line available (the telephone is said to
have seized the line). When off hook, the voltage across the telephone drops
considerably, to about 5 to 10 volts, due to the resistance of the telephone
line. Resistance can also be added at the central office, if necessary, to main-
tain the loop current in the desired range of approximately 20 to 80 mA. Fig-
ure 5.3 illustrates this capability.
EXAMPLE 5.1 Y
A local loop has a resistance of 1 kΩ, and the telephone connected to it has
an off-hook resistance of 200 Ω. Calculate the loop current and the voltage
across the telephone when the phone is:
(a) on hook
(b) off hook
FIGURE 5.3 Local loop
SOLUTION
(a) When the telephone is on hook, its dc resistance is infinite so the current
is zero. Since there will be no voltage drop around the loop, except at the
phone itself, the full battery voltage will appear across the phone.
(b) When the phone is off hook, the total loop resistance is
RT = 1000 Ω + 200 Ω = 1200 Ω
Then the loop current is

48 V
I = = 40 mA
1200 Ω
The voltage across the telephone is
V = IR = 40 mA × 200 Ω = 8 V
X
Once a line has been assigned, the office signals the user to proceed by
transmitting a dial tone, which consists of 350 Hz and 440 Hz signals added
together.
Dialing can be accomplished in one of two ways. The old-fashioned ro-
tary dial functions by breaking the loop circuit at a 10 Hz rate, with the num-
ber of interruptions equal to the number dialed. That is, dialing the number
5 causes five interruptions (pulses) in the loop current. This technique is
called pulse dialing and can be emulated by some electronic telephones. The
second and much more efficient way is for the phone to transmit a combina-
tion of two tones for each number. This is officially known as dual-tone
150 ! CHAPTER 5
multi-frequency (DTMF) dialing and is commonly referred to as Touch-

Tone or just tone dialing (the term Touch-Tone is a registered trade mark of
AT&T). Table 5.1 shows the combinations of tones used for each digit. This
system improves efficiency because digits can be transmitted in much less
time than with pulse dialing.
TABLE 5.1 DTMF Frequencies
Frequencies (Hz) 1209 1336 1477 1633
697 1 2 3 A
770 4 5 6 B
852 7 8 9 C
941 * 0 # D
The letters A through D are included in the system specifications but are
not present on ordinary telephones. Some wireless equipment uses them for
special functions.
EXAMPLE 5.2 Y
What frequencies are generated by a telephone using DTMF signaling when
the number 9 is pressed?
SOLUTION
Use Table 5.1. Go across from 9 to find 852 Hz; go up to find 1477 Hz. There-
fore, the output frequencies are 852 Hz and 1477 Hz.
X
Assume for now that the called party is connected to the same central of-
fice as the calling party; that is, they have the same exchange, and the first
three numbers in a typical seven-digit telephone number are the same.
When the switch connects to the called party, it must send an intermittent
ringing signal to that telephone. The standard for the ringing voltage at the
central office is 100 V ac at a frequency of 20 Hz, superimposed on the 48 V
dc battery voltage. Of course, the voltage at the telephone will be less than
that, due to the resistance of the wire in the local loop. In order to respond to
the ac ringing signal when on-hook, without allowing dc current to flow, the
telephone ac-couples the ringer to the line. In a conventional telephone
with an electromechanical ringer, the ringer consists of two coils and a
capacitor in series across the line.
While the called telephone is ringing, the central office switch sends
a pulsed ac voltage, called a ringback signal, to the calling telephone. The
ringback signal consists of 440 and 480 Hz signals added together. When
the called phone goes off-hook, the circuit is complete, the ringing volt-
ages are switched off, and conversation can begin. If the circuit correspond-
ing to the called telephone is in use, a busy signal will be returned to the
caller.
See Table 5.2 for a summary of the signals described so far.
TABLE 5.2 Local-Loop Voltages and Currents
On-hook voltage 48 V dc
Off-hook voltage (at phone) 5–10 V dc, depending on loop resistance
Off-hook current 23–80 mA dc, depending on loop resistance
Dial tone 350 and 440 Hz
Ringing voltage (at office) 100 V ac, 20 Hz, superimposed on 48 V dc
Ringing voltage (at phone) Approximately 80 V ac, superimposed on

48 V dc
Ringback voltage 440 and 480 Hz; pulsed 2 s on, 4 s off
Busy signal 480 Hz and 620 Hz; pulsed 0.5 s on, 0.5 s off
The single twisted-pair local loop is required to carry both sides of the
conversation simultaneously, providing full-duplex communication. This is
called a two-wire system. The rest of the network uses separate transmission
paths for each direction, a topology called a four-wire system. Converting be-
tween the two systems is done using a circuit called a hybrid coil, shown in
Figure 5.4. The same thing can be done electronically. Signals from the trans-
mitter will add at the line and cancel at the receiver. Similarly, signals com-
ing in on the line will cancel at the transmitter and add at the receiver.
Deliberately unbalancing the circuit allows a small portion of the transmit-
ter signal to reach the receiver, creating a sidetone that lets the user know
the line is active and hear what is being transmitted. Hybrid coils are used
both in the phone and in the central office line cards.
152 ! CHAPTER 5
FIGURE 5.4
Hybrid coil
' 5.4 Digital Telephony

The telephone system was completely analog at its beginning in the nine-
teenth century, of course. Over the past 30 years or so, it has gradually been
converted to digital technology and the process is not yet complete. The dig-
ital techniques were originally designed to work in conjunction with the
existing analog system. This, and the fact that many of the standards for dig-
ital technology are quite old, should help the reader understand some of the
rather peculiar ways things are done in digital telephony.
The basics of digital transmission of analog signals by pulse-code modu-
lation (PCM) were discussed in Chapter 3. We can summarize the results
here as they apply to the North American system. The numbers vary slightly
in some other countries but the principles are the same.
The analog voice signal is low-pass filtered at about 3.4 kHz and then
digitized, using 8-bit samples at a sampling rate of 8 kHz. The signal is com-
pressed, either before or after digitization, to improve its signal-to-noise
ratio. The bit rate for one voice signal is then
ƒ b (voice) = 8 bits/sample × 8000 samples/second = 64 kb/s (5.1)

The sample rate is determined by the maximum frequency to be trans-
mitted, which was chosen for compatibility with existing analog FDM trans-
mission (which uses SSBSC AM with a bandwidth of 4 kHz per channel,
including guardbands between channels). An upper frequency limit of
about 3.4 kHz has long been considered adequate for voice transmission.
In Chapter 3 we found that a much lower bit rate could be used for
telephone-quality voice using data compression and vocoders. These tech-
niques are not employed in the ordinary telephone system, though data
compression is used in special situations where bandwidth is limited and

expensive, as in intercontinental undersea cables.
The connection of wireless systems to the PSTN requires conversion of
standards in both directions. In general, wireless systems use ordinary tele-
phone quality as a guide, though many of these systems fall short of what
has been considered toll quality, that is, good enough to charge long-distance
rates for. Until now, users have been so delighted to have portable tele-
phones that they have been willing to put up with lower quality. This situa-
tion is changing quickly now that wireless phones are everywhere.
Time-Division In Chapter 3 we discussed the simplest form of time-division multiplexed

Multiplexing telephone signal. The DS-1 signal frame has one sample (8 bits) from each of
24 telephone channels plus one framing bit. This gives it a bit rate of
ƒ b (DS-1) = 24 × 64 kb/s + 1 × 8 kb/s = 1.544 Mb/s. (5.2)

When this signal is transmitted over copper wire, the result is known as a
T-1 carrier. That is, the signal includes only the coding into ones and zeros,
while the carrier also includes the voltage levels used. See Figure 5.5 for a re-
view of the DS-1 frame, which we looked at in Chapter 4.
FIGURE 5.5
DS-1 signal
A DS-1 signal can equally well be used for data communication. A single
time slot may be used or any number of time slots can be combined for
higher bit rates.
The framing bits are used to enable the receiver to determine which bit
(and in what sample) is being received at a given time. In addition, the re-
ceiver must often distinguish between frames in order to decode signaling
information. In one frame out of every six, each of the least significant bits
in the 24 samples may be used for signaling information rather than as part
of the PCM signal. This information includes on/off hook status, dial tone,
dialed digits, ringback, and busy signal. This bit robbing for signaling re-
sults in a very slight degradation of voice signal quality; for instance, the
signal-to-noise ratio is reduced by about two decibels.
154 ! CHAPTER 5
The frames are divided into groups of 12 with different signaling infor-
mation in the sixth and twelfth frames, known as the A and B frames, in a
sequence. A group of twelve frames is called a superframe. As a result, the
receiver is required to count frames up to 12. To allow the receiver to accom-
plish this, the framing bit alternates between two sequences, 100011 and
011100. The underlined bits indicate the A and B signaling frames, respec-
tively. The “stolen” signaling bits can be used to indicate basic line states
such as on-hook and off-hook, ringing, and busy signals.
Unfortunately the effect of bit robbing on data transmission is much
greater than it is for voice. Occasional bit errors are acceptable in voice sig-
nals, but certainly not in data. To avoid errors when bit robbing is used, one
bit from each 8-bit sample is discarded in every frame. This reduces the data
capacity of one voice channel from 64 kb/s to 56 kb/s. Of course, channels
can be combined for higher rates, but the loss in throughput is very substan-
tial. Bit robbing can be eliminated by using common-channel signaling,
which is described in the next section.
Digital Signal The DS-1 signal and T-1 carrier described earlier represent the lowest level in
Hierarchy a hierarchy of TDM signals with higher bit rates. All of these signals contain
PCM audio signals, each sampled 8,000 times per second. As the number of
multiplexed voice signals increases, so does the bit rate. This requires that
the channel have a wider frequency response and that variations of time de-
lay with frequency be held to a low level. Twisted-pair lines, when specially
conditioned, can be used for the T1 and T2 carriers, but higher data rates
require channels with greater bandwidth, such as coaxial cable, microwave
radio, or optical fiber. See Table 5.3 for more details.
TABLE 5.3 Digital Signal Hierarchy
Carrier Signal Voice Channels Bit Rate (Mb/s)
T1 DS-1 24 1.544
T1C DS-1C 48 3.152
T2 DS-2 96 6.312
T3 DS-3 672 44.736
T4 DS-4 4032 274.176
T5 DS-5 8064 560.16

A glance at the table shows that the math does not seem to be exact. For
instance, a DS-1C signal carries as many voice channels as two DS-1 signals,
but the bit rate is more than twice as great. The difference is:
3.152 Mb/s − 2 × 1.544 Mb/s = 64 kb/s
The extra bits have several uses. They provide synchronization and
framing for the demultiplexer. There are also extra bits called stuff bits which
are added during multiplexing to compensate for differences between the
clock rates of the tributaries and the multiplexer. If the tributary clock rate is
slow, more stuff bits will be added to build up the bit rate; if it is fast, fewer
stuff bits are needed. This bit stuffing is more formally called justification.
Figure 5.6 is an example of the creation of a DS-3 signal by multiplexing
other signals.
FIGURE 5.6 Creation of DS-3 signal

156 ! CHAPTER 5
' 5.5 Telephone-Network Signaling

We have already looked at some of the control and supervisory signals used
with the telephone system. The local-loop signals such as DTMF tones, dial
tone, busy signal, and ringback signal are examples. All of these signals use
the same channel as the voice, but not at the same time. Since they use the
same channel, they are called in-channel signals. Their frequencies are also
in the same range as voice frequencies; they can be heard by the user, so they
are also referred to as in-band signals.
We also noted that the telephone instrument communicates its off-hook
or on-hook status to the central office by the presence or absence, respec-
tively, of a dc current. This signal is in-channel because it uses the same pair
of wires as the voice, but it is out-of-band because the dc current is not in the
same frequency range as voice signals. Consequently, the central office can
receive the off-hook signal continuously, even during the call.
Traditionally, similar methods have been used within the network for
such purposes as communicating the number of the calling party to billing
equipment, determining which trunk lines are idle (that is, ready for use),
and so on. Early systems used dc currents and dial pulses, as with local loops.
Later versions used either a switched single-frequency tone at 2600 Hz
(called SF signaling) or, still later, combinations of tones similar to but not
the same as the DTMF system and known as MF, (multi-frequency) signal-
ing. For a time, that system was plagued by fraud as people ranging from am-
ateur “phone phreakers” to members of organized crime used so-called “blue
boxes” to duplicate network signaling tones and make long-distance calls
without paying for them. However, changes to the network soon eliminated
or at least greatly reduced this problem.
In-channel but out-of-band signaling is also used. A tone at 3825 Hz can
be sent along a long-distance network. It will pass through the allotted
4-kHz channel for each call, but will be filtered out before it reaches the cus-
tomer telephone. This type of signal replaces the dc loop current as a means
of indicating whether the line is in use, since long-distance circuits, unlike
local loops, do not have dc continuity.
We have seen that the digital signal hierarchy makes available a few
bits, “stolen” from the voice channel, for signaling. These can be used to
replace the in-channel tones. However, there are insufficient bits to use for
call routing, call display, billing, and other functions of a modern telephone
network.
Signaling System Recently the trend has been to use a completely separate data channel to
Seven transmit control information between switches. This common-channel sig-
naling reduces fraud, since users have no access to the control channels, and
also allows a call to be set up completely before any voice channels are used.
The state of the whole network can be known to the control equipment, and
the most efficient routes for calls can be planned in advance. Com-
mon-channel signaling also makes such services as calling-number identifi-
cation much more practical.
The current version of common-channel signaling is signaling system
seven (SS7). It was introduced to the Bell System in the United States in 1980
and has become, with minor variations, a worldwide system. SS7 is a packet-
switched data network linking central offices to each other, to long-distance
switching centers, and to centralized databases used for such purposes as call
display, credit card validation, voice mail, 800 and 900 number routing and
(most interesting for our purposes) cellular and PCS telephone roaming in-
formation. SS7 allows much more data to be sent more quickly, and with less
interference with voice signals, than older signaling schemes involving
in-channel signals.
SS7 uses dedicated 64 kb/s data channels. Usually one digital voice chan-
nel in each direction is reassigned for this purpose; the data rate is the same
as for a voice channel to accommodate this. If necessary, an analog channel
with modems can be used. One 64 kb/s signaling channel can handle the sig-
naling requirements of many voice channels. Figure 5.7 shows how SS7 con-
nects to the rest of the network.
With SS7, calls can be set up with no need to tie up a long-distance voice
channel until the connection is made. Since analog local loops do not sup-
port common-channel signaling, it is necessary to tie up a voice connection
from the subscriber to the central office. If ISDN, described in the next sec-
tion, is used, voice and control signals can be kept completely separate.
FIGURE 5.7
Signaling system
seven (SS7)
158 ! CHAPTER 5
' 5.6 Digital Local Loops

The analog local loop dates back to the earliest days of telephony, and is cer-
tainly outdated. Subscribers now need much greater bandwidth for uses
such as high-speed internet access or interactive video than was the case a
few years ago. One way to get this bandwidth is to replace the twisted-pair
loop with coaxial cable or optical fiber. However, the capacity of a twisted
pair is much greater than is required for the single analog signal, with a
bandwidth under 4 kHz that it usually carries. Most of the cost of the local
loop is in the labor to install it, rather than the cost of the wire itself, so when
greater bandwidth is required it makes good sense to redesign the system
to increase the capacity of the existing wiring rather than replace the wire.
Here is a brief look at two ways of increasing the capacity of twisted-pair
local loops.
Integrated The integrated services digital network (ISDN) concept is designed to

Services Digital allow voice and data to be sent in the same way along the same lines. Cur-
Network (ISDN) rently, most subscribers are connected to the switched telephone network
by local loops and interface cards designed for analog signals. This is reason-
ably well-suited to voice communication, but data can be accommodated
only by the use of modems.
As the telephone network gradually goes digital, it seems logical to send
data directly over telephone lines without modems. If the local loop could
be made digital, with the codec installed in the telephone instrument, there
is no reason why the 64 kb/s data rate required for PCM voice could not also
be used for data, at the user’s discretion.
The integrated services digital network concept provides a way to stan-
dardize the above idea. The standard encompasses two types of connections
to the network. Large users connect at a primary-access point with a data rate
of 1.544 Mb/s. This, you will recall, is the same rate as for the DS-1 signal
described earlier. It includes 24 channels with a data rate of 64 kb/s each.
One of these channels is the D (data) channel and is used for common-
channel signaling, that is, for setting up and monitoring calls. The other 23
channels are called B (bearer) channels and can be used for voice or data, or
combined, to handle high-speed data or digitized video signals, for example.
Individual terminals connect to the network through a basic interface at
the basic access rate of 192 kb/s. Individual terminals in a large organization
use the basic access rate to communicate with a private branch exchange
(PBX), a small switch dedicated to that organization. Residences and small
businesses connect directly to the central office by way of a digital local
loop. Two twisted pairs can be used for this, though more use of fiber optics
is expected in the future.
Basic-interface users have two 64 kb/s B channels for voice or data, one
16 kb/s D channel, and 48 kb/s for network overhead. The D channel is used
to set up and monitor calls and can also be employed for low-data-rate appli-
cations such as remote meter-reading. All channels are carried on one physi-
cal line, using time-division multiplexing. Two pairs are used, one for signals
in each direction.
Figure 5.8 shows typical connections to the ISDN. The primary interface
is known as a T type interface, and the basic interface has the designation S.
Terminal equipment, such as digital telephones and data terminals, de-
signed especially for use with ISDN, is referred to as TE1 (terminal equipment
type 1), and connects directly to the network at point S. The network termi-
nation equipment designated NT2 could be a PBX, a small computer net-
work called a local area network, or a central office. Terminal equipment
not especially designed for ISDN is designated TE2 (terminal equipment type
2) and would need a terminal adapter (TA) to allow it to work with the ISDN.
Examples of type two equipment would be ordinary analog telephones, ordi-
nary fax machines, and personal computers with serial ports. Each of these
would need a different type of terminal adapter.
FIGURE 5.8
ISDN access
Implementation of the ISDN has been slow, leading some telecommuni-

cations people to claim, tongue-in-cheek, that the abbreviation means “it
still does nothing.” Several reasons can be advanced for this. First, convert-
ing local loops to digital technology is expensive, and it is questionable
whether the results justify the cost for most small users. Residential tele-
phone users would not notice the difference (except that they would have
to replace all their telephones or buy terminal adapters), and analog lines
with low-cost modems or fax machines are quite satisfactory for occasional
data users. Newer techniques like Asymmetrical Digital Subscriber Line
(ADSL), which is described in the next section, and modems using ca-
ble-television cable have higher data rates and are more attractive for resi-
dential and small-office data communication. Very large data users often
need a data rate well in excess of the primary interface rate for ISDN. They
are already using other types of networks. It appears possible that the ISDN
standard is becoming obsolete before it can be fully implemented.
160 ! CHAPTER 5
With this in mind, work has already begun on a revised and improved
version of ISDN called broadband ISDN (B-ISDN). The idea is to use much
larger bandwidths and higher data rates, so that high-speed data and video
can be transmitted. B-ISDN uses data rates of 100 to 600 Mb/s.
Asymmetrical The idea behind the asymmetrical digital subscriber line (ADSL) is to use the
Digital Subscriber frequencies above the voice range for high-speed data while leaving the use
Line (ADSL) of the local loop for analog telephony intact. This allows the subscriber to
use conventional analog telephones without special adapters, while simulta-
neously sending and receiving high-speed data. The word asymmetrical in
the name refers to the fact that the system is designed for faster communica-
tion from the network to the subscriber than from the subscriber to the net-
work. Typical uses for ADSL include internet access and interactive
television; for both of these the subscriber needs to receive data at a faster
rate than it needs to be transmitted.
There are many types of ADSL using different-frequency carriers for
downstream (to the subscriber) and upstream (from the subscriber) data.
Downstream data rates vary from about 1 to 8 Mb/s, with upstream rates
from 160 to 640 kb/s. Most systems use FDM to separate upstream from
downstream data, as illustrated in Figure 5.9. Note that the downstream
signal has wider bandwidth, as would be expected considering its higher
data rate.
Most ADSL systems require the installation of a splitter at the customer
premises to separate voice and data signals, but one variety, known as DSL
Lite, requires no splitter and provides a downstream rate of up to 1.5 Mb/s.
ADSL has the advantage over ISDN in that data signals do not have to go
through the central office switch. This means that a user can be connected
to the internet on a continuous rather than a dial-up basis. It also reduces
traffic on the switched network.
FIGURE 5.9
Spectrum of a
typical ADSL system
( The public switched telephone network is of great importance for wireless

communication because it links wireless users with other users—both
wireless and wired.
( Many of the specifications for telephone equipment have their basis in
nineteenth century technology and have remained almost unchanged for
reasons of compatibility.
( Common-channel signaling allows the telephone network to set up calls
economically, without tying up voice lines, and allows the system to pro-
vide a considerable amount of data about calls.
( Wireless systems that connect with the telephone system can use its com-
mon-channel signaling system to carry information about subscribers
roaming out of their local service areas.
( Modern telephone systems are digital except for the subscriber loop, and
progress is being made towards digitizing the local loop as well.
( Digital telephone signals are time-division multiplexed with a data rate of
64 kb/s per voice channel.
( Signaling can be done by robbing bits from the voice signal but com-
mon-channel signaling using Signaling System Seven is more efficient
and allows more information to be shared.
( ISDN allows the telephone system to be completely digital from end
to end.
( ADSL allows for a conventional analog voice local loop and a high-speed
data link to be combined on one twisted pair.
( Key Terms
Asymmetrical Digital Subscriber Line (ADSL) method of providing
high-speed data transmission on twisted-pair telephone loops by
using high-frequency carriers
B (bearer) channels in ISDN, channels that carry subscriber
communication
bit robbing use of bits that normally carry payload information for
other purposes, such as controlling the communication system
bit stuffing addition of bits to a bitstream to compensate for timing
variations
call blocking failure to connect a telephone call because of lack of
system capacity
162 ! CHAPTER 5
central office switch in a telephone system that connects to local

subscriber lines
circuit-switched network communication system in which a dedicated
channel is set up between parties for the duration of the
communication
common-channel signaling use of a separate signaling channel in a
telephone system, so that voice channels do not have to carry
signaling information
crosstalk interference between two signals multiplexed into the same
channel
D (data) channel in ISDN, a communication channel used for setting up
calls and not for user communication
dual-tone multi-frequency (DTMF) dialing signaling using combinations
of two audio tones transmitted on the voice channel
end office see central office
fiber-in-the-loop (FITL) use of optical fiber for telephone connections to
individual customers
hybrid coil a specialized transformer (or its electronic equivalent) that
allows telephone voice signals to travel in both directions
simultaneously on a single twisted-pair loop
in-band signals control signals sent in a voice channel at voice frequencies
in-channel signals control signals using the same channel as a voice signal
integrated services digital network (ISDN) telephone system using
digital local loops for both voice and data, with the codec in the
telephone equipment
justification addition of bits to a digital signal to compensate for
differences in clock rates; informally known as bit stuffing
local access and transport area (LATA) in a telephone system, the area
controlled by one central office switch
local area network a small data network, usually confined to a building
or cluster of buildings
local loop in a telephone system, the wiring from the central office to an
individual customer
out-of-band in telephone signaling, a control signal that is outside the
voice frequency range
packet-switched network a communication system that works using data
divided into relatively short transmissions called packets; these are
routed through the system without requiring a long-term connection

between sender and receiver
point of presence (POP) place where one telephone network connects to
another
private branch exchange (PBX) small telephone switch located on
customer premises
public switched telephone system (PSTN) the ordinary public wireline
phone system
ringback signal in telephony, a signal generated at the central office and
sent to the originating telephone to indicate that the destination
telephone is ringing
sidetone in telephony, the presence in the receiver of sounds picked up
by the transmitter of the same telephone
signaling system seven (SS7) system used in telephony which transmits
all call setup information on a packet-data network that is separate
from the voice channels used for telephone conversations
tandem office telephone switch that connects only to other switches,
and not to individual customers
trunk lines transmission line carrying many signals, either on multiple
pairs or multiplexed together on a single twisted-pair, coaxial cable, or
optical fiber
( Questions
1. Explain briefly how the telephone network differs from a broadcasting
network.
2. Explain the difference between circuit-switched and packet-switched
networks. Is the PSTN mainly circuit-switched or packet-switched?
3. What is a central office?
4. What is the difference between a tandem office and an end office?
5. What is a trunk line?
6. What is meant by a LATA?
7. How do wireless telephone providers connect to the wired telephone
network?
8. How has the breakup of the Bell monopoly changed the North Ameri-
can telephone network?
9. What is meant by call blocking, and why does it happen?
164 ! CHAPTER 5
10. How many wires are needed (for a single line) from the individual tele-
phone set to the central office?
11. How many wires are normally contained in the cable from an individual
residence subscriber to the network? Why is this number different from
the answer to Question 10 above?
12. Explain the meaning of the terms tip and ring. Which has negative
polarity?
13. Explain how pulse dialing works.
14. What is meant by DTMF dialing, and why is it better than pulse dialing?
15. What is the function of the hybrid coil in a telephone instrument?
16. What is sidetone and why is it used in a telephone instrument?
17. Approximately how much bandwidth, at baseband, is needed for one
channel of telephone-quality audio?
18. List the steps required in originating a local call. Include the appropri-
ate voltages and frequencies that appear at the telephone instrument.
19. Describe the difference between in-band and out-of-band signaling,
and give an example of each.
20. Describe the difference between in-channel and common-channel sig-
naling. Which is the more modern system?
21. How does common-channel signaling reduce the vulnerability of the
telephone system to fraudulent use?
22. Name some types of information that are carried by Signaling System
Seven.
23. What type of data channel is used by SS7?
24. What type of modulation is used in FDM telephony?
25. What type of modulation is used in TDM telephony?
26. What is meant by bit robbing? What is its function and why is it undesir-
able for data connections?
27. What is meant by bit stuffing? When and why is it necessary?
28. Compare basic-rate ISDN and ADSL as technologies for voice telephony.
Compare number of lines and type of equipment needed.
29. Compare basic-rate ISDN and ADSL as technologies for data communi-
cation. Compare data rates and connection type.
( Problems
1. Suppose the voltage across a telephone line, at the subscriber, drops
from 48 V to 10 V when the phone goes off hook. If the telephone in-
strument has a resistance of 200 ohms when off hook and represents an
open circuit when on hook, calculate:
(a) the current that flows when the phone is off hook
(b) the combined resistance of the local loop and the power source at
the central office
2. The local loop has a resistance of 650 ohms and the telephone instru-
ment has a ringer voltage of 80 volts when the voltage at the central of-
fice is 100 V. Calculate the impedance of the ringer in the telephone.
3. Find the DTMF frequencies for the number 8.
4. What number is represented by tones of 770 and 1209 Hz, in the DTMF
system?
5. Calculate the overhead of a DS-4 signal:
(a) in bits per second
(b) as a percentage of the total bit rate
6. By what percentage does the use of bit robbing reduce the data capacity
of a DS-1 signal?
7. What is the proportion of overhead in a basic rate ISDN signal? (Assume
the D channel is part of the overhead.) Compare with the overhead in a
DS-1 signal.
8. The fastest modem for use on an analog telephone line operates at 56
kb/s in both directions. If an ADSL system has an upstream rate of 640
kb/s and a downstream rate of 1.5 Mb/s, by what factor does it exceed
the modem data rate in each direction?
9. Compare the data rate available with basic-rate ISDN and that available
with the ADSL system described in the previous problem. By what factor
does ADSL exceed basic-rate ISDN in each direction:
(a) if a voice call is being made simultaneously with data?
(b) if no voice call is being made so that the whole available ISDN rate
can be used for data?
10. Telephone signals can be carried by radio using either analog or digital
modulation schemes. Compare the bandwidth required to carry each of
the following signals:
(a) an analog voice signal with a baseband bandwidth of 4 kHz using
SSBSC AM (often used for terrestrial microwave links where many
voice signals are transmitted together)
166 ! CHAPTER 5
(b) the same analog voice signal using FM with a frequency deviation of
12 kHz (used for cellular phones)
(c) a standard digital voice signal using QPSK and assuming a channel
with a signal-to-noise ratio of 20 dB. (Your answer to this question
may suggest why data compression and vocoders are often used
when digital signals are to be transmitted by radio.)
Transmission Lines
and Waveguides
6
Objectives
( Give several examples of transmission lines and explain what parameters of
a transmission line must be considered as the frequency increases.
( Define characteristic impedance and calculate the impedance of a coaxial
or open-wire transmission line.
( Define reflection coefficient and standing-wave ratio and calculate them in
practical situations.
( Calculate losses on transmission lines.
( Name the dominant mode in a rectangular waveguide and describe its
characteristics.
( Calculate the cutoff frequency, phase and group velocity, guide
wavelength, and characteristic impedance, for the TE10 mode in a
rectangular waveguide.
( Describe several methods for coupling power into and out of waveguides.
( Explain the operation of passive microwave components, including
waveguide bends and tees, resonant cavities, attenuators, and loads.
( Describe the operation and use of circulators and isolators.
( Explain the importance of impedance matching with respect to
transmission lines.
166 ! CHAPTER 6
' 6.1 Introduction

Radio communication systems rely on the propagation of electromagnetic
waves through air and space. The waves also travel along transmission lines
and waveguides en route to and from antennas. In order to understand these
systems, we must first learn something of the nature of electromagnetic radi-
ation and then consider the propagation of these waves along conductors
and in space. In this chapter we will study the propagation of electromag-
netic waves on transmission lines and in waveguides. The next chapter deals
with propagation in space.
' 6.2 Electromagnetic Waves

Radio waves are one form of electromagnetic radiation. Other forms include
infrared, visible light, ultraviolet, x-radiation, and gamma rays. Figure 6.1
shows the place of radio waves at the low-frequency end of the frequency
spectrum.
Electromagnetic radiation, as the name implies, involves the creation of
electric and magnetic fields in free space or in some physical medium. When
the waves propagate in a transmission line consisting of a pair of conduc-
tors, the fields can be represented as voltages and currents. When the propa-
gation is in a dielectric medium or in free space, we need to work with the
fields directly. Both types of wave propagation are essential to wireless com-
munication.
Sinusoidal waves of voltage and current should be familiar from basic
electricity and from our studies of various signals earlier in this book. When
a sine wave propagates along a line or through space, it varies sinusoidally in
both time and space. Just as the time taken for one repetition of a wave is
called the period, the distance between identical points on the wave, at the
FIGURE 6.1 The electromagnetic spectrum

TRANSMISSION LINES AND WAVEGUIDES ! 167
same instant of time, is known as the wavelength of the signal. Another way
of putting this is to say that the wavelength is the distance a wave travels in
one period.
To arrive at an expression for wavelength on a line, we start with the def-
inition of velocity:
d
v =
t
where
v = velocity in meters per second
d = distance in meters
t = time in seconds
Since we are trying to find wavelength, we substitute the period T of the
signal for t, the wavelength λ for d, and v p , the propagation velocity, for v.
This gives us
λ
vp =
T
More commonly, frequency is used rather than period. Since
1
ƒ =
T
the above equation becomes
v p = ƒλ (6.1)
Equation (6.1) applies to any kind of wave, not just radio waves. The
propagation velocity varies greatly for different types of waves. For electro-
magnetic waves in free space, vp is usually represented as c, where c is approx-
imately equal to 300 × 106 meters per second.
EXAMPLE 6.1 Y
Find the wavelength of a radio wave in free space, with a frequency of 100 MHz.
SOLUTION
From Equation (1.1), and noting that here v = c:
c
c = ƒλ λ =
ƒ
300 × 106 m/s
= = 3m
100 × 106 Hz
X
168 ! CHAPTER 6
' 6.3 Propagation on Transmission Lines

When dc or low-frequency signals are considered, conducting paths such
as wire cables are usually ignored. At most, the resistance of the conductors
may be taken into consideration. When the path length is an appreciable
fraction of the signal wavelength, however, the techniques described in
this section must be used to analyze the path as a transmission line. Thus,
transmission line techniques must be used for long lines, high frequen-
cies, or a combination of the two. In general, if a line is longer than about
one-sixteenth of the signal wavelength, it should be considered a transmis-
sion line.
Almost any configuration of two or more conductors can operate as a
transmission line, but there are several types that have the virtues of being
relatively easy to study as well as very commonly used in practice.
Figure 6.2 shows some coaxial cables, in which the two conductors are
concentric, separated by an insulating dielectric. Figure 6.2(a) shows a
solid-dielectric cable; the one in Figure 6.2(b) uses air (and occasional plastic
FIGURE 6.2
Coaxial cables
(a) Solid dielectric
Air dielectric.
(Courtesy of Andrew
Corporation. All Rights
Reserved. Reprinted with
Permission.) (b) Air dielectric
spacers) for the dielectric. When high power is used it is important to keep
the interior of the line dry, so these lines are sometimes pressurized with
nitrogen to keep out moisture. Coaxial cables are referred to as unbalanced
lines because of their lack of symmetry with respect to ground (usually the
outer conductor is grounded).
Parallel-line transmission lines are also common. They are not used for
signal transmission as frequently as coaxial cable because they tend to radi-
ate energy into space and also to absorb interfering signals. However, paral-
lel lines are often used as part of antenna arrays. Ordinary television twin-
lead, as shown in Figure 6.3(a), is an example of a parallel transmission line.
Parallel lines can be shielded to reduce interference problems, as illustrated
in Figure 6.3(b).
Parallel lines are usually operated as balanced lines; that is, the imped-
ance to ground from each of the two wires is equal. This ensures that the
currents in the two wires will be equal in magnitude and opposite in sign, re-
ducing radiation from the cable and its susceptibility to outside interfer-
ence. Coaxial cable, on the other hand, is inherently unbalanced and should
be used with the outer shield grounded. When properly used, there is no
connection between the signal currents in the cable, which travel on the
outside of the inner conductor and the inside of the outer conductor,
and any interference currents, which are only on the outside of the outer
conductor.
Twisted pairs of wires are often used as transmission lines for relatively
low frequencies, because of their low cost and because of the large amount
of twisted-pair line that is already installed as part of the telephone sys-
tem. Transmission-line techniques do not need to be used at audio frequencies
FIGURE 6.3 Parallel-line cables

170 ! CHAPTER 6
unless the distance is great. When higher frequencies are used, as when tele-
phone signals are multiplexed using TDM or twisted-pair line is used for
local-area computer networks, these techniques are required. Twisted-pair line
is illustrated in Figure 6.3(c).
FIGURE 6.3
(continued)
(c) Twisted-pair cable
In the microwave region, wavelengths are short (millimeters to centi-

meters) and even relatively short printed-circuit board traces exhibit trans-
mission-line effects. Often these traces are carefully designed to have known
characteristics. There are two main types of such lines. Microstrip lines
consist of a circuit-board trace on one side of a dielectric substrate (the
circuit board) and a ground plane on the other side. Striplines have a
ground plane on each side of the board with a conducting trace in the center.
Both of these lines are illustrated in Figure 6.4. Of the two, microstrips are
easier to build and to connect to components, but striplines radiate less
energy.
FIGURE 6.4 Stripline and microstrip lines

Electrical In the analysis of transmission lines it is necessary to consider the resistance

Model of a of the conductors, the conductance of the dielectric separating them, and
Transmission Line the inductance and capacitance of the line. However, it is not sufficient to
measure the total value of each quantity for the length of line in use. Rather,
the line must be divided into small (ideally, infinitesimally small) sections
and the factors listed above must be considered for each section. Another
way of saying this is that we must use distributed rather than lumped con-
stants in our analysis.
Models for infinitesimal sections of both balanced and unbalanced lines
are shown in Figure 6.5. The figure allows for R, the resistance of the wire;
G, the conductance of the dielectric; L, the series inductance; and C, the
shunt capacitance. The values for all of these must be given per unit length;
for example, the resistance can be expressed in ohms per meter. How-
ever, the length of line we are considering is certainly much smaller than a
meter. In fact, the idea is to allow the section to shrink until its length is
infinitesimal.
At dc and low frequencies, the inductance has no effect because its
reactance is very small compared with the resistance of the line. Similarly,
the reactance of the shunt capacitance is very large, so the effect of the
capacitance is negligible as well. The line is characterized by its resistance
and possibly by the conductance of the dielectric, though this can usually be
neglected.
FIGURE 6.5
Transmission-line
models
172 ! CHAPTER 6
As the frequency increases, the inductance and capacitance have more

effect. The higher the frequency, the larger the series inductive reactance
and the lower the parallel capacitive reactance. In fact, it is often possible at
high frequencies to simplify calculations by neglecting the resistive ele-
ments and considering only the inductance and capacitance of the line.
Such a line is called lossless, since the inductive and capacitive reactances
store energy but do not dissipate it.
Characteristic Consider a lossless line as described in the previous section. Suppose that a
Impedance voltage step is applied to one end of an infinite length of such a line, as
shown in Figure 6.6. Energy from the source will move outward along the
line as the capacitors and inductors gradually become charged with energy.
Since the line is infinite in length, the charging process will never be com-
pleted, and a constant current and voltage will appear at the source.
FIGURE 6.6
Step input to
infinite line
The ratio between voltage and current has the dimensions of imped-
ance, expressed in ohms. For a lossless line, this ratio is a real number repre-
senting a resistance, even though the lossless line has no actual resistance.
Instead of being dissipated as heat, as in a resistor, the energy from the
source continues to move down the line forever, since its length is infinite.
The ratio just described is called the characteristic impedance of the
line. It can be shown that for any transmission line,
R + jωL
Z0 = (6.2)
G + jωC
where
Z 0 = characteristic impedance in ohms
R = conductor resistance in ohms per unit length
j = √−1
L = inductance in henrys per unit length
G = dielectric conductance in siemens per unit length

C = capacitance in farads per unit length
ω = frequency in radians per second
In general the impedance is complex and is a function of frequency as
well as the physical characteristics of the line. For a lossless line, however, R
and G are zero and Equation (6.2) simplifies to
L
Z0 = (6.3)
C
Of course, there is no such thing as a completely lossless line, but many
practical lines approach the ideal closely enough that the characteristic im-
pedance can be approximated by Equation (6.3). This is especially true at
high frequencies: as ω gets larger, the values of R and G become less signifi-
cant in comparison with ωL and ωC. For this reason, Equation (6.3) is often
referred to as the high-frequency model of a transmission line. Equation (6.3)
gives a characteristic impedance that is a real number and does not depend
on frequency or the length of the line, but only on such characteristics as the
geometry of the line and the permittivity of the dielectric. For coaxial cable,
the characteristic impedance is given by:
138 D
Z0 = log (6.4)
∈r d
where
Z 0 = characteristic impedance of the line
D = inside diameter of the outer conductor
d = diameter of the inner conductor
∈ r = relative permittivity of the dielectric, compared with that of
free space. ∈ r is also called the dielectric constant.
See Figure 6.7(a) for an illustration of the geometry. Here it is quite obvi-
ous that increasing the dielectric constant, or reducing the diameter of the
cable, reduces the characteristic impedance.
FIGURE 6.7
Cable cross
sections
174 ! CHAPTER 6
EXAMPLE 6.2 Y
Find the characteristic impedance of a coaxial cable using a solid polyethyl-
ene dielectric having ∈ r = 2.3, with an inner conductor 2 mm in diameter
and an outer conductor 8 mm in inside diameter.
SOLUTION
138 D
Z0 = log
∈r d
138 8
= log
2.3 2
= 54.8 Ω
X
In practice, it is usually not necessary to use Equation (6.4) to find the

impedance of a coaxial cable, since this is part of the cable specifications. In
fact, the impedance is often marked right on the insulating jacket. For coax-
ial cable, there are a few standard impedances that fulfill most requirements.
Some examples are shown in Table 6.1.
TABLE 6.1 Coaxial Cable Applications
Impedance Typical Type

(ohms) Application Numbers
50 radio transmitters RG-8/U
50 communication receivers RG-58/U
75 cable television RG-59/U
93 computer networks RG-62/U
Parallel line normally has a dielectric consisting mostly of air. Spacers

used with open-wire line or the plastic separating the conductors in twin-
lead will raise the effective dielectric constant slightly, but a reasonable ap-
proximation to the impedance can be derived by assuming an air dielectric:
D
Z 0 = 276 log (6.5)
r
where
D = separation between conductors (center to center)
r = conductor radius
See Figure 6.7(b).
EXAMPLE 6.3 Y
An open-wire line uses wire with a diameter of 2 mm. What should the wire
spacing be for an impedance of 150 Ω?
SOLUTION
First convert the diameter to a radius of 1 mm. Next rearrange Equation (6.5)
to express D as the unknown.
D
Z 0 = 276 log
r
D Z0
log =
r 276
D Z
= log −1 0
r 276
Z0
D = r log −1
276
150
= 1 mm × log −1
276
= 35
. mm
X
The impedance equations for stripline and microstrip transmission lines

are somewhat more complicated. A number of approximations can be used,
but practically all design work with these lines is done with the aid of spe-
cialized computer programs, which have the equations built in. The approx-
imations will not be used here.
Propagation The speed at which energy is propagated along a transmission line is always
Velocity and less than the speed of light. It varies from about 66% of this velocity on coax-
Velocity Factor ial cable with solid polyethylene dielectric, through 78% for polyethylene
foam dielectric, to about 95% for air-dielectric cable. Rather than specify the
176 ! CHAPTER 6
actual velocity of propagation, it is normal for manufacturers to specify the

velocity factor, which is given simply by
vp 1
vƒ = = (6.6)
c ∈r
where
v ƒ = velocity factor, as a decimal fraction
v p = propagation velocity on the line
c = speed of light in free space
∈ r = relative permittivity of the dielectric
The velocity factor for a transmission line depends entirely on the di-
electric used. It is commonly expressed as a percentage found by multiplying
the value from Equation (6.6) by 100.
EXAMPLE 6.4 Y
Find the velocity factor and propagation velocity for a coaxial cable with a
Teflon dielectric (∈r = 2.1).
SOLUTION
1 1
vƒ = = = 0.69
∈r 21
.
v p = v ƒ c = 0.69 × 300 × 106 m/s = 207 × 106 m/s

X
Traveling Suppose that a sinusoidal wave is applied to one end of a transmission line.
Waves on a The line can be infinite in length, or it can be terminated with a resistance
Transmission Line equal to its characteristic impedance. This will ensure that any signal reach-
ing the end of the line will disappear into the resistance, so that, from the
source the line will look as though it were of infinite length. Let the source
resistance also be equal to Z0; the reason for this will soon become apparent.
The voltage applied to the line at a given instant moves down the line and
appears farther away as time goes on. See Figure 6.8 for the setup.
If the input is sinusoidal, any point on the line will also show a sinusoi-
dal voltage, delayed more in time as we move farther down the line. The
FIGURE 6.8
AC applied to a
matched line
signal appears as a wave that travels down the line. In fact, such a signal is
called a traveling wave. The signal at any point along the line is the same as
that at the source except for a time delay.
With a sine wave, of course, a time delay is equivalent to a phase shift. A
time delay of one period will cause a phase shift of 360 degrees, or one com-
plete cycle; a wave that has been delayed that much will be indistinguishable
from one that has not been delayed at all. As mentioned earlier, the length of
line that causes a delay of one period is known as a wavelength, for which
the usual symbol is λ. If we could look at the voltage along the line at one
instant of time, the resulting “snapshot” would look like the input sine
wave, except that the horizontal axis would be distance rather than time,
and one complete cycle of the wave would occupy one wavelength instead
of one period.
Figure 6.9 shows how the voltage varies along the line for four instants
of time. Note how the crest of the wave moves down the line from one frame
to the next.
Lengths of line that are not equal to a wavelength provide a phase delay
proportional to their length. Since a length λ produces a phase shift of 360
degrees, the phase delay produced by a given line is simply
360L
φ = (6.7)
λ
where
φ = the phase shift in degrees
L = the length of the line
λ = the wavelength on the line
In many applications the time delay and phase shift due to a length of
transmission line are of no concern. However, there are times (for example,
when two signals must arrive at a given point in phase with each other)
when it is important to consider phase shift. Transmission lines can also be
used to deliberately introduce phase shifts and time delays where they are
required.
178 ! CHAPTER 6
FIGURE 6.9 Traveling waves
EXAMPLE 6.5 Y
What length of standard RG-8/U coaxial cable would be required to obtain a
45° phase shift at 200 MHz?
SOLUTION
The velocity factor for this line is 0.66 so
vp = vƒc
= 0.66 × 300 × 106 m/s
= 198 × 106 m/s
The wavelength on the line of a 200 MHz signal is found from Equa-
tion (6.1):
vp = ƒλ
vp
λ =
ƒ
198 × 106 m/s
=
200 × 106 Hz
= 0.99 m
The required length for a phase shift of 45° would be
. m × 45
099
L =
360
= 124 mm
X
' 6.4 Reflections and Standing Waves

It has been shown that a sine wave applied to a matched line results in an
identical sine wave, except for phase, appearing at every point on the line as
the incident wave travels down it. If the line is unmatched, a reflected wave
from the load adds to the incident wave from the source.
First, let us consider a transmission line terminated in an open circuit.
Assume that the line is reasonably long, say one wavelength. The situation is
sketched in Figure 6.10. This figure shows the situation at several points in
the cycle. Since the open circuit cannot dissipate any power, all the energy in
the incident wave must be reflected. The net current at the open circuit must
be zero, and the reflected current must have the same amplitude as, and op-
posite polarity from, the incident voltage at the load. Therefore, the re-
flected voltage has the same amplitude and polarity as the incident voltage.
The reflected wave propagates down the line until it is dissipated in the
source impedance, which has been chosen to match the characteristic im-
pedance of the line. (If there is a mismatch at both ends, the situation will be
complicated by multiple reflections.)
At every point on the line, the instantaneous values of incident and re-
flected voltage add algebraically to give the total voltage. From the figure, it
can be inferred that the voltage at every point on the line varies sinusoidally,
but due to constructive and destructive interference between the incident
and reflected waves, the peak amplitude of the voltage varies greatly. At the
180 ! CHAPTER 6
FIGURE 6.10
Waves on an
open-circuited line
open-circuited end of the line the peak voltage is at maximum. One-quarter

wavelength away from that end, the incident and reflected voltages cancel
exactly, because the two signals have equal amplitude and opposite phase.
At a distance of one-half wavelength from the open-circuited end, there
is another voltage maximum, and the process continues for every half-
wavelength segment of line.
A sketch showing the variation of peak (or RMS) voltage along the line
is shown in Figure 6.11. It is important to realize that this figure does not
represent either instantaneous or dc voltages. There is no dc on this line at
all. The figure shows how the amplitude of a sinusoidal voltage varies along
the line.
FIGURE 6.11
Standing waves on an
open-circuited line
The interaction between the incident and reflected waves, which are
both traveling waves, causes what appears to be a stationary pattern of waves
on the line. It is customary to call these standing waves because of this ap-
pearance. Of course, waves do not really stand still on a line; it is only the
interference pattern that stands still.
For comparison, Figure 6.12 shows the standing waves of voltage on a
line with a shorted end. Naturally, there is no voltage at the shorted end,
since there can never be a voltage at a short circuit. The reflected wave must
therefore have a voltage equal in magnitude and opposite in sign to the inci-
dent voltage. A voltage maximum occurs one-quarter wavelength from the
end, another null occurs at one-half wavelength, and so on.
The current responds in just the opposite way from the voltage. For
the open line, the current must be zero at the open end. It is a maximum
FIGURE 6.12
Standing waves on
a short-circuited line
182 ! CHAPTER 6
one-quarter wavelength away, zero again at a distance of one-half wave-

length, and so on. In other words, the graph of current as a function of posi-
tion for an open-circuited line looks just like the graph of voltage versus
position for a short-circuited line. Similarly, the current on a short-circuited
line has a maximum at the termination, a minimum one-quarter wave-
length away, and another maximum at a distance of a half wavelength, just
like the voltage curve for the open-circuited line.
What happens when the line is mismatched, but not so drastically as dis-
cussed above? There will be a reflected wave, but it will not have as large an
amplitude as the incident wave since some of the incident signal will be dis-
sipated in the load. The amplitude and phase angle will depend on the load
impedance compared to that at the line. The incident and reflected voltages
are related by the coefficient of reflection:
Vr
Γ = (6.8)
Vi
where
Γ = coefficient of reflection
V r = reflected voltage
V i = incident voltage
It can be shown that

ZL − Z0
Γ = (6.9)
ZL + Z0
where
Z L = load impedance
In general, Γ is complex, but for a lossless line it is a real number if the

load is resistive. It is positive for ZL > Z0 and negative for ZL < Z0. For ZL = Z0,
the reflection coefficient is of course zero. A positive real coefficient means
that the incident and reflected voltages are in phase at the load.
When a reflected signal is present but of lower amplitude than the inci-
dent wave, there will be standing waves of voltage and current, but there will
be no point on the line where the voltage or current remains zero over the
whole cycle. See Figure 6.13 for an example.
It is possible to define the voltage standing-wave ratio (VSWR or just
SWR) as follows:
Vmax
SWR = (6.10)
Vmin
FIGURE 6.13
Standing waves on a
mismatched line
where
SWR = voltage standing-wave ratio
V max = maximum rms voltage on the line
V min = minimum rms voltage on the line
The SWR concerns magnitudes only and is thus a real number. It must be
positive and greater than or equal to 1. For a matched line the SWR is 1 (some-
times expressed as 1:1 to emphasize that it is a ratio), and the closer the line is to
being matched, the lower the SWR. The SWR has the advantage of being easier
to measure than the reflection coefficient, but the latter is more useful in many
calculations. Since both are essentially measures of the amount of reflection on
a line, it is possible to find a relationship between them.
The maximum voltage on the line occurs where the incident and re-
flected signals are in phase, and the minimum voltage is found where they
are out of phase. Therefore, using absolute value signs to emphasize the lack
of a need for phase information, we can write
Vmax = Vi + Vr (6.11)
and
Vmin = Vi − Vr (6.12)
Combining Equations (6.10), (6.11), and (6.12), we get
Vmax
SWR = (6.13)
Vmin
Vi + Vr
=
Vi − Vr
 V 
Vi  1 + r 
 Vi 
=
 V 
Vi  1 − r 
 Vi 
184 ! CHAPTER 6
Vr
1+
Vi
=
Vr
1−
Vi
1+ Γ
=
1− Γ
A little algebra will show that |Γ| can also be expressed in terms of SWR:
SWR − 1
Γ = (6.14)
SWR + 1
For the special, but important, case of a lossless line terminated in a resis-
tive impedance, it is possible to find a simple relationship between stand-
ing-wave ratio and the load and line impedances. First, suppose that ZL > Z0.
Then from Equation (6.13),
1+ Γ (6.15)
SWR =
1− Γ
ZL − Z0
1+
ZL + Z0
=
Z − Z0
1− L
ZL + Z0
ZL + Z0 + ZL − Z0
=
ZL + Z0 − ZL + Z0
ZL
=
Z0
It is easy to show, in a similar way, that if Z0 > ZL, then

Z0
SWR = (6.16)
ZL
Use of the appropriate equation will always give an SWR that is greater than
or equal to one, and positive.
EXAMPLE 6.6 Y
A 50-Ω line is terminated in a 25-Ω resistance. Find the SWR.
SOLUTION
In this case, Z0 > ZL so the solution is given by Equation (6.16).
Z0
SWR =
ZL
50
=
25
= 2
X
The presence of standing waves causes the voltage at some points on the
line to be higher than it would be with a matched line, while at other points
the voltage is low but the current is higher than with a matched line. This sit-
uation results in increased losses. In a transmitting application, standing
waves put additional stress on the line and can result in failure of the line or
of equipment connected to it. For instance, if the transmitter happens to be
connected at or near a voltage maximum, the output circuit of the transmit-
ter may be subjected to a dangerous overvoltage condition. This is especially
likely to damage solid-state transmitters, which for this reason are often
equipped with circuits to reduce the output power in the presence of an SWR
greater than about 2:1.
Reflections can cause the power delivered to the load to be less than
it would be with a matched line for the same source, because some of
the power is reflected back to the source. Since power is proportional to the
square of voltage, the fraction of the power that is reflected is Γ 2, that is,
Pr = Γ2Pi (6.17)
where
P r = power reflected from the load
P i = incident power at the load
Γ = voltage reflection coefficient
Sometimes Γ2 is referred to as the power reflection coefficient.

The amount of power absorbed by the load is the difference between the
incident power and the reflected power, that is,
PL = Pi − Γ2Pi (6.18)
= P i (1 − Γ )2
186 ! CHAPTER 6
EXAMPLE 6.7 Y
A generator sends 50 mW down a 50-Ω line. The generator is matched to the
line but the load is not. If the coefficient of reflection is 0.5, how much
power is reflected and how much is dissipated in the load?
SOLUTION
The amount of power that is reflected is, from Equation (6.17):
Pr = Γ2Pi
= 0.52 × 50 mW
= 12.5 mW
The remainder of the power reaches the load. This amount is
PL = Pi − Pr
= 50 mW − 12.5 mW
= 37.5 mW
Alternatively, the load power can be calculated directly from Equa-

tion (6.18):
PL = Pi (1 − Γ2)
= 50 mW(1 − 0.52)
= 37.5 mW
X
Since SWR is easier to measure than the reflection coefficient, an expres-

sion for the power absorbed by the load in terms of the SWR would be useful.
It is easy to derive such an expression by using the relationship between Γ
and SWR given in Equation (6.13). The derivation is left as an exercise; the
result is
4 SWR
PL = Pi (6.19)
(1 + SWR )2
EXAMPLE 6.8 Y
A transmitter supplies 50 W to a load through a line with an SWR of 2. Find
the power absorbed by the load.
SOLUTION
4 SWR
PL = Pi
(1 + SWR )2
4×2
= × 50 W
(1 + 2 )2
= 44.4 W
X
Reflections on transmission lines can cause problems in receiving appli-

cations as well. For instance, reflections on a television antenna feedline can
cause a double image or “ghost” to appear. In data transmission, reflections
can distort pulses, causing errors.
Variation of A matched line presents its characteristic impedance to a source located

Impedance Along any distance from the load. If the line is not matched, however, the imped-
a Line ance seen by the source can vary greatly with its distance from the load.
At those points where the voltage is high and the current low, the impedance
is higher than at points with the opposite current and voltage characteris-
tics. In addition, the phase angle of the impedance can vary. At some points,
a mismatched line may look inductive, at others capacitive, and at a few
points, resistive. Very near the load, the impedance looking into the line is
close to ZL. This is one reason why transmission-line techniques need to be
used only with relatively high frequencies and/or long lines: a line shorter
than about one-sixteenth of a wavelength can usually be ignored.
The impedance that a lossless transmission line presents to a source var-
ies in a periodic way. We have already noticed that the standing-wave pat-
tern repeats itself every one-half wavelength along the line; the impedance
varies in the same fashion. At the load and at distances from the load that are
multiples of one-half wavelength, the impedance looking into the line is
that of the load.
The impedance at any point on a lossless transmission line is given by
the equation
Z L cos θ + jZ 0 sin θ
Z = Z0 (6.20)
Z 0 cos θ + jZ L sin θ
188 ! CHAPTER 6
where
Z = impedance looking from the source toward the load
Z L = load impedance
θ = distance to the load in degrees (for example a quarter
wavelength would be 90°)
Provided that cos θ ≠ 0, this simplifies to
Z L + jZ 0 tan θ
Z = Z0 (6.21)
Z 0 + jZ L tan θ
These equations can be fairly tedious to work with, especially when ZL is

complex. Most transmission-line impedance calculations are done using
computers. Many of the computer programs give their results in the form of
a Smith chart, which will be described in Appendix A.
Characteristics of Though a section of transmission line that is terminated in an open or short

Shorted and circuit is useless for transmitting power, it can serve other purposes. Such a
Open Lines line can be used as an inductive or capacitive reactance or even as a resonant
circuit. In practice, short-circuited sections are much more useful, because
open-circuited lines tend to radiate energy from the open end.
The impedance of a short-circuited line can be found from Equation
(6.21), by setting ZL equal to zero. The impedance looking toward the short
circuit is
Z = jZ 0 tan θ (6.22)
Note that the impedance has no resistive component, since no power

can be dissipated in this line. For short lengths (less than one-quarter wave-
length or 90°), the impedance is inductive. At one-quarter wavelength, the
line looks like a parallel-resonant circuit, with infinite impedance. The line
is capacitive for lengths between one-quarter and one-half wavelength,
since the tangent of the corresponding angle is negative. At a length of
one-half wavelength, or 180°, tan θ is zero, and the line behaves like a
series-resonant circuit having zero impedance. For longer lines the cycle
repeats.
Figure 6.14 shows graphically how the impedance varies with length. It
is easy to remember that a short length of shorted line is inductive by visual-
izing the shorted end as a loop or coil hence an inductance.
The open-circuited line, on the other hand, is capacitive in short
lengths. You can remember this by thinking of the two parallel lines at the
FIGURE 6.14
Impedance of a
shorted line
end of the transmission line as a capacitor. It is series-resonant at a length of

one-quarter wavelength, inductive between one-quarter and one-half wave-
length, and parallel-resonant at a length of one-half wavelength. For longer
lines the cycle repeats. Figure 6.15 shows graphically how the impedance
varies with length for this line.
FIGURE 6.15
Impedance of an
open-circuited line
Short transmission-line sections, called stubs, can be substituted for ca-

pacitors, inductors or tuned circuits in applications where lumped-constant
components (conventional inductors or capacitors) would be inconvenient
or impractical. At VHF and UHF frequencies the required values of induc-
tance and capacitance are often very small. It would be difficult to build a
1 pF capacitor, for instance; the capacitance between its leads might well be
more than that. In addition, physically small components are difficult to use
where large amounts of power are involved, as in transmitters, because of
190 ! CHAPTER 6
the large voltages and currents that must be handled. A short section of
transmission line can avoid all these problems. Also, an air-dielectric trans-
mission line can have much higher Q than a typical lumped-constant reso-
nant circuit.
EXAMPLE 6.9 Y
A series-tuned circuit operating at a frequency of 1 GHz is to be constructed
from a shorted section of air-dielectric coaxial cable. What length should
be used?
SOLUTION
The velocity factor of an air-dielectric line is about 0.95 so the propagation
velocity is, from Equation (6.6),
vp = vƒc
= 0.95 × 300 × 106 m/s
= 285 × 106 m/s
The wavelength on the line is given by Equation (6.1):
vp = ƒλ
vp
λ =
ƒ
285 × 106 m/s

=
1000 × 106 Hz
= 285 mm
Since this is a shorted stub, a half-wavelength section will be series resonant.

Therefore the length will be
λ
L =
2
285 mm
=
2
= 143 mm
X
Shorted stubs can be used anywhere a capacitance, inductance, or tuned

circuit is required. Of course, this technique is more practical at VHF
frequencies and higher where the length of the stub is reasonable and it is
often hard to find discrete components with sufficiently small values.
' 6.5 Transmission Line Losses

Most of the losses in a properly constructed and used transmission line are
due to conductor resistance and, at higher frequencies, dielectric conduc-
tance. Both of these losses increase with frequency. Some transmission lines
also radiate energy. This is particularly true of lines that are improperly con-
nected; for instance, a coaxial line that is used in a balanced circuit will radi-
ate energy from its shield. When parallel lines are used at very high
frequencies, such that the distance between the wires is a substantial portion
of the wavelength, they too can radiate.
Transmission line loss is usually specified in decibels for a given length,
for instance, 100 m or sometimes 100 feet. The decibel loss is proportional to
distance, so it is easy to find the loss for any other length of line.
EXAMPLE 6.10 Y
A transmission line is specified to have a loss of 4 dB/100 m at 800 MHz.
(a) Find the loss in decibels of 130 m of this cable.
(b) Suppose that a transmitter puts 100 W of power into this cable. How
much power reaches the load?
SOLUTION
4 dB
(a) L(dB ) = × 130 m
100 m
= 5.2 dB
(b) For this part we need the loss as a power ratio.

 L(dB) 
L = antilog  
 10 
 5.2 
= antilog  
 10 
= 3.31
192 ! CHAPTER 6
Pi
L =
Po
Pi
Po =
L
100 W
=
331
.
= 30.2 W
X
' 6.6 Waveguides

Transmission line losses increase rapidly with frequency. At frequencies of
several gigahertz or more, the losses in conventional transmission lines are
such as to make long cable runs impractical. Luckily, there is a different form
of transmission line that is impractical at low frequencies but very useful in
the microwave region. This type of line, called a waveguide, generally con-
sists of a hollow, air-filled tube made of conducting material. Rectangular
waveguides of brass or aluminum, sometimes silver-plated on the inside, are
most common, but elliptical and circular cross-sections are also used. Ellipti-
cal waveguides can be semi-flexible and are used extensively for connecting
equipment at the base of a tower to antennas on the tower. Circular guides
are useful when rotating antennas are used, as in radar antennas, because of
their symmetry. Examples are shown in Figure 6.16.
Electromagnetic waves are launched into one end and propagate down
the guide. The waves fill the entire space within the guide, and can be visual-
ized as rays reflecting from the walls of the guide. Since the waves spend
most of their time in the air within the guide, losses due to currents in the
walls are relatively small. The waves are completely contained within the
conducting walls, so there is no loss due to radiation from the guide. Wave-
guides, then, solve the problem of high transmission-line losses for micro-
wave signals.
It is possible to build a waveguide for any frequency, but waveguides op-
erate essentially as high-pass filters, that is, for a given waveguide cross-
section there is a cutoff frequency, below which waves will not propagate. At
frequencies below the gigahertz range, waveguides are too large to be practi-
cal for most applications.
FIGURE 6.16
Waveguides
Modes and There are a number of ways (called modes) in which electrical energy can propa-
Cutoff Frequency gate along a waveguide. All of these modes must satisfy certain boundary con-
ditions. For instance, assuming an ideal conductor for the guide, there cannot
be any electric field along the wall of the waveguide. If there were such a field,
there would have to be a voltage gradient along the wall, and that is impossible
since there cannot be any voltage across a short circuit.
Modes are most easily understood by thinking of a wave moving
through the guide as if it were a ray of light. Figure 6.17 shows the idea. For
each different mode, the ray will strike the walls of the waveguide at a differ-
ent angle. As the angle a ray makes with the wall of the guide becomes larger,
the distance the ray must travel to reach the far end of the guide becomes
greater. Though the propagation in the guide is at the speed of light,
FIGURE 6.17
Multimode
propagation
194 ! CHAPTER 6
the greater distance traveled causes the effective velocity down the guide to
be reduced.
It is desirable to have only one mode propagating in a waveguide. To see
the effect of multimode propagation (more than one mode propagating at a
time), consider a brief pulse of microwave energy applied to one end of
a waveguide. The pulse will arrive at the far end at several different times,
one for each mode. Thus a brief pulse will be spread out over time, an effect
called dispersion. If another pulse follows close behind, there may be inter-
ference between the two. For this reason it is undesirable to have more than
one mode propagating.
Each mode has a cutoff frequency below which it will not propagate. Sin-
gle-mode propagation can be achieved by using only the mode with the low-
est cutoff frequency, which is called the dominant mode. The waveguide is
used at frequencies between the cutoff frequency for the dominant mode
and that of the mode with the next lowest cutoff frequency.
Modes are designated as transverse electric (TE) or transverse magnetic
(TM) modes according to the pattern of electric and magnetic fields within
the waveguide. Figure 6.18 shows several examples of TE modes in a rectan-
gular waveguide. The electric field strength is represented by the arrows,
with the length of the arrows proportional to the field strength. Note that in
all cases the field strength is zero along the walls of the guide. The field
strength varies sinusoidally across the guide cross section. The first number
FIGURE 6.18
TE modes in
rectangular
waveguide
following the TE designation represents the number of half-cycles of the

wave along the long dimension (a) of the rectangular guide, and the second
represents the number of variations along the short dimension (b). Of
course, a and b refer to inside dimensions.
In a rectangular waveguide, TE10 is the dominant mode. In a typical rect-
angular waveguide with a = 2b, the TE10 and TE20 modes each have a cutoff
frequency twice that of the TE10 mode, giving an approximate 2:1 frequency
range for the waveguide in its dominant mode. In what follows, a rectangu-
lar waveguide operating in the TE10 mode will be assumed unless otherwise
stated.
The cutoff frequency for the TE 10 mode can easily be found. For this
mode to propagate, there must be at least one-half wavelength along the
wall. Therefore, at the cutoff frequency,
λc
a = (6.23)
2
where
a = the longer dimension of the waveguide cross section
λ c = cutoff wavelength
Assuming that the waveguide has an air dielectric, the cutoff frequency can
be found as follows. From Equation (6.23),
λ c = 2a
From earlier work we know that for propagation in free space,
c
λ =
ƒ
where
c = 300 × 10 6 meters per second
Therefore, at the cutoff frequency ƒc,
c
λc = (6.24)
ƒc
c
2a =
ƒc
c
ƒc =
2a
196 ! CHAPTER 6
EXAMPLE 6.11 Y
Find the cutoff frequency for the TE 10 mode in an air-dielectric waveguide
with an inside cross section of 2 cm by 4 cm. Over what frequency range is
the dominant mode the only one that will propagate?
SOLUTION
The larger dimension, 4 cm, is the one to use in calculating the cutoff fre-
quency. From Equation (6.24), the cutoff frequency is
c
ƒc =
2a
300 × 106 m/s

=
2 × 4 × 10−2 m
= 3.75 GHz
The dominant mode is the only mode of propagation over a two-to-one fre-
quency range, so the waveguide will be usable to a maximum frequency of
3.75 × 2 = 7.5 GHz.
X
Group and Phase Assuming an air dielectric, the wave travels inside the waveguide at the
Velocity speed of light. However, it does not travel straight down the guide but re-
flects back and forth from the walls. The actual speed at which a signal trav-
els along the guide is called the group velocity, and it is considerably less
than the speed of light. The group velocity in a rectangular waveguide is
given by the equation
2
ƒ 
vg = c 1 −  c  (6.25)
 ƒ
where
ƒ c = cutoff frequency
ƒ = operating frequency
From the above equation, it can be seen that the group velocity is a func-
tion of frequency, becoming zero at the cutoff frequency. At frequencies
below cutoff, of course, there is no propagation, so the equation does not
apply. The physical explanation of the variation of group velocity is that
the angle the wave makes with the wall of the guide varies with frequency. At
frequencies near the cutoff value, the wave moves back and forth across the
guide more often while traveling a given distance down the guide than
it does at higher frequencies. Figure 6.19 gives a qualitative idea of the effect.
FIGURE 6.19 Variation of group velocity with frequency
EXAMPLE 6.12 Y
Find the group velocity for the waveguide of Example 6.10, at a frequency of
5 GHz.
SOLUTION
From Example 6.10, the cutoff frequency is 3.75 GHz. Therefore, from Equa-
tion (6.25), the group velocity is
2
ƒ 
vg = c 1 −  c 
 ƒ
2
 3.75 
= 300 × 106 m/s 1 −  
 5 
= 198 × 106 m/s

X
Since the group velocity varies with frequency, dispersion exists even for
single-mode propagation. If a signal has components of different
198 ! CHAPTER 6
frequencies, the higher-frequency components will travel faster than the

lower-frequency components. This can be a real problem for pulsed and
other wideband signals. For instance, the upper sideband of an FM or AM
signal travels faster than the lower sideband.
EXAMPLE 6.13 Y
A waveguide has a cutoff frequency for the dominant mode of 10 GHz. Two
signals with frequencies of 12 and 17 GHz respectively propagate down a 50
m length of the guide. Calculate the group velocity for each and the differ-
ence in arrival time for the two.
SOLUTION
The group velocities can be calculated from Equation (6.25):
2
ƒ 
vg = c 1 −  c 
 ƒ
For the 12 GHz signal,
2
 10 
vg = 300 × 106 m/s 1 −  
 12 
= 165.8 × 106 m/s
Similarly, the 17-GHz signal has vg = 242.6 × 106 m/s. The 12-GHz signal will
travel the 50 m in:
50 m
t1 = = 3016
. ns
. × 106 m/s
1658
Similarly, it can be shown that the travel time for the 17-GHz signal is:
t2 = 206.1 ns
The difference in travel times for the two signals is:
∆t = 301.6 ns − 206.1 ns = 95.5 ns

X
It is often necessary to calculate the wavelength of a signal in a wave-

guide. For instance, it may be required for impedance matching. It might
seem that the wavelength along the guide could be found using the group
velocity, in much the same way that the velocity factor of a transmission line
is used. However, this common-sense approach does not work because what
is really important in impedance-matching calculations is the change in
phase angle along the line.
Figure 6.20 shows how the angle varies along the guide. The guide wave-
length shown represents 360 degrees of phase variation. The guide wave-
length is always larger than the free-space wavelength. Interestingly, the
more slowly the waves in the guide propagate along it, the more quickly the
phase angle varies along the guide.
FIGURE 6.20
Variation of
phase angle
along a
waveguide
The term phase velocity is used to describe the rate at which the wave
appears to move along the wall of the guide, based on the way the phase
angle varies along the walls. Surprisingly, the phase velocity in a waveguide
is always greater than the speed of light. Of course, the laws of physics pre-
vent anything from actually moving that fast (except in science fiction
stories). A phase velocity greater than the speed of light is possible because
phase velocity is not really the velocity of anything.
A similar effect can be seen with water waves at a beach. If the waves ap-
proach the shore at an angle, as in Figure 6.21, the crest of the wave will appear
to run along the shore at a faster rate than that at which the waves approach the
beach. Once again, there is nothing physical moving at the higher velocity.
FIGURE 6.21
Water waves at
a shoreline
200 ! CHAPTER 6
The relationship between phase velocity and group velocity is very sim-
ple. The speed of light is the geometric mean of the two, that is,
vgvp = c2 (6.26)
It is easy to modify the equations given earlier for group velocity, so that
the phase velocity can be calculated without first finding the group velocity.
The derivation is left as an exercise; the result is given below.
c
vp = (6.27)
2
ƒ 
1−  c
 ƒ
EXAMPLE 6.14 Y
Find the phase velocity for the waveguide used in Example 6.10, at a fre-
quency of 5 GHz.
SOLUTION
The answer can be found directly from Equation (6.27):
c
vp =
2
ƒ 
1−  c
 ƒ
300 × 106 m/s

=
2
. 
 375
1−  
 5 
= 454 × 106 m/s
Alternatively, it can be found from Equation (6.26), since the group velocity
is already known:
vgvp = c2
c2
vp =
vg
(300 × 106 m/s)2

=
198 × 106 m/s
= 454 × 106 m/s

X
Waveguide Like any transmission line, a waveguide has a characteristic impedance.

Impedance Unlike wire lines, however, the waveguide impedance is a function of
frequency. You might expect that the impedance of a waveguide with an air
dielectric would have some relationship to the impedance of free space,
which is 377 Ω, and that is true. The characteristic impedance of a wave-
guide, in ohms, is given by
377
Zo = (6.28)
2
ƒ 
1−  c
 ƒ
EXAMPLE 6.15 Y
Find the characteristic impedance of the waveguide used in the previous ex-
amples, at a frequency of 5 GHz.
SOLUTION
377
Z0 =
2
ƒ 
1−  c
 ƒ
377
=
 3.75 
2
1−  
 5 
= 570 Ω
X
Impedance matching with waveguides can be accomplished as for other

types of transmission lines. The only difference in principle is that the phase
velocity vp must be used in calculating λg, the wavelength in the guide. That
is, the wavelength is given by
vp
λg = (6.29)
ƒ
202 ! CHAPTER 6
Another way to find the guide wavelength, given the free-space wave-
length, is to use the equation
λ
λg = (6.30)
2
ƒ 
1−  c
 ƒ
where
λ = the free-space wavelength
λ g = the wavelength in the guide
EXAMPLE 6.16 Y
Find the guide wavelength for the waveguide used in the previous examples.
SOLUTION
Since the phase velocity is already known, the easiest way to solve this prob-
lem is to use Equation (6.29).
vp
λg =
ƒ
454 × 106 m/s

=
5 × 10 9 Hz
= 9.08 cm
X
Techniques for matching impedances using waveguides also differ from

those used with conventional transmission lines. Shorted stubs of adjustable
length can be used, but a simpler method is to add capacitance or inductance
by inserting a tuning screw into the guide, as shown in Figure 6.22. As the
FIGURE 6.22
Tuning screw
screw is inserted farther into the guide, the effect is first capacitive, then
series-resonant, and finally inductive.
Coupling Power There are three basic ways to launch a wave down a guide. Figure 6.23 shows
into and out of all three. Figure 6.23(a) shows the use of a probe. The probe couples to the
Waveguides electric field in the guide and should be located at an electric-field maxi-
mum. For the TE 10 mode, it should be in the center of the a (wide) dimen-
sion. The probe launches a wave along the guide in both directions.
Assuming that propagation in only one direction is desired, it is only neces-
sary to place the probe a quarter-wavelength from the closed end of the
guide. This closed end represents a short circuit, and following the same
logic as for transmission lines, there will be an electric-field maximum at a
distance of one-quarter wavelength from the guide. This is the waveguide
equivalent of the voltage maximum one-quarter wavelength from the
shorted end of a transmission line.
The wave emitted by the probe reflects from the shorted end of the
guide. Just as with conventional lines, there is a 180° phase shift at the re-
flecting surface and this, combined with the 180° shift due to the total path
length of one-half wavelength, results in the reflected wave being in phase
with (and adding to) the direct wave in the direction along the length of the
guide. The distance to the end of the guide will have to be calculated using
the guide wavelength described earlier.
Figure 6.23(b) shows another way to couple power to a guide. A loop is
used to couple to the magnetic field in the guide. It is placed in a location of
maximum magnetic field, which for the TE10 mode occurs close to the end
wall of the guide. It may help you understand this if you think of the mag-
netic field in a waveguide as equivalent to current in a conventional trans-
mission line and consider the electric field as the equivalent of voltage.
FIGURE 6.23 Coupling power to a waveguide

204 ! CHAPTER 6
There is, of course, a current maximum at the short-circuited end of a trans-

mission line.
A third way of coupling energy, shown in Figure 6.23(c), is simply to put
a hole in the waveguide, so that electromagnetic energy can propagate into
or out of the guide from the region exterior to it.
Waveguides are reciprocal devices, just like transmission lines. There-
fore, the same means can be used to couple power into and out of a wave-
guide.
Variations on the above methods are also possible. For instance, it is
possible to use two holes spaced one-quarter wavelength apart to launch a
wave in only one direction. Figure 6.24 shows how this works. A signal mov-
ing down the main guide in the direction of the arrow will be coupled to
the secondary guide where it will again move in the direction indicated
by the arrow. It will not propagate in the other direction because the signals
from the two holes are 180 degrees out of phase with each other in that
direction. The signal in the main guide reaches hole B 90 degrees after it gets
to hole A. The wave propagating to the left from hole B is delayed another 90
degrees before joining the wave propagated from hole A. A wave propagating
down the main guide in the opposite direction will also be coupled to the
secondary guide, but it will propagate in the opposite direction and will be
absorbed by the resistive material in the end of the secondary guide. Though
only two holes are shown, in practice there is often a greater number to
provide more coupling between the guides. Devices of this type are called di-
rectional couplers.
FIGURE 6.24
Two-hole
directional
coupler
Directional couplers are characterized by their insertion loss, coupling,

and directivity. All three are normally specified in decibels. The insertion loss
is the amount by which a signal in the main guide will be attenuated. The
coupling specification gives the amount by which the signal in the main
guide is greater than that coupled to the secondary waveguide. The directivity
refers to the ratio between the power coupled to the secondary guide, for sig-
nals travelling in the two possible directions along the main guide.
EXAMPLE 6.17 Y
A signal with a level of 20 dBm enters the main waveguide of a directional
coupler in the direction of the arrow. The coupler has an insertion loss of 1
dB, coupling of 20 dB, and directivity of 40 dB. Find the strength of the
signal emerging from each guide. Also find the strength of the signal that
would emerge from the secondary guide, if the signal in the main guide were
propagating in the other direction.
SOLUTION
The signal level in the main guide is
20 dBm − 1 dB = 19 dBm
The signal level in the secondary guide is
20 dBm – 20 dB = 0 dBm
If the signal direction in the main guide were reversed, the signal level in the
secondary guide would be reduced by 40 dB to
0 dBm − 40 dB = −40 dBm

X
Waveguide The use of waveguides requires redesign of some of the ordinary compo-
Components nents that are used with feedlines. The lowly tee connector is an example. In
addition, other components, such as resonant cavities, are too large to be
practical at lower frequencies. Several other varieties of microwave passive
components will also be described.
Bends and Tees Anything that changes the shape or size of a waveguide has an effect on the
electric and magnetic fields inside. If the disturbance is sufficiently large,
there will be a change in the characteristic impedance of the guide. However,
as long as any bend or twist is gradual, the effect will be minimal. Figure 6.25
shows examples of bends and twists. The bends are designated as E-plane
bends, Figure 6.25(a), or H-plane bends shown in Figure 6.25(b). Since the
rectangular guide shown normally operates in the TE 10 mode, the electric
field lines are perpendicular to the long direction. Therefore the E-plane
bend changes the direction of the electric field lines. Similar logic holds for
the H-plane bend.
206 ! CHAPTER 6
FIGURE 6.25
Waveguide bends
Rigid waveguide, with its carefully designed gradual bends, resembles

plumbing and is just as tricky to install. Flexible waveguide, as shown in Fig-
ure 6.26, is used for awkward installations.
FIGURE 6.26
Flexible waveguide
One of the more common components used with ordinary transmission

line is the tee, which allows one line to branch into two. Tees can also be
built for waveguides. Figure 6.27 shows E-plane and H-plane tees, named in
the same way as the bends described above. A signal applied to port 1
FIGURE 6.27
Waveguide tees
appears at each of the other ports. For the H-plane tee, the signal is in phase
at the two outputs, while the E-plane tee produces two out-of-phase signals.
Sometimes the H-plane tee is referred to as a shunt tee, and the E-plane tee is
called a series tee.
The hybrid or “magic” tee shown in Figure 6.28 is a combination of
E-plane and H-plane tees. It has some interesting features. In particular, it
can provide isolation between signals. An input at port 3 will result in equal
and in-phase outputs at ports 1 and 2 but no output at port 4. On the other
hand, a signal entering via port 4 will produce equal and out-of-phase out-
puts at ports 1 and 2 but no output at port 3.
FIGURE 6.28
Hybrid tee
Cavity In our discussion of waveguides it was noted that the waves reflect from the
Resonators walls as they proceed down the guide. Suppose that instead of using a con-
tinuous waveguide, we were to launch a series of waves in a short section of
guide called a cavity, as illustrated in Figure 6.29. The waves would, of
course, reflect back and forth from one end to the other. A cavity of random
size and shape would have random reflections with a variety of phase angles
and a good deal of cancellation. However, suppose that the cavity had a
FIGURE 6.29
Rectangular
cavity
208 ! CHAPTER 6
length of exactly one-half wavelength. Waves would reflect from one end to
the other and would be in phase with the incident signal. There would be a
buildup of field strength within the cavity. In a perfectly lossless cavity, this
could continue forever. Of course, losses in the walls of a real cavity would
result in the signal eventually dying out unless sustained by new energy in-
put.
This description should sound familiar: it is a description of resonance.
Like any other resonant device, a waveguide cavity has a Q; the Q for reso-
nant cavities is very high, on the order of several thousand. Cavities can be
tuned by changing their size (for instance, by moving a short-circuiting
plate at one end).
The rectangular cavity described above is not the only possible type. Fig-
ure 6.30 shows several other types. Resonant cavities of various sorts are
found in many types of microwave devices and are even used in the VHF re-
gion when very high Q is required and the considerable size of the cavities is
acceptable.
FIGURE 6.30
Resonant cavities
Waveguide At lower frequencies a load is simply a resistor, and an attenuator is a combi-

Attenuators and nation of resistors designed to preserve the characteristic impedance of a sys-
Loads tem while reducing the amount of power applied to a load. Once again, this
lumped-constant approach becomes less practical as the frequency in-
creases. At microwave frequencies, an ordinary carbon or metal-film resistor
would have a complex equivalent circuit involving a good deal of distrib-
uted inductance and capacitance. Nonetheless, the basic idea of using resis-
tive material to absorb energy is still valid.
Figure 6.31 shows waveguide versions of attenuators. The flap attenu-
ator shown in Figure 6.31(a) uses a carbon flap that can be inserted to a
greater or lesser extent into the waveguide. The fields inside the guide are, of
course, present in the carbon vane as well. A current flows in the flap, caus-
ing power loss.
FIGURE 6.31
Waveguide
attenuators
The attenuator shown in Figure 6.31(b) uses a rotating vane. When the
vane is rotated so that the electric field is perpendicular to its surface, little
loss occurs, but when the field runs along the surface of the vane, a much
larger current is induced, causing greater loss. Both fixed and variable vane
attenuators are available with a variety of attenuation values.
Figure 6.32 illustrates a terminating load for a waveguide. The carbon in-
sert is designed to dissipate the energy in the guide without reflecting it.
FIGURE 6.32
Waveguide
termination
Circulators and Isolators and circulators are useful microwave components that generally
Isolators use ferrites in their operation. The theory of operation is beyond the scope of
this text, but a brief description of their characteristics is in order.
An isolator is a device that allows a signal to pass in only one direction.
In the other direction, it is greatly attenuated. An isolator can be used to
shield a source from a mismatched load. Energy will still be reflected from
the load, but instead of reaching the source, the reflected power is dissipated
in the isolator. Figure 6.33 illustrates this application of an isolator.
The circulator, shown schematically in Figure 6.34, is a very useful de-
vice that allows the separation of signals. The three-port circulator shown
allows a signal introduced at any port to appear at, and only at, the next port
in counterclockwise rotation. For instance, a signal entering at port 1 ap-
pears at port 2 but not at port 3. Circulators can have any number of ports
from three up, but three- and four-port versions are the most common.
One simple example of an application for a circulator is as a transmit-
receive switch. The transmitter output is connected to port 1, the antenna to
210 ! CHAPTER 6
FIGURE 6.33
Isolator
FIGURE 6.34
Circulator
port 2, and the receiver input to port 3. The transmitter output signal is
applied to the antenna, and a received signal from the antenna reaches the
receiver. The transmitter signal does not reach the receiver, however; if it
did, it would probably damage the receiver input circuitry.
( Radio waves are a form of electromagnetic radiation similar to light ex-
cept for a lower frequency and longer wavelength.
( Any pair of conductors can act as a transmission line. The methods pre-
sented in this chapter need to be used whenever the line is longer than ap-
proximately one-sixteenth wavelength at the highest frequency in use.
( Any transmission line has a characteristic impedance determined by its
geometry and its dielectric. If a transmission line is terminated by an im-
pedance that is different from its characteristic impedance, part of a sig-
nal travelling down the line will be reflected. Usually this is undesirable.
( Reflections on lines are characterized by the reflection coefficient and the

standing-wave ratio. The latter is easier to measure. A matched line has a
reflection coefficient of 0 and an SWR of 1.
( Lines that are terminated by either a short or an open circuit can be used
as reactances or as either series- or parallel-resonant circuits, depending
on their length.
( Waveguides are a very practical means of transmitting electrical energy at
microwave frequencies, as they have much lower losses than coaxial ca-
ble. They are not very useful at lower frequencies because they must be
too large in cross section.
( Waveguides are generally useful over only a 2:1 frequency range. They
have a lower cutoff frequency that depends on their dimensions, and
they exhibit dispersion due to multimode propagation at high frequen-
cies.
( Two velocities must be calculated for waveguides. The group velocity,
which is lower than the speed of light, is the speed at which signals travel
down the guide. The phase velocity, which is greater than the speed of
light, is used for calculating the wavelength in the guide.
( Power can be coupled into and out of waveguides using probes, loops, or
holes in the guide.
( Ferrites have special properties that make them just as useful at micro-
wave frequencies as at lower frequencies. Applications of ferrites include
circulators, isolators, attenuators and resonators.
( Equation List
vp = ƒλ (6.1)
L
Z0 = (6.3)
C
138 D
Z0 = log (6.4)
∈r d
D
Z 0 = 276 log (6.5)
r
vp 1
vƒ = = (6.6)
c ∈r
212 ! CHAPTER 6
360L
φ = (6.7)
λ
Vr
Γ = (6.8)
Vi
ZL − Z0
Γ = (6.9)
ZL + Z0
Vmax
SWR = (6.10)
Vmin
1+ Γ
SWR = (6.13)
1− Γ
SWR − 1
Γ = (6.14)
SWR + 1
ZL
SWR = (6.15)
Z0
Z0
SWR = (6.16)
ZL
Pr = Γ2Pi (6.17)
PL = Pi (1 − Γ2) (6.18)
4 SWR
PL = Pi (6.19)
(1 + SWR )2
Z L + jZ 0 tan θ
Z = Z0 (6.21)
Z 0 + jZ L tan θ
Z = jZ0 tan θ (6.22)
λc
a = (6.23)
2
c
ƒc = (6.24)
2a
2
ƒ 
vg = c 1 −  c  (6.25)
 ƒ
vgvp = c2 (6.26)
c
vp = (6.27)
2
ƒ 
1−  c
 ƒ
377
Z0 = (6.28)
2
ƒ 
1−  c
 ƒ
vp
λg = (6.29)
ƒ
λ
λg = (6.30)
2
ƒ 
1−  c
 ƒ
( Key Terms
characteristic impedance ratio between voltage and current on a
transmission line, on a waveguide, or in a medium
directional coupler device that allows signals to travel between ports in
one direction only
dispersion pulse spreading caused by variation of propagation velocity
with frequency
group velocity speed at which signals move down a transmission line or
waveguide
microstrip transmission line consisting of a circuit-board trace on one
side of a dielectric substrate (the circuit board) and a ground plane on
the other side
phase velocity the apparent velocity of waves along the wall of a
waveguide
stripline transmission line consisting of a circuit board having a ground
plane on each side of the board, with a conducting trace in the center
214 ! CHAPTER 6
voltage standing-wave ratio (VSWR) ratio between maximum and

minimum peak or rms voltage on a transmission line
waveguide metallic tube down which waves propagate
wavelength distance a wave travels in one period
( Questions
1. Explain why sound waves can have the same wavelength as radio waves
at a much lower frequency.
2. Why is an ordinary extension cord not considered a transmission line,
while a television antenna cable of the same length would be?
3. Explain the difference between balanced and unbalanced lines, and
give an example of each.
4. What is the difference between a microstrip and a stripline?
5. Draw the equivalent circuit for a short section of transmission line, and
explain the physical meaning of each circuit element.
6. What is meant by the characteristic impedance of a transmission line?
7. What factors determine the characteristic impedance of a line?
8. Define the velocity factor for a transmission line, and explain why it
can never be greater than one.
9. Explain what is meant by the SWR on a line, and state its value when a
line is perfectly matched.
10. Why is a high SWR generally undesirable?
11. Why are shorted stubs preferred to open ones for impedance matching?
12. Draw a sketch showing how the impedance varies with distance along a
lossless shorted line.
13. What would be the effect of placing a shorted, half-wave stub across a
matched line?
14. What are the major contributors to transmission-line loss?
15. How does transmission-line loss vary with frequency, and why?
16. Explain what is meant by the dominant mode for a waveguide, and why
the dominant mode is usually the one used.
17. Name the dominant mode for a rectangular waveguide, and explain
what is meant by the numbers in the designation.
18. Explain the difference between phase velocity and group velocity in a
waveguide. State which one of these is greater than the speed of light,
and explain how this is possible.
19. State what is meant by dispersion, and show how it can arise in two dif-
ferent ways in a waveguide.
20. Draw diagrams showing the correct positions in which to install a probe
and a loop to launch the dominant mode in a rectangular waveguide.
21. Explain the function of a directional coupler, and draw a sketch of a di-
rectional coupler for a waveguide.
22. Explain the operation of a hybrid tee.
23. Sketch a four-port circulator, and show what happens to a signal enter-
ing at each port.
( Problems
1. Find the wavelength for radio waves in free space at each of the follow-
ing frequencies:
(a) 160 MHz (VHF marine radio range)
(b) 800 MHz (cell phone range)
(c) 2 GHz (PCS range)
2. How far does a radio wave travel through space in one microsecond?
3. Visible light has a range of wavelengths from approximately 400
nanometers (violet) to 700 nm (red). Express this as a frequency range.
4. Calculate the characteristic impedance of an open wire transmission
line consisting of two wires with diameter 1 mm and separation 1 cm.
5. Calculate the characteristic impedance of a coaxial line with a polyeth-
ylene dielectric, if the diameter of the inner conductor is 3 mm and the
inside diameter of the outer conductor is 10 mm.
6. Repeat problem 1 for waves on a coaxial cable with a solid polyethylene
dielectric.
7. Repeat problem 2 for a radio wave propagating along a coaxial cable
with polyethylene foam dielectric.
8. How long a line is required to produce a 45° phase shift at 400 MHz if
the dielectric is
(a) air?
(b) solid polyethylene?
216 ! CHAPTER 6
9. A generator is connected to a short-circuited line 1.25 wavelengths

long.
(a) Sketch the waveforms for the incident, reflected, and resultant volt-
ages at the instant the generator is at its maximum positive voltage.
(b) Sketch the pattern of voltage standing waves on the line.
(c) Sketch the variation of impedance along the line. Be sure to note
whether the impedance is capacitive, inductive, or resistive at each
point.
10. An open-circuited line is 0.75 wavelengths long.
(a) Sketch the incident, reflected, and resultant voltage waveforms at
the instant the generator is at its peak negative voltage.
(b) Draw a sketch showing how RMS voltage varies with position along
the line.
11. A 75-Ω source is connected to a 50-Ω load (a spectrum analyzer) with a
length of 75-Ω line. The source produces 10 mW. All impedances are re-
sistive.
(a) Calculate the SWR.
(b) Calculate the voltage reflection coefficient.
(c) How much power will be reflected from the load?
12. A transmitter delivers 50 W into a 600-Ω lossless line terminated with
an antenna that has an impedance of 275 Ω, resistive.
(a) What is the coefficient of reflection?
(b) How much of the power actually reaches the antenna?
13. A shorted stub acts as a parallel-resonant circuit at 100 MHz. What is its
function at 200 MHz?
14. A properly matched transmission line has a loss of 1.5 dB per 100 m. If a
signal power of 10 W is supplied to one end of the line, how much
power reaches the load, 27 m away?
15. A receiver requires a signal of at least 0.5 µV for satisfactory reception.
How strong (in µV) must the signal be at the antenna if the receiver is
connected to the antenna by 25 m of matched line having an attenua-
tion of 6 dB per 100 m ?
16. A transmitter with an output power of 50 W is connected to a matched
load by means of 32 m of matched coaxial cable. It is found that only 35
W of power is dissipated in the load. Calculate the loss in the cable in
dB/100 m.
17. RG-52/U waveguide is rectangular with an inner cross-section of 22.86
by 10.16 mm. Calculate the cutoff frequency for the dominant mode
and the range of frequencies for which single-mode propagation is

possible.
18. For a 10-GHz signal in RG-52/U waveguide (described in problem 17),
calculate:
(a) the phase velocity
(b) the group velocity
(c) the guide wavelength
(d) the characteristic impedance
19. A signal with a center frequency of 10 GHz has a bandwidth of 200 MHz.
It propagates down 100 m of RG-52/U waveguide (described in problem
17). Calculate the difference in arrival times between the highest-fre-
quency and lowest-frequency components of the signal.
20. A signal with a power of 10 mW enters port 1 of a directional coupler.
The output power is 7 mW at port 2 and 100 µW at port 3. Calculate the
coupling to each output port and the insertion loss for the device.
21. (a) Draw a sketch showing how a hybrid tee could be used to connect a
transmitter and a receiver to the same antenna in such a way that
the transmitter power would not reach the receiver.
(b) Draw a sketch showing how a circulator could be arranged to do the
same thing.
(c) What advantage does the circulator version of this device have over
the hybrid-tee version?
22. A transmission line of unknown impedance is terminated with two dif-
ferent resistances and the SWR is measured each time. With a 75-Ω ter-
mination the SWR measures 1.5. With a 300-Ω termination it measures
2.67. What is the impedance of the line?
Radio Propagation 7
Objectives
( Describe the propagation of radio waves in free space and over land.
( Calculate power density and electric and magnetic field intensity for waves
propagating in free space.
( Calculate free-space attenuation and path loss.
( Perform the necessary calculations to determine the maximum
communication range for line-of-sight propagation.
( Calculate path loss in a mobile environment, and explain how such an
environment differs from free space.
( Explain rapid fading and calculate the fade period for a moving vehicle.
( Describe the use of repeaters to increase communication range.
( Explain the cellular concept and calculate the signal-to-interference ratio for
cellular systems.
( Describe how cell-splitting can increase the capacity of a system.
( Describe means to reduce the effects of fading in mobile systems.
( Distinguish between geostationary satellites and those in lower orbits, and
explain the advantages and disadvantages of each.
220 ! CHAPTER 7
' 7.1 Introduction

In the previous chapter we studied the propagation of electromagnetic
waves on transmission lines and in waveguides. That is, we looked at the
“wired” part of wireless communication. In this chapter we consider the
“wireless” part. Waves propagate through space as transverse electromagnetic
(TEM) waves. This means that the electric field, the magnetic field, and the
direction of travel of the wave are all mutually perpendicular. The sketch in
Figure 7.1 is an attempt to represent this three-dimensional process in two
dimensions. The polarization of a wave is the direction of the electric field
vector. Polarization may be horizontal or vertical and can also be circular or
elliptical if the electric field vector rotates as it moves through space.
Radio waves are generated by electrons moving in a conductor, or set of
conductors, called an antenna. Antennas are the subject of the next chapter.
Once launched, electromagnetic waves can travel through free space
and through many materials. Any good dielectric will pass radio waves; the
material does not have to be transparent to light. The waves do not travel
well through lossy conductors, such as seawater, because the electric fields
cause currents to flow that dissipate the energy of the wave very quickly.
Radio waves reflect from good conductors, such as copper or aluminum.
The speed of propagation of radio waves in free space is the same as that
of light, approximately 300 × 106 m/s. In other media, the velocity is lower.
FIGURE 7.1
Transverse
electromagnetic
waves
RADIO PROPAGATION ! 221
The propagation velocity is given by

c
vp = (7.1)
∈r
where
v p = propagation velocity in the medium
c = 300 × 10 6 m/s, the propagation velocity in free space
∈ r = the relative permittivity (dielectric constant) of the medium
This is the same equation given in Chapter 6 for the velocity of signals
along a coaxial cable. One way of thinking of such a line is to imagine an
electromagnetic wave propagating through the dielectric, guided by the
conductors. It should be no surprise that the wave equation in Chapter 6
also applies to waves in free space and dielectrics. Radio waves are refracted
as they pass from one medium to another with a different propagation veloc-
ity, just as light is.
' 7.2 Free-Space Propagation

The simplest source of electromagnetic waves would be a point in space.
Waves would radiate equally from this source in all directions. A wavefront,
that is, a surface on which all the waves have the same phase, would be the
surface of a sphere. Such a source is called an isotropic radiator and is
shown in Figure 7.2. Of course, an actual point source is not possible, but the
approximation is good at distances that are large compared with the dimen-
sions of the source. This is nearly always true of radio propagation.
If only a small area on the sphere shown in Figure 7.2 is examined and
if the distance from the center of the sphere is large, the area in question
FIGURE 7.2
Isotropic
radiator
222 ! CHAPTER 7
resembles a plane. In the same way, we experience the earth as flat, though
we know it is roughly spherical. Consequently, waves propagating in free
space are usually assumed to be plane waves, which are often simpler to deal
with than spherical waves.
There is no loss of energy as radio waves propagate in free space, but
there is attenuation due to the spreading of the waves. If a sphere were
drawn at any distance from the source and concentric with it, all the energy
from the source would pass through the surface of the sphere. Since no en-
ergy would be absorbed by free space, this would be true for any distance, no
matter how large. The energy would be spread over a larger surface as the dis-
tance from the source increased.
Since an isotropic radiator radiates equally in all directions, the power
density, in watts per square meter, is simply the total power divided by the
surface area of the sphere. Put mathematically,
Pt
PD = (7.2)
4πr 2
where
P D = power density in W/m 2
P t = transmitted power in W
r = distance from the antenna in meters
Not surprisingly, this is the same “square-law” attenuation that applies
to light and sound, and, in fact, to any form of radiation. It is important to
realize that this attenuation is not due to any loss of energy in the medium,
but only to the spreading out of the energy as it moves farther from the
source. Any actual losses will be in addition to this.
EXAMPLE 7.1 Y
A power of 100 W is supplied to an isotropic radiator. What is the power den-
sity at a point 10 km away?
SOLUTION
Pt
PD =
4 π r2
100 W
=
4 π(10 × 103 m)2
= 79.6 nW/m2
X
Real antennas do not radiate equally in all directions, of course. Equa-

tion (7.2) can easily be modified to reflect this, if we define antenna gain as
follows:
PDA
Gt = (7.3)
PDI
where
G t = transmitting antenna gain
P DA = power density in a given direction from the real antenna
P DI = power density at the same distance from an isotropic
radiator with the same P t
Antennas are passive devices and do not have actual power gain. They
achieve a greater power density in certain directions at the expense of re-
duced radiation in other directions. In this capacity, an antenna resembles
the reflector in a flashlight more than it does an amplifier. The ways in
which antennas can be constructed will be examined in the next chapter.
Now we can modify Equation (7.2) to include antenna gain:
PG
PD = t t
(7.4)
4 π r2
Usually, antenna gain is specified in dBi, where the “i” indicates gain
with respect to an isotropic radiator. The gain must be converted to a power
ratio to be used with Equation (7.4).
We can define the effective isotropic radiated power (EIRP) of a transmitting
system in a given direction as the transmitter power that would be needed,
with an isotropic radiator, to produce the same power density in the given
direction. Therefore, it is apparent that
EIRP = P t G t (7.5)
and Equation (7.4) can be modified for use with EIRP.
EIRP
PD = (7.6)
4 π r2
The use of Equation (7.4) or Equation (7.6) in a given problem depends

on whether the transmitter power and antenna gain are specified separately
or combined into one EIRP rating.
EXAMPLE 7.2 Y
The transmitter of Example 7.1 is used with an antenna having a gain of
5 dBi. Calculate the EIRP and the power density at a distance of 10 km.
224 ! CHAPTER 7
SOLUTION
First convert the gain to a power ratio.
 5
Gt = log −1  
 10 
= 316
.
This means that the EIRP in the given direction is about three times the ac-
tual transmitter power. More precisely,
EIRP = GtPt = 3.16 × 100 W = 316 W
The power density is

EIRP
PD =
4 π r2
316
=
4 π(10 × 103 )2
= 251.5 nW/m2
X
The strength of a signal is more often given in terms of its electric field
intensity than power density, perhaps because the former is easier to mea-
sure. There is a simple relationship between electric field intensity and
power density. Power density is analogous to power in a lumped-constant
system, and electric field intensity is the equivalent of voltage. The familiar
equation from basic electricity:
V2
P =
R
becomes
E2
PD =
Z
where
E = electric field intensity in volts per meter
Z = characteristic impedance of the medium in ohms
The characteristic impedance of free space is 377 Ω so, in free space,
E2
PD = (7.7)
377
It is quite easy to find a direct relationship between EIRP, distance, and

electric field strength. First we rearrange Equation (7.7) so that E is the un-
known:
E = 377PD (7.8)
Now we can substitute the expression for PD found in Equation (7.6) into this
equation to get:
377 EIRP
E = (7.9)
4 π r2
30EIRP
=
r
EXAMPLE 7.3 Y
Find the electric field intensity for the signal of Example 7.2, at the same dis-
tance (10 km) from the source.
SOLUTION
There are two ways to do this. Since PD is already known, we could use Equa-
tion (7.8):
E = 377PD
= . × 10−9
377 × 2515
= 9.74 mV/m
Alternatively we could begin again with Equation (7.9):
30EIRP
E =
r
30 × 316
=
10 × 103
= 9.74 mV/m
X
226 ! CHAPTER 7
Receiving A receiving antenna absorbs some of the energy from radio waves that pass
Antenna Gain and it. Since the power in the wave is proportional to the area through which it
Effective Area passes, a large antenna will intercept more energy than a smaller one (other
things being equal) because it intercepts a larger area. Antennas are also
more efficient at absorbing power from some directions than from others.
For instance, a satellite dish would not be very efficient if it were pointed at
the ground instead of the satellite. In other words, receiving antennas have
gain, just as transmitting antennas do. In fact, the gain is the same whether
the antenna is used for receiving or transmitting.
The power extracted from the wave by a receiving antenna depends both
on its physical size and on its gain. The effective area of an antenna can be
defined as
Pr
Ae ƒƒ = (7.10)
PD
where
A eƒƒ = effective area of the antenna in m 2
P r = power delivered to the receiver in W
P D = power density of the wave in W/m 2
Equation (7.10) simply tells us that the effective area of an antenna is the
area from which all the power in the wave is extracted and delivered to the
receiver. Combining Equation (7.10) with Equation (7.4) gives
Pr = Ae ƒƒ PD (7.11)
Ae ƒƒ PG
= t t
4 π r2
It can be shown that the effective area of a receiving antenna is
λ2G r
Ae ƒƒ = (7.12)
4π
where
G r = antenna gain, as a power ratio
λ = wavelength of the signal
Path Loss Combining Equations (7.11) and (7.12) gives an expression for the receiver
power in terms of antenna gain, which is much more commonly found in
specifications than is effective area.
Ae ƒƒ PG
Pr = t t
(7.13)
4 π r2
λ2G r PG
= t t
(4 π )(4 π r 2 )
λ2 PG
t t Gr
=
16π 2r 2
While accurate, this equation is not very convenient. Gain and attenua-
tion are usually expressed in decibels rather than directly as power ratios;
the distance between transmitter and receiver is more likely to be given in
kilometers than meters; and the frequency of the signal, in megahertz,
is more commonly used than its wavelength. It is quite easy to perform the
necessary conversions to arrive at a more useful equation. The work in-
volved is left as an exercise; the solution follows.
P r = P t + G t + G r − (32.44 + 20 log d + 20 log ƒ) (7.14)

where
P r = received power in dBm
P t = transmitted power in dBm
G t = transmitting antenna gain in dBi
G r = receiving antenna gain in dBi
d = distance between transmitter and receiver, in km
ƒ = frequency in MHz
Note that Pt and Pr are the power levels at the transmitting and receiving
antennas, respectively. Attenuation due to transmission-line losses or mis-
match is not included; these losses (in decibels) can be found separately and
subtracted from the result for P r given above to give the actual received
power.
Sometimes it is convenient to have an expression for the free-space at-
tenuation, often called path loss, that is independent of antenna gain. This
is easily obtained by extracting the loss part of Equation (7.14), thereby con-
verting Equation (7.14) into two equations:
P r = P t + G t + G r − L ƒs (7.15)
L ƒs = 32.44 + 20 log d + 20 log ƒ (7.16)
where
L ƒs = free-space loss in decibels

228 ! CHAPTER 7
Using Equation (7.15) allows us to construct a diagram of a radio link, show-

ing all the power levels, gains, and losses, and then merely add and subtract
to find the received power. It also allows for the use of a different expression
for path loss, necessary if the medium is not free space. The following exam-
ples show the use of path loss in calculating received signal strength.
First let us consider a very straightforward application. Then look at
Example 7.5, which is a little more complex, involving transmission-line
loss and mismatch.
EXAMPLE 7.4 Y
A transmitter has a power output of 150 W at a carrier frequency of 325 MHz.
It is connected to an antenna with a gain of 12 dBi. The receiving antenna is
10 km away and has a gain of 5 dBi. Calculate the power delivered to the
receiver, assuming free-space propagation. Assume also that there are no
losses or mismatches in the system.
SOLUTION
In all problems of this sort, it is a good idea to begin by sketching the system.
This example can be done easily enough without such a sketch, but many
real-world situations are more complex. See Figure 7.3 for the setup.
FIGURE 7.3
The next step is to convert the transmitter power into dBm:

 Pt 
Pt(dBm) = 10 log  
 1 mW 
 150 W 
= 10 log  
 0.001 W 
= 51.8 dBm
Marking the transmitter power and antenna gains on the sketch shows
us that the only missing link is the path loss. We can find this from Equa-
tion (7.16).
Lƒs = 32.44 + 20 log d + 20 log ƒ

= 32.44 + 20 log 10 + 20 log 325
= 102.7 dB
Now we can easily find the received power from Equation (7.15).
Pr = Pt + Gt + Gr − Lƒs
= 51.8 + 12 + 5 − 102.7
= −33.9 dBm
EXAMPLE 7.5 Y
A transmitter has a power output of 10 W at a frequency of 250 MHz. It is
connected by 20 m of a transmission line having a loss of 3 dB/100 m to an
antenna with a gain of 6 dBi. The receiving antenna is 25 km away and has a
gain of 4 dBi. There is negligible loss in the receiver feedline, but the receiver
is mismatched: the antenna and line are designed for a 50 Ω impedance, but
the receiver input is 75 Ω. Calculate the power delivered to the receiver,
assuming free-space propagation.
SOLUTION
We begin with the sketch of Figure 7.4. A glance at the sketch shows what
needs to be done. First, as in the previous example, we convert the transmit-
ter power to dBm.
 Pt 
Pt(dBm) = 10 log  
 1 mW 
 10 W 
= 10 log  
 0.001 W 
= 40 dBm
FIGURE 7.4
230 ! CHAPTER 7
Next, it is necessary to find the loss at each stage of the system. If all the
losses are in decibels, it will only be necessary to add them to get the total loss.
We use Equation (7.16) to find the path loss.
Lƒs = 32.44 + 20 log d + 20 log ƒ

= 32.44 + 20 log 25 + 20 log 250
= 108.3 dB
For the transmitter feedline, the loss is
3 dB
Ltx = 20 m × = 0.6 dB
100 m
The receiver feedline is lossless, but some of the power reflects from the
receiver back into the antenna due to the mismatch. This power will be
reradiated by the antenna, and will never reach the receiver. Therefore, for
our purposes it is a loss. Remember from Chapter 6 that the proportion of
power reflected is the square of the reflection coefficient and the reflection
coefficient is given by
ZL − Z0
Γ =
ZL + Z0
75 − 50
=
75 + 50
= 0.2
Γ2 = 0.22
= 0.04
The proportion of the incident power that reaches the load is
1 − Γ2 = 0.96
In decibels, the loss due to mismatch is
Ltx = − 10 log 0.96

= 0.177 dB
Now we can easily find the received signal strength. We start with the
transmitter power in dBm, add all the gains, and subtract all the losses (in
dB) to get the result. Thus we find that
Pr = Pt − Ltx + Gt − Lƒs + Gr − Lrx

= 40 − 0.6 + 6 − 108.3 + 4 − 0.177
= −59.1 dBm
X
' 7.3 Terrestrial Propagation

The simplified free-space model described above has some applicability
to propagation over the surface of the earth. Air has very little loss at fre-
quencies below about 20 GHz, and its dielectric constant is close to one.
The most obvious differences between free-space and terrestrial propagation
are that range is often limited by the horizon, signals may reflect from the
earth itself, and various obstacles may exist between transmitter and
receiver.
At low-to-medium frequencies (up to about 3 MHz), radio waves can fol-
low the curvature of the earth, a phenomenon known as ground-wave propa-
gation, and in the high-frequency range (about 3 to 30 MHz) the waves may
be returned to earth from an ionized region in the atmosphere called the
ionosphere. Both ground-wave and ionospheric propagation can result in
reception of signals at much greater distances than the horizon, but neither
is generally applicable at the frequencies (from VHF up) generally employed
in wireless communication systems. At these frequencies, propagation is
generally line-of-sight, as explained in the next section.
Line-of-Sight The practical communication distance for line-of-sight propagation is lim-

Propagation ited by the curvature of the earth. In spite of the title of this section, the
maximum distance is actually greater than the eye can see because refraction
in the atmosphere tends to bend radio waves slightly toward the earth. The
dielectric constant of air usually decreases with increasing height, because
of the reduction in pressure, temperature, and humidity with increasing
distance from the earth. The effect varies with weather conditions, but
it usually results in radio communication being possible over a distance
approximately one-third greater than the visual line of sight.
Just as one can see farther from a high place, the height above average
terrain of both the transmitting and receiving antennas is very important in
calculating the maximum distance for radio communication. Figure 7.5
shows the effect of increased antenna height on maximum range. Antenna
heights are greatly exaggerated in the figure, of course.
FIGURE 7.5 Line-of-sight propagation

232 ! CHAPTER 7
An approximate value for the maximum distance between transmit-

ter and receiver, over reasonably level terrain, is given by the following
equation:
d = 17ht + 17hr (7.17)
where
d = maximum distance in kilometers
h t = height of the transmitting antenna in meters
h r = height of the receiving antenna in meters
EXAMPLE 7.6 Y
A taxi company uses a central dispatcher, with an antenna at the top of a
15 m tower, to communicate with taxicabs. The taxi antennas are on the
roofs of the cars, approximately 1.5 m above the ground. Calculate the maxi-
mum communication distance:
(a) between the dispatcher and a taxi
(b) between two taxis
SOLUTION
(a) d = 17ht + 17hr
= 17 × 15 + 17 × 15
.
= 21 km
(b) d = 17ht + 17hr
= 17 × 15
. + 17 × 15
.
= 10.1 km
X
The maximum range calculated from Equation (7.17) will only be

achieved if the received signal strength, as calculated in the previous sec-
tion, is sufficient. Maximum range is achieved by using a combination of
reasonably large transmitter power and high-gain antennas located as high
as possible.
It is not always desirable to achieve maximum distance. There are situa-

tions, as in the cellular systems to be described later, where it is common
to limit the effective range to a distance smaller than predicted by Equa-
tion (7.17).
Multipath Although line-of-sight propagation uses a direct path from transmitter to re-
Propagation ceiver, the receiver can also pick up reflected signals. Probably the simplest
case is reflection from the ground, as shown in Figure 7.6. If the ground is
rough, the reflected signal is scattered and its intensity is low in any given
direction. If, on the other hand, the reflecting surface is relatively smooth—
a body of water, for instance—the reflected signal at the receiver can have a
strength comparable to that of the incident wave, and the two signals will
interfere. Whether the interference is constructive or destructive depends
on the phase relationship between the signals: if they are in phase, the re-
sulting signal strength is increased, but if they are 180° out of phase, there is
partial cancellation. When the surface is highly reflective, the reduction in
signal strength can be 20 dB or more. This effect is called fading. The exact
phase relationship depends on the difference, expressed in wavelengths,
between the lengths of the transmission paths for the direct and reflected
signals. In addition, there is usually a phase shift of 180° at the point of
reflection.
FIGURE 7.6
Ground
reflections
If the transmitter and receiver locations are fixed, the effect of reflec-
tions can often be reduced by carefully surveying the proposed route and ad-
justing the transmitter and receiver antenna heights so that any reflection
takes place in wooded areas or rough terrain, where the reflection will be dif-
fuse and weak. Where most of the path is over a reflective surface such as
desert or water, fading can be reduced by using either frequency diversity or
spatial diversity. In the former method, more than one frequency is avail-
able for use; the difference, in wavelengths, between the direct and incident
path lengths will be different for the two frequencies. In spatial diversity
there are two receiving antennas, usually mounted one above the other on
the same tower. The difference between direct and reflected path length is
different for the two antennas.
234 ! CHAPTER 7
Diffraction from obstacles in the path can also be a problem for

line-of-sight radio links, when the direct wave and the diffracted wave have
opposite phase and tend to cancel. Figure 7.7 shows this effect of diffraction.
FIGURE 7.7 Diffraction
The solution to the problem of interference due to diffraction is to ar-

range for the direct and diffracted signals to be in phase. Again, this requires
a careful survey of the proposed route and adjustment of the transmitting
and receiving antenna heights to achieve this result.
Problems can also occur when the signal reflects from large objects like
cliffs or buildings, as shown in Figure 7.8. There may be not only phase can-
cellation but also significant time differences between the direct and re-
flected waves. These can cause a type of distortion called, not surprisingly,
multipath distortion, in FM reception. Directional receiving antennas
aimed in the direction of the direct signal can reduce the problem of reflec-
tions for fixed receivers.
FIGURE 7.8 Multipath reception

The Mobile In the previous section, we looked at the basics of multipath propagation. In
Environment an environment where both transmitter and receiver are fixed, it is often
possible, as noted above, to position the antennas in such a way as to reduce,
if not eliminate, the effects of multipath interference. In the usual wireless
situation, however, either the transmitter, the receiver, or both, are in con-
stant motion, and the multipath environment is therefore in a constant
state of flux. The luxury of carefully sited, directional antennas does not ex-
ist. In addition, the mobile and portable environment is often very cluttered
with the potential for multiple reflections from vehicles and buildings.
There may also be places where the direct signal is blocked by a tall building,
for example. At such times a reflected signal may actually allow communica-
tion to take place where it would otherwise be impossible. Figure 7.9 shows
some of the possibilities.
FIGURE 7.9
Multipath reception in a
mobile environment
Mobile environments are often so cluttered that the square-law attenua-

tion of free space, which is the basis for Equation (7.16), the path loss equa-
tion, no longer applies because of multipath propagation and shadowing by
obstacles. In fact, the situation is so complex and so dependent on the actual
environment that no one equation can cover all situations. Some approxi-
mations are used, however.
To understand how the mobile environment changes path loss, let us
first have another look at Equation (7.16):
L ƒs = 32.44 + 20 log d + 20 log ƒ

Because this equation is set up to give loss in decibels, the square-law attenu-
ation due to distance is represented by its logarithmic equivalent: 20 log d.
In mobile environments the attenuation increases much more quickly
with distance; values range from 30 log d to a more typical 40 log d. That is,
236 ! CHAPTER 7
attenuation is roughly proportional to the fourth power of distance because

of reflections and obstacles.
Equation (7.16) had no term for antenna height, since this is irrelevant
in free space where there is no ground to measure height from. We looked at
antenna height separately in Equation (7.17), as a limiting factor governing
the maximum distance for line-of-sight propagation. Even for distances
much shorter than the maximum, antenna height is an important determi-
nant of signal strength in mobile systems. The greater the antenna height,
the less important ground reflections become, and the more likely the trans-
mitting and receiving antennas are to have an unobstructed path between
them. Empirical observations show that doubling the height of a base-
station antenna gives about a 6 dB decrease in path loss, while doubling
the height of a (much lower) portable or mobile antenna reduces loss by
about 3 dB.
Path loss increases with frequency according to Equation (7.16), but
only because the effective area of an antenna with a given gain decreases
with frequency. For an ideal isotropic radiator and a receiving antenna of
constant effective area, there is no dependence of free-space attenuation on
frequency. In fact, as we pointed out earlier, the basic square-law attenuation
is valid for waves as different as sound and light. The situation is different in
a mobile environment because the amount of reflection and diffraction of a
wave from an object is a function of the object’s dimensions, in wave-
lengths. At the low-frequency end of the VHF range, only very large objects
such as the ground and buildings cause reflections. As the frequency in-
creases, the wavelength becomes smaller and relatively small objects begin
to reflect the waves. The increase in multipath interference due to this con-
dition may be compensated for in portable operation because signals at UHF
and up have wavelengths small enough to allow them to propagate through
windows, improving the results in such situations as the use of a portable
wireless phone in a car or building.
There is always a considerable loss in signals as they penetrate buildings.
The loss varies with frequency, building construction, size and location of
windows, and so on, but a rough estimate would be 20 dB at 800 MHz (cell
phone range) in a typical steel-reinforced concrete office building, reduced
to about 6 dB if the user is near a window facing the base station.
From the foregoing, it should be obvious that the modeling of terrestrial
propagation is very difficult, especially in a mobile/portable environment.
There are many models; most of them combine theory with observations of
actual propagation in different areas such as open flat terrain, hilly terrain,
suburbs, and cities. Usually a computer is used to calculate signal strengths
in the coverage region. Even so, the models are all quite approximate; they
must be used with care and verified by actual field-strength measurements.
Errors of 10 dB or so are common.
As an example, we shall look at a simplified version of a model, proposed

by Hata, for propagation in a dense urban mobile environment in the fre-
quency range between 150 and 1000 MHz. A mobile antenna height of 2 m is
assumed in this simplified version.
L p = 68.75 + 26.16 log ƒ − 13.82 log h (7.18)

+ (44.9 − 6.55 log h) log d
where
L p = path loss in dB
h = base station antenna height in m
d = distance in km
Several of the factors we have already mentioned can be seen in this
equation. First, note that loss increases with frequency at a greater rate than
in the free-space equation. Next, notice that increased antenna height re-
duces loss. Finally, the loss increases much more quickly with distance than
in the free-space model. Antenna height affects this loss factor as well.
EXAMPLE 7.7 Y
Find the propagation loss for a signal at 800 MHz, with a transmitting an-
tenna height of 30 m, over a distance of 10 km, using:
(a) the free-space model (Equation 7.16)
(b) the mobile-propagation model (Equation 7.18)
SOLUTION
(a) From Equation (7.16),
Lƒs = 32.44 + 20 log d + 20 log ƒ
= 32.44 + 20 log 10 + 20 log 800
= 110.5 dB
(b) From Equation (7.18),
Lp = 68.75 + 26.16 log ƒ − 13.82 log h + (44.9 − 6.55 log h) log d
= 68.75 + 26.16 log 800 − 13.82 log 30
+ (44.9 − 6.55 log 30) log 10
= 159.5 dB
There is almost 50 dB more attenuation in the mobile environment than in

free space.
X
238 ! CHAPTER 7
Another mobile-specific problem is fast fading. As the mobile user trav-

els through the environment, the signal strength tends to increase and de-
crease as the mobile moves between areas of constructive and destructive
interference. As an example, suppose a mobile receiver moves directly away
from the transmitting antenna and toward a reflecting surface, as shown in
Figure 7.10. If the two signals are in phase at a given point, they will add.
As the mobile moves forward a distance of λ/4, the direct path is increased
and the reflected path is reduced by the same amount, resulting in a total
phase shift of 180° and cancellation of the signal. The cancellation will
likely not be complete, but it can be quite severe (up to 50 dB in extreme
cases). When the vehicle moves another distance of λ/4, the signals are once
again in phase. Thus the fades occur each time the car moves a distance of
λ/2. Given the frequency of the signal and the speed of the vehicle, it is easy
to estimate the time between fades. The time between fades is
λ/2 (7.19)
T =
v
λ
=
2v
c
=
2ƒ v
FIGURE 7.10 Fast fading in a mobile environment
EXAMPLE 7.8 Y
An automobile travels at 60 km/hr. Find the time between fades if the car
uses:
(a) a cell phone at 800 MHz
(b) a PCS phone at 1900 MHz
SOLUTION
First convert the car’s speed to m/s.
60 km 60 × 103 m
=
hr 3.6 × 103 s
= 16.7 m/s
Now we can find the fading period using Equation (7.19).

c
(a) T =
2ƒ v
300 × 106
=
2 × 800 × 106 × 167
.
= 11.2 ms
c
(b) T =
2ƒ v
300 × 106
=
2 × 1900 × 106 × 167
.
= 4.7 ms
Notice that the rapidity of the fading increases with both the frequency of
the transmissions and the speed of the vehicle.
X
Repeaters and The previous two sections have shown that antenna height is important for
Cellular Systems line-of-sight communication. If it is desired to achieve maximum communi-
cation range, increasing antenna height increases the distance to the radio
horizon. Even if the distances involved are small, increased antenna height
reduces multipath interference and avoids radio shadows.
Unfortunately, in a typical wireless communication system, the users are
mobile or portable, and there is not much that can be done about their an-
tenna height. Communication directly between mobile or portable users
therefore has limited range and is subject to a great deal of multipath and
shadowing, even at close range. A good example of such a system is citizens’
band (CB) radio. Even with transmitter power of about 4 W, communication
tends to be poor in urban areas.
Modern wireless systems use base stations with elevated antennas. The
use of base stations is necessary to provide a connection with the telephone
240 ! CHAPTER 7
network, and the elevated antennas improve propagation. In addition, it is

possible when required, to use higher transmitter power at the base than at
the mobile or portable unit. Portable units especially should use as little
power as possible to reduce the size and weight of the unit, and especially its
batteries.
The simplest form of base station has a transmitter and receiver on the
same frequency. The mobiles use the same frequency as well. Typically all
units have moderately high transmitter power (on the order of 30 W). The
base-station antenna is located as high as practical in order to obtain wide
coverage. Figure 7.11 shows the setup, which is commonly used for services
such as taxicab dispatching. Most communication is between mobiles and
the base, though mobiles can communicate directly with each other if they
are close enough together. Communication is half-duplex; that is, each
station can both talk and listen, but not at the same time.
FIGURE 7.11
Dispatcher
system
If mobile units need to talk to each other through the base station to
achieve greater range, or if they need full-duplex access to the telephone
system through the base, at least two frequencies must be used. The base
station in this system is called a repeater, and must transmit and receive
simultaneously on at least two frequencies. Normally the same antenna
is used for both transmitting and receiving, and a high-Q filter called a
duplexer, using resonant cavities, is used to separate the transmit and re-
ceive frequencies. Mobiles can communicate with each other through the
repeater, and if the repeater is connected to the PSTN, phone calls can be
made through the repeater.
There is an extensive network of amateur radio repeaters operating on
this principle. The frequencies shown in Figure 7.12 are typical of one of
these systems. Early mobile-phone systems used this method, and it is still in
common use for fixed microwave links.
FIGURE 7.12
Typical repeater
system
The major problem with the dispatcher and repeater systems just de-
scribed is, since they are constructed for maximum coverage using antennas
at large elevations and high-power transmitters, they use spectrum ineffi-
ciently. One telephone conversation ties up two channels over the complete
coverage area of the system, out to the radio horizon. The simplicity of being
able to cover an entire city with one repeater is paid for by this great waste of
spectrum.
Modern cellular systems do not use the radio horizon as the limit of cov-
erage. Antennas may still be mounted quite high in order to reduce
multipath and shadowing, but the range is deliberately limited by using as
low a transmitter power as possible. This enables the same frequencies to be
reused at distances much closer than the radio horizon. Cellular systems are
complex but much more efficient in their use of spectrum.
Figure 7.13 shows a typical cellular system. (The term cellular, here, also
applies to PCS systems, wireless LANs, and so on, because they all use the
same general principle.)
Each repeater is responsible for coverage in a small cell. As shown, the
cells are hexagons with the repeater in the center, but of course in a real situ-
ation the antenna patterns will not achieve this precision and the cells will
have irregular shapes with some overlap.
Since each transmitter operates at low power, it is possible to reuse
frequencies over a relatively short distance. Typically a repeating pattern
of either twelve or seven cells is used, and the available bandwidth is divided
among these cells. The frequencies can then be reused in the next pattern.
To find out how many cells are needed in a pattern, it is necessary
to make some assumptions. First, we assume that all signals are well above
the noise level, so that cochannel interference—interference from transmitters
242 ! CHAPTER 7
FIGURE 7.13
Cellular system
on the same frequency in other cells—is the range-limiting factor. Next,

we assume that only the nearest cells with the same frequency will cause
serious problems. This is usually true. Further, assume that all transmit-
ters have equal power. This is actually a worst-case assumption, because
most cellular systems reduce transmitter power when possible to reduce in-
terference.
We need a criterion for an acceptable level of interference. Zero interfer-
ence is not an option, since the next cell with the same frequency is gener-
ally above the horizon, and its signal strength may be well above the noise
level. The usual assumption is that a signal-to-interference (S/I) ratio of
at least 18 dB is sufficient. This number is based on subjective tests of com-
munication quality with FM voice transmission, and may vary with other
systems.
We should also know how rapidly signals drop off with distance. The
greater the attenuation, within reason, the better in a cellular system be-
cause the interfering signal, coming from a greater distance, will be affected
more than the desired signal. Earlier in this chapter we saw that square-law
attenuation is the rule in free space, but in mobile propagation, especially in
urban areas, the attenuation is more likely to be proportional to the fourth
power of the distance. Let us make that assumption for now, recognizing
that the attenuation may sometimes be less than this, leading to more inter-
ference.
With this information and a knowledge of geometry, we can analyze any
repeating pattern of cells to find the probable S/I ratio. In particular, let us
look at a pattern of seven cells, as shown in Figure 7.14, since it is the one
most often used in practice.
In a seven-cell pattern, there will be six interfering signals with equal
distances from our chosen cell. The interference powers from these six cells
will add. All other interfering signals will be much farther away and can be
FIGURE 7.14
7-Cell repeating
pattern
ignored. Let us place ourselves in the center of a cell, at the cell site, and
assume that a mobile at the edge of our cell, a distance r from the center, is
being interfered with by transmitters in the six nearest cells with the same
number. Since the interfering signals can be anywhere in these cells, we will
assume that they are all at the center of their respective cells, at a distance d
from us. From geometry, the ratio between these distances, which is usually
called q, is
d
q = = 4.6
r
Now we can find the signal-to-interference ratio, S/I. Assuming equal

transmitter powers, and fourth-law attenuation
S d4
= 4
I 6r
q4
=
6
4.64
=
6
= 74.6
= 18.7 dB
The result appears satisfactory at first, but it is actually marginal. Fading

of the desired signal or less attenuation for any of the interfering signals
could easily degrade the signal to the point where it is unusable. The
situation can be improved by reusing the frequencies less often, for instance,
244 ! CHAPTER 7
using a twelve-cell repeating pattern. For such a pattern, q = 6.0, there

are eleven significant interfering signals, and the theoretical performance is
S d4
=
I 11r 4
q4
=
11
6.04
=
11
= 117.8
= 20.7 dB
This represents about a 2 dB improvement, but at the expense of much

less frequency reuse. Another possibility, widely employed in practice, is to
use sectorization. Three directional antennas are located at each base sta-
tion in such a way that each antenna covers an angle of 120 degrees. Each
sector uses a different set of channels. The effect of this is to reduce the num-
ber of interfering signals that must be considered from six to two. The effect
on the S/I ratio is quite dramatic:
S d4
=
I 2r 4
q4
=
2
4.64
=
2
= 223.8
= 23.5 dB
Another way of seeing this is to note that S/I improves by

 6
10 log  
 2
= 10 log 3
= 4.8 dB
The division of cells into sectors has a downside. Since each cell now has
three sectors with different channels, each cell acts like three cells with only
one-third the number of channels. This gives the equivalent in frequency
reuse of a 21-cell repeating pattern; so it is actually less efficient in terms of

frequency reuse than the 12-cell pattern just described. Since cell sites are
expensive, however, the 7-cell pattern is more economical to build.
As the number of users increases, cell sizes can be made smaller by in-
stalling more cell sites, and the frequencies can be reused over closer inter-
vals. This technique is called cell-splitting, and gives cellular systems great
flexibility to adapt to changes in demand over both space and time.
Here is how cell-splitting works. The number of cells required for a given
area is given by:
A
N = (7.20)
a
where
N = number of cells
A = total area to be covered
a = area of one cell
This assumes that all cells are of equal size and that there is no overlap be-
tween cells, but it serves as a reasonably good estimate.
If hexagonal cells are used, there need be no overlap, at least in theory.
(In practice, the cells will not be perfectly hexagonal.) The area of a regular
hexagon (one with all six sides equal) is given by the equation
A = 3.464r 2 (7.21)
where
A = the area of the hexagon
r = the radius of a circle inscribed in the hexagon, as in
Figure 7.15
FIGURE 7.15
Area of hexagon
The required number of cells is:

A
N = (7.22)
a
A
=
3.464 r 2
246 ! CHAPTER 7
The number of cells required is inversely proportional to the square of

the cell radius. Reducing the cell radius by a factor of two increases the num-
ber of cells and therefore the system capacity by a factor of four. Unfortu-
nately, reducing the cell radius by a factor of two also increases the number
of cell sites required by a factor of four. As cell sizes are reduced, the available
spectrum is used more efficiently, but cell-site equipment is not, since the
maximum number of users per cell remains constant.
EXAMPLE 7.9 Y
A metropolitan area of 1000 square km is to be covered by cells with a radius
of 2 km. How many cell sites would be required, assuming hexagonal cells?
SOLUTION
A
N =
3.464 r 2
1000
=
3.464 × 2 2
= 73
X
Users of earlier mobile-radio systems could use the same frequencies
throughout a conversation, but since the cells in a cellular radio system are
relatively small, many calls from or to moving vehicles must be transferred
or handed off from one cell site to another as the vehicles proceed. This re-
quires a change in frequency, since frequencies are not reused in adjacent
cells. The system has to instruct the mobile units to change frequency. The
details will be covered later, but it is worth noting here that cellular schemes
were not practical for portable/mobile use until microprocessors were cheap
and small enough to be incorporated in the phones themselves.
Control of Fading Though mobile and portable communication is more problematic than for
in Mobile Systems fixed locations, there are many techniques that can be used to reduce the prob-
lem of fading. Probably the most obvious is to increase the transmitter power. If
the signal is subject to 20-dB fades, then increasing power by 20 dB should
solve the problem. The trouble with this is that a 20-dB increase requires multi-
plying transmitter power by 100. This may be practical for base stations, but
size and battery-life considerations make it a highly undesirable solution for
mobile and especially portable equipment. A typical portable cell phone pro-
duces about 700 mW of RF power. Increasing this to 70 W would be completely
impractical: the equipment would have to be mounted in a vehicle. Increasing
power can also cause interference problems.
Frequency diversity of the conventional type, where two channels are
used in place of one in each direction, is similarly impractical. Mobile sys-
tems seldom have enough bandwidth to afford this luxury. There is an ex-
ception, however: spread spectrum systems achieve frequency diversity
without an increase in bandwidth. They do this by, in effect, sharing multi-
ple RF channels among multiple voice channels so that each voice channel
has its data bits distributed over many frequencies. Fading of a narrow band
of frequencies causes some loss of data, but this can often be corrected by
error-correcting codes.
CDMA spread-spectrum systems have another advantage in the presence
of multipath interference. Using a special receiver called a rake receiver,
they can receive several data streams at once. These actually contain the
same data, displaced in time because of the different propagation times that
arise from the reflections. The receiver combines the power from the various
streams. This adds another possible type of diversity to the system.
Space diversity can be used with portable and mobile systems. Usually
this consists of positioning two receiving antennas at the base station (two
per sector in a sectorized system), though it is possible to use space diversity
on a vehicle by installing an antenna at each end. Since multipath interfer-
ence is very dependent on the exact phase difference between signals travel-
ing along different paths, moving the antenna, which changes the path
length for both direct and reflected signals, can often cause the interference
to become constructive rather than destructive.
As a vehicle or pedestrian moves, all the path lengths are constantly
changing, causing the signal strength to vary rhythmically as the interfer-
ence cycles between constructive and destructive. The frequency of these
fades depends on the signal wavelength and the vehicle’s speed. Spreading a
digital signal in time by transmitting it more than once or by scrambling bits
so that consecutive bits in the baseband signal are not transmitted in the
same order, can help to reduce the destructive effects of multipath fading in
a mobile environment.
' 7.4 Satellite Propagation

Satellites have been used for many years as repeaters for broadcast and
point-to-point transmission, but their use in wireless communication is rela-
tively new. Several wireless satellite systems are now in operation, however.
248 ! CHAPTER 7
In this chapter we look at the basics of propagation to and from satellites.

The details of practical systems will be covered in Chapter 12.
Many communication satellites use the geosynchronous orbit: that is,
they occupy a circular orbit above the equator, at a distance of 35,784 km
above the earth’s surface. At this height, the satellite’s orbital period is equal
to the time taken by the earth to rotate once, approximately 24 hours. If the
direction of the satellite’s motion is the same as that of the earth’s rotation,
the satellite appears to remain almost stationary above one spot on the
earth’s surface. The satellite can then be said to be geostationary, as illus-
trated in Figure 7.16.
Geostationary satellites are convenient in several respects. Antennas at
fixed earth stations can be aimed once and left in place. Satellites are usable
at all times; they never disappear below the horizon.
Geostationary satellites have many disadvantages for mobile use, how-
ever. Their distance from earth is so great that high-gain antennas, which are
impossible to use in handheld phones, are highly desirable. The transmitter
power required puts a strain on batteries. The path length also causes a time
delay of about 0.25 seconds for a round trip, which is inconvenient. Some
mobile communication systems do employ geostationary satellites, but
most of the new proposals use constellations of satellites at much lower alti-
tudes where signals are stronger and propagation time is shorter.
Satellites that are not geostationary are called orbital satellites, though
of course all satellites are in orbit. They are further categorized by the height
of their orbit above the earth. Low earth orbit (LEO) satellites have orbits
below about 1500 km while medium earth orbit (MEO) systems are from
FIGURE 7.16
Geostationary
satellite orbit
10,000 to 15,000 km in altitude. LEO satellites, because they are much closer
to earth than the others, can have relatively strong signals and short propa-
gation times. However, they do not stay above the horizon very long, so
many satellites are needed for continuous coverage. MEO satellites are a
compromise between LEOs and geostationary satellites.
Satellite propagation generally obeys the free-space equation, provided
the mobile unit has a clear view of the satellite. At VHF and UHF frequencies
the ionosphere tends to change the axis of polarization of radio waves in a
random way so circular polarization is commonly used at these frequencies.
An example will demonstrate the differences among LEO, MEO, and
geostationary satellites in terms of path loss.
EXAMPLE 7.10 Y
Calculate the free-space path loss at a frequency of 1 GHz, for each of the fol-
lowing paths, to a point on the earth directly under the satellite:
(a) from a LEO satellite at an altitude of 500 km

(b) from a MEO satellite at an altitude of 12,000 km
(c) from a geostationary satellite at 36,000 km
SOLUTION
(a) From Equation (7.16),
Lƒs = 32.44 + 20 log d + 20 log ƒ

= 32.44 + 20 log 500 + 20 log 1000
= 146.4 dB
(b) Using the same equation, but d = 12,000 km:
Lƒs = 32.44 + 20 log d + 20 log ƒ

= 32.44 + 20 log 12,000 + 20 log 1000
= 174 dB
(c) Now use d = 36,000 km.
Lƒs = 32.44 + 20 log d + 20 log ƒ

= 32.44 + 20 log 36,000 + 20 log 1000
= 184 dB
Note the great difference in loss between the LEO satellite and the other two.
X
250 ! CHAPTER 7
( Radio waves are transverse electromagnetic waves very similar, except for
frequency, to light waves.
( The impedance of a medium is the ratio of the electrical and magnetic
field strengths in that medium. The impedance of free space is 377 Ω.
( The polarization of a radio wave is the direction of its electric field vector.
( As a radio wave propagates in space, its power density reduces with dis-
tance due to spreading of the waves. Directional antennas can partially
offset this effect.
( Reflection, refraction, and diffraction are possible for radio waves and fol-
low rules similar to those for light.
( Line-of-sight propagation is the system most commonly used at VHF and
above. The radio range is actually somewhat greater than the visible range
and is very dependent on the height of the transmitting and receiving
antennas.
( In mobile systems, reflections and radio shadows generally cause signal
strength to decrease much more rapidly with distance from the transmit-
ter than in free space.
( In cellular systems, propagation distances are deliberately kept shorter
than line-of-sight range in order to allow frequency reuse at shorter dis-
tances.
( Propagation via satellite is useful for long-distance communication.
Geostationary systems are convenient but the very large path loss and
propagation time limit their usefulness for wireless communication.
Low-earth-orbit satellites reduce these problems, but several are needed
for continuous service.
( Equation List
c
vp = (7.1)
∈r
EIRP
PD = (7.6)
4 π r2
30EIRP
E = (7.9)
r
Pr = Pt + Gt + Gr − Lƒs (7.15)
Lƒs = 32.44 + 20 log d + 20 log ƒ (7.16)
d = 17ht + 17hr (7.17)
Lp = 68.75 + 26.16 log ƒ − 13.82 log h + (44.9 − 6.55 log h) log d (7.18)
c
T = (7.19)
2ƒ v
A
N = (7.22)
3.464 r 2
( Key Terms
duplexer combination of filters to separate transmit and receive signals
when both use the same antenna simultaneously
fading reduction in radio signal strength, usually caused by reflection or
absorption of the signal
frequency diversity use of two or more RF channels to transmit the same
information, usually done to avoid fading
geostationary orbit satellite orbit in which the satellite appears to
remain stationary at a point above the equator
geosynchronous orbit satellite orbit in which the satellite’s period of
revolution is equal to the period of rotation of the earth
isotropic radiator an antenna that radiates all power applied to it,
equally in all directions. It is a theoretical construct and not a
practical possibility.
low-earth-orbit (LEO) satellite an artificial satellite orbiting the earth at
an altitude less than about 1500 kilometers
medium-earth-orbit (MEO) satellite a satellite in orbit at a distance
above the earth’s surface of between approximately 8,000 and 20,000
km
multipath distortion distortion of the information signal resulting from
the difference in arrival time in signals arriving via multiple paths of
different lengths
orbital satellite any satellite that is not in a geostationary orbit
252 ! CHAPTER 7
rake receiver a radio receiver that is capable of combining several

received signals with different time delays into one composite signal
sectorization division of a cell into several sections all radiating outward
from a cell site
spatial diversity use of two or more physically separated antennas at one
end of a communication link
( Questions
1. What are the similarities between radio waves and light waves?
2. What is meant by the characteristic impedance of a medium? What is
the characteristic impedance of free space?
3. State the difference between power and power density. Explain why
power density decreases with the square of the distance from a source.
4. A radio wave propagates in such a way that its magnetic field is parallel
with the horizon. What is its polarization?
5. What is an isotropic radiator? Could such a radiator be built? Explain.
6. State three factors that determine the amount of power extracted from a
wave by a receiving antenna.
7. The equation given for calculating path loss in decibels shows the loss
increasing with frequency. Why is this?
8. State two undesirable effects that can be caused by reflections in
line-of-sight communication and explain how they arise.
9. Why is the attenuation greater for mobile communication than for free
space?
10. What is meant by space diversity? How can it be used to improve the
reliability of a communication system?
11. Explain how spread-spectrum systems automatically take advantage of
frequency diversity.
12. What is a rake receiver? How does it reduce the effects of multipath
propagation?
13. What is fast fading and how is it caused?
14. Explain how cellular systems allow for frequency reuse.
15. What is meant by cell-splitting, and why is it done?
16. Why are all geostationary communication satellites at the same dis-
tance from the earth?
17. Why is circular polarization commonly used with VHF and UHF satel-
lite communication?
18. Why are high-gain antennas needed with geostationary satellites?
19. Explain the difference between LEO, MEO, and geostationary satellites.
20. What advantages and disadvantages do LEO satellite systems have for
mobile communication?
( Problems
1. Find the propagation velocity of radio waves in glass with a relative
permittivity of 7.8.
2. Find the wavelength, in free space, for radio waves at each of the follow-
ing frequencies:
(a) 50 kHz
(b) 1 MHz
(c) 23 MHz
(d) 300 MHz
(e) 450 MHz
(f) 12 GHz
3. An isotropic source radiates 100 W of power in free space. At a distance
of 15 km from the source, calculate the power density and the electric
field intensity.
4. A certain antenna has a gain of 7 dB with respect to an isotropic radiator.
(a) What is its effective area if it operates at 200 MHz?
(b) How much power would it absorb from a signal with a field strength
of 50 µV/m?
5. Find the characteristic impedance of glass with a relative permittivity
of 7.8.
6. A transmitter has an output power of 50 W. It is connected to its an-
tenna by a feedline that is 25 meters long and properly matched. The
loss in the feedline is 5 dB/100 m. The antenna has a gain of 8.5 dBi.
(a) How much power reaches the antenna?
(b) What is the EIRP in the direction of maximum antenna gain?
(c) What is the power density 1 km from the antenna in the direction of
maximum gain, assuming free space propagation?
(d) What is the electric field strength at the same place as in (c)?
254 ! CHAPTER 7
7. A satellite transmitter operates at 4 GHz with an antenna gain of 40 dBi.

The receiver, 40,000 km away, has an antenna gain of 50 dBi. If the
transmitter has a power of 8 W (ignoring feedline losses and mismatch),
find:
(a) the EIRP in dBW
(b) the power delivered to the receiver
8. A paging system has a transmitting antenna located 50 m above average
terrain. How far away could the signal be received by a pager carried 1.2
m above the ground?
9. A boat is equipped with a VHF marine radio, which it uses to communi-
cate with other nearby boats and shore stations. If the antenna on the
boat is 2.3 m above the water, calculate the maximum distance for com-
munication with:
(a) another similar boat
(b) a shore station with an antenna on a tower 22 m above the water
level
(c) another boat, but using the shore station as a repeater
10. A PCS signal at 1.9 GHz arrives at an antenna via two paths differing in
length by 19 m.
(a) Calculate the difference in arrival time for the two paths.
(b) Calculate the phase difference between the two signals. (Hint: 360°
is the same as 0° as far as phase is concerned, so multiples of 360°
can be ignored.)
11. Use the mobile-propagation model given in Equation (7.18) to calculate
the loss over a path of 5 km, with a base antenna 25 m above the
ground, for a
(a) cellular telephone at 800 MHz
(b) PCS at 1900 MHz
12. How rapidly would the signal fade if a cell phone (at 800 MHz) is in a car
moving at 100 km/h?
13. Suppose that cells have a 3 km radius and that the vehicle in the previ-
ous problem is moving directly across a cell. How long will it remain in
that cell before it has to be handed off to the next cell?
14. How many hexagonal cells with a 3 km radius would be needed to cover
an area of 600 km2?
15. Two points 50 km apart on earth can communicate using direct radio
link, with a LEO satellite of 500 km or a geostationary satellite with a
path length to the satellite of 36,000 km. Calculate the one-way time
delay for each mode assuming that each satellite is directly over the cen-
ter of the path.
16. For each situation in the preceding question, calculate the path loss at a
frequency of 1 GHz. Assume free-space propagation.
17. A map indicates that a certain spot on the earth is in the 40 dBW con-
tour of a satellite beam, that is, the EIRP from the satellite is 40 dBW in
that direction. What is the actual signal strength at the earth’s surface,
in W/m2, if the satellite is 37,000 km away?
Antennas 8
Objectives
( Explain the basic principles of operation of antenna systems.
( Define radiation resistance and use it to calculate the efficiency of an
antenna.
( Define antenna gain, beamwidth, and front-to-back ratio, and determine
them from a plot of an antenna’s radiation pattern.
( Convert between antenna gains given in dBi and dBd, and use either to
calculate effective isotropic radiated power (EIRP) and effective radiated
power (ERP) for an antenna-transmitter combination.
( Calculate the dimensions of simple practical antennas for a given frequency.
( Identify, explain the operation of, and sketch the approximate radiation
patterns for common types of antennas and antenna arrays.
( Calculate the gain and beamwidth for selected antenna types.
( Explain the use of diversity and downtilt in cellular base station antennas.
258 ! CHAPTER 8
' 8.1 Introduction

So far, we have considered the ways in which electromagnetic waves can
propagate along transmission lines and through space. The antenna is the
interface between these two media and is a very important part of the com-
munication path. In this chapter we will study the basic operating principles
of antennas and look at some of the parameters that describe their perfor-
mance. Some representative examples of practical antennas used in wireless
communication will be analyzed.
Before we begin, you should have two ideas firmly in mind. First, anten-
nas are passive devices. Therefore, the power radiated by a transmitting an-
tenna cannot be greater than the power entering from the transmitter. In
fact, it is always less because of losses. We will speak of antenna gain, but re-
member that gain in one direction results from a concentration of power
and is accompanied by a loss in other directions. Antennas achieve gain the
same way a flashlight reflector increases the brightness of the bulb: by con-
centrating energy.
The second concept to keep in mind is that antennas are reciprocal
devices; that is, the same design works equally well as a transmitting or a re-
ceiving antenna and in fact has the same gain. In wireless communication,
often the same antenna is used for both transmission and reception.
Essentially, the task of a transmitting antenna is to convert the electrical
energy travelling along a transmission line into electromagnetic waves in
space. This process, while difficult to analyze in mathematical detail, should
not be hard to visualize. The energy in the transmission line is contained in
the electric field between the conductors and in the magnetic field sur-
rounding them. All that is needed is to launch these fields (and the energy
they contain) into space.
At the receiving antenna, the electric and magnetic fields in space cause
current to flow in the conductors that make up the antenna. Some of the
energy is thereby transferred from these fields to the transmission line con-
nected to the receiving antenna, and thence to the receiver.
' 8.2 Simple Antennas

The simplest antenna, in terms of its radiation pattern, is the isotropic radia-
tor that was introduced in Chapter 7. Recall that it has zero size, is perfectly
efficient, and radiates power equally in all directions. Though merely a theo-
retical construct, the isotropic radiator makes a good reference with which
ANTENNAS ! 259
to compare the gain and directionality of other antennas. That is because,

even though this antenna cannot be built and tested, its characteristics are
simple and easy to derive.
The half-wave dipole antenna, on the other hand, is a simple, practical an-
tenna which is in common use. An understanding of the half-wave dipole is
important both in its own right and as a basis for the study of more complex
antennas. A half-wave dipole is sketched in Figure 8.1.
FIGURE 8.1
Half-wave dipole
antenna
The word dipole simply means it has two parts, as shown. A dipole
antenna does not have to be one-half wavelength in length like the one
shown in the figure, but this length is handy for impedance matching, as we
shall see. Actually, in practice its length should be slightly less than one-half
the free-space wavelength to allow for capacitive effects. A half-wave dipole
is sometimes called a Hertz antenna, though strictly speaking the term
Hertzian dipole refers to a dipole of infinitesimal length. This, like the iso-
tropic radiator, is a theoretical construct; it is used in the calculation of
antenna radiation patterns.
Typically the length of a half-wave dipole, assuming that the conductor
diameter is much less than the length of the antenna, is 95% of one-half the
wavelength measured in free space. Recall that the free-space wavelength is
given by
c
λ =
ƒ
where
λ = free-space wavelength in meters
c = 300 × 10 6 m/s
260 ! CHAPTER 8
Therefore, the length of a half-wave dipole, in meters, is approximately

c
L = 0.95 × 0.5
ƒ
300 × 106
= 0.95 × 0.5 ×
ƒ
142.5 × 106
=
ƒ
If the frequency is given in megahertz, as is the usual case, this equation

becomes
142.5
L = (8.1)
ƒ
where
L = length of a half-wave dipole in meters
ƒ = frequency in megahertz
For length measurements in feet, the equivalent equation is
468
L = (8.2)
ƒ
where
L = length of a half-wave dipole in feet
ƒ = operating frequency in megahertz
EXAMPLE 8.1 Y
Calculate the length of a half-wave dipole for an operating frequency of
200 MHz.
SOLUTION
142.5
L =
ƒ
142.5
=
200
= 0.7125 m
X
ANTENNAS ! 261
One way to think about the half-wave dipole is to consider an open-

circuited length of parallel-wire transmission line, as shown in Figure 8.2(a).
The line has a voltage maximum at the open end, a current maximum
one-quarter wavelength from the end, and a very high standing-wave ratio.
Now, suppose that the two conductors are separated at a point one-quarter
wavelength from the end, as in Figure 8.2(b). The drawing shows how the
electric field seems to stretch away from the wires. If the process continues,
as in Figure 8.2(c), some of the field detaches itself from the antenna and
helps to form electromagnetic waves that propagate through space.
FIGURE 8.2
Development of the
half-wave dipole
Radiation The radiation of energy from a dipole is quite apparent if we measure the im-
Resistance pedance at the feedpoint in the center of the antenna. An open-circuited lossless
transmission line, as described in Chapter 6, would look like a short circuit at a
distance of one-quarter wavelength from the open end. At distances slightly
greater than or less than one-quarter wavelength, the line would appear reac-
tive. There would never be a resistive component to the feedpoint impedance,
since an open-circuited line has no way of dissipating power.
A lossless half-wave dipole does not dissipate power either, but it does ra-
diate power into space. The effect on the feedpoint impedance is the same as
if a loss had taken place. Whether power is dissipated or radiated, it disap-
pears from the antenna and therefore causes the input impedance to have
a resistive component. The half-wave dipole looks like a resistance of about
70 Ω at its feedpoint.
262 ! CHAPTER 8
The portion of an antenna’s input impedance that is due to power radi-

ated into space is known, appropriately, as the radiation resistance. It is im-
portant to understand that this resistance does not represent loss in the
conductors that make up the antenna.
The ideal antenna just described radiates all the power supplied to it into
space. A real antenna, of course, has losses in the conductor and therefore
has an efficiency less than one. This efficiency can be defined as
Pr
η = (8.3)
Pt
where
P r = radiated power
P t = total power supplied to the antenna
Recalling that P = I 2R, we have
I 2 Rr
η =
I 2 Rt
Rr
= (8.4)
Rt
where
R r = radiation resistance, as seen from the feedpoint
R t = total resistance, as seen from the feedpoint
EXAMPLE 8.2 Y
A dipole antenna has a radiation resistance of 67 Ω and a loss resistance of
5 Ω, measured at the feedpoint. Calculate the efficiency.
SOLUTION
Rr
η =
Rt
67
=
67 + 5
= 0.93 or 93%
X
ANTENNAS ! 263
Unlike the isotropic radiator, the half-wave dipole does not radiate uni-
formly in all directions. The field strength is at its maximum along a line at a
right angle to the antenna and is zero off the ends of the antenna. In order to
describe more precisely the radiation of this and other more complex anten-
nas, we need some way of graphing radiation patterns. These techniques will
be developed in the next section.
' 8.3 Antenna Characteristics

Now that two simple antennas have been described, it is already apparent
that antennas differ in the amount of radiation they emit in various
directions. This is a good time to introduce some terms to describe and quan-
tify the directional characteristics of antennas and to demonstrate methods
of graphing some of them. These terms will be applied to the isotropic and
half-wave dipole antennas here and will also be applied to other antenna
types as each is introduced.
Radiation Pattern The diagrams used in this book follow the three-dimensional coordinate sys-
tem shown in Figure 8.3. As shown in the figure, the x–y plane is horizontal,
and the angle φ is measured from the x axis in the direction of the y axis. The
z axis is vertical and the angle θ is usually measured from the horizontal
plane toward the zenith. This vertical angle, measured upward from the
ground, is called the angle of elevation. Most work with antennas uses posi-
tive angles of elevation, but sometimes (as when the transmitting antenna is
on a tall tower and the receiving antenna is close to it and much lower) we
are interested in angles below the horizon. Different manufacturers handle
below-horizon angles differently as shown in Figure 8.4 and described in the
following paragraphs.
FIGURE 8.3
Three-dimensional
coordinate system
264 ! CHAPTER 8
Figure 8.4 shows two ways in which the radiation pattern of a dipole
can be represented. The three-dimensional picture in Figure 8.4(a) is useful
in showing the general idea and in getting a feel for the characteristics of
the antenna. The two views in Figures 8.4(b) and (c), on the other hand, are
less intuitive but can be used to provide quantitative information about the
FIGURE 8.4 Radiation pattern of half-wave dipole

ANTENNAS ! 265
FIGURE 8.4
(continued)
performance of the antenna. It is the latter type of antenna pattern that is

usually found in manufacturers’ literature, and it is this type of depiction
that will be used in this book.
Notice that polar graph paper is used in Figures 8.4(b) and (c). The angle is
measured in one plane from a reference axis. Because the three-dimensional
space around the antenna is being represented in two dimensions, at least two
views are required to give the complete picture. It is important to choose the
axes carefully to take advantage of whatever symmetry exists. Usually, the hori-
zontal and vertical planes are used; most often the antenna is mounted either
vertically or horizontally, so the axis of the antenna itself can be used as one of
the reference axes.
The two graphs in Figures 8.4(b) and (c) can be thought of as slices of the
three-dimensional pattern of Figure 8.4(a). By convention, the horizon-
tal-plane diagram is oriented so that the direction of maximum radiation (or
one of these directions, if the antenna is bidirectional like a dipole) is toward
the top of the page, and the vertical-plane drawing, not surprisingly, has the
zenith at the top. All manufacturers draw the horizontal scale in the same
way, with 0° at the top, but practices differ for the vertical angle scale. The
266 ! CHAPTER 8
three numbering methods shown on Figure 8.4(b) are all in use by various
manufacturers. No confusion should result if you simply remember that the
top of the diagram represents the zenith (“up” in the drawing represents
“up” in real life). Sometimes the radiation plot for angles below the horizon
is omitted.
In Figure 8.4(a), the dipole itself is drawn to help visualize the antenna
orientation. This is not done in the polar plots, however. These radia-
tion patterns are valid only in the far-field region; an observer must be
far enough away from the antenna that any local capacitive or inductive
coupling is negligible. In practice this means a distance of at least several
wavelengths, and generally an actual receiver is at a much greater distance
than that. From this distance the antenna would be more accurately repre-
sented as a dot in the center of the graph. The space close to the antenna is
called the near-field region and does not have the same directional charac-
teristics.
The distance from the center of the graph represents the strength of the ra-
diation in a given direction. The scale is usually in decibels with respect to some
reference. Often the reference is an isotropic radiator, as in Figure 8.4. Note that
the farthest point on the graph from the center is at 2.14 dBi; that is, the gain of
a lossless dipole, in its direction of maximum radiation, is 2.14 decibels with
respect to an isotropic radiator. Usually, but not always, the pattern is drawn
so that the point of maximum radiation is at the outside of the chart and the
reference level for that point is stated, as shown in the figure.
The half-wave dipole itself is sometimes used as a reference. In that case,
the gain of an antenna may be expressed in decibels with respect to a half-
wave dipole, or dBd for short. Since the gain of a dipole is 2.14 dBi, the gain
of any antenna in dBd is 2.14 dB less than the gain of the same antenna
expressed in dBi. That is,
G(dBd) = G(dBi) − 2.14 dB (8.5)

where
G(dBd) = gain of antenna in decibels with respect to a half-wave
dipole
G(dBi) = gain of the same antenna in decibels with respect to an
isotropic radiator
Obviously, when comparing antennas, it is important to know which
reference antenna was used in gain calculations.
EXAMPLE 8.3 Y
Two antennas have gains of 5.3 dBi and 4.5 dBd, respectively. Which has
greater gain?
ANTENNAS ! 267
SOLUTION
Convert both gains to the same standard. In this case, let us use dBi. Then,
for the second antenna,
G = 4.5 dBd
= 4.5 + 2.14 dBi
= 6.64 dBi
Therefore, the second antenna has higher gain.

X
Gain and The sense in which a half-wave dipole antenna can be said to have gain can
Directivity be seen from Figure 8.5. This sketch shows the pattern of a dipole, from Fig-
ure 8.4(c), superimposed on that of an isotropic radiator. It can be seen that
while the dipole has a gain of 2.14 dBi in certain directions, in others its gain
is negative. If the antennas were to be enclosed by a sphere that would ab-
sorb all the radiated power, the total radiated power would be found to be
the same for both antennas. Remember that for antennas, power gain in one
direction is at the expense of losses in others.
FIGURE 8.5
Isotropic and dipole
antennas
268 ! CHAPTER 8
Sometimes the term directivity is used. This is not quite the same as
gain. Directivity is the gain calculated assuming a lossless antenna. Real an-
tennas have losses, and gain is simply the directivity multiplied by the effi-
ciency of the antenna, that is:
G = Dη (8.6)
where
D = directivity, as a ratio (not in dB)
G = gain, as a ratio (not in dB)
η = efficiency
When an antenna is used for transmitting, the total power emitted by
the antenna is less than that delivered to it by the feedline. Earlier we de-
fined the efficiency as
Pr
η =
Pt
where
η = antenna efficiency
P t = power supplied to the antenna
The figure of 2.14 dBi we have been using for the gain of a lossless dipole
is also the directivity for any dipole. To find the gain of a real (lossy) dipole, it
is necessary first to convert the decibel directivity to a power ratio, then to
multiply by the efficiency.
EXAMPLE 8.4 Y
A dipole antenna has an efficiency of 85%. Calculate its gain in decibels.
SOLUTION
The directivity of 2.14 dBi can be converted to a power ratio:
.
214
D = log −1
10
= 1.638
Now, find the gain:
G = Dη
= 1.638 × 0.85
= 1.39
ANTENNAS ! 269
This gain can easily be converted to dBi:
G(dBi) = 10 log 1.39

= 1.43 dBi
X
Beamwidth Just as a flashlight emits a beam of light, a directional antenna can be said to
emit a beam of radiation in one or more directions. The width of this beam is
defined as the angle between its half-power points. These are also the points
at which the power density is 3 dB less than it is at its maximum point. An in-
spection of Figures 8.4(b) and (c) will show that the half-wave dipole has a
beamwidth of about 78° in one plane and 360° in the other. Many antennas
are much more directional than this, with a narrow beam in both planes.
Front-to-Back As you might expect, the direction of maximum radiation in the horizontal
Ratio plane is considered to be the front of the antenna, and the back is the direc-
tion 180° from the front. For a dipole, the front and back have the same radi-
ation, but this is not always the case. Consider the unidirectional antenna
shown in Figure 8.6: there is a good deal more radiation from the front of
FIGURE 8.6
Unidirectional
antenna
270 ! CHAPTER 8
this antenna than from the back. The ratio between the gains to the front
and back is the front-to-back ratio. It is generally expressed in dB, in which
case it can be found by subtracting the gains in dBi or dBd.
Major and The dipole antenna of Figure 8.4 viewed in the horizontal plane has two
Minor Lobes equal lobes of radiation. The more complex antenna of Figure 8.6, how-
ever, has one major lobe and a number of minor ones. Each of these lobes
has a gain and beamwidth which can be found separately using the diagram.
EXAMPLE 8.5 Y
For the antenna pattern sketched in Figure 8.6, find:
(a) the antenna gain in dBi and dBd
(b) the front-to-back ratio in dB
(c) the beamwidth for the major lobe
(d) the angle, gain and beamwidth for the most important minor lobe
SOLUTION
(a) Since the major lobe just reaches the outer ring of the chart, which is
specified as 13.95 dBi, that is the antenna gain. Antenna plots are usu-
ally, but not always, done this way. The gain of a dipole is 2.14 dBi so the
gain of this antenna with respect to a dipole is:
G(dBd) = G(dBi) − 2.14 dB

= 13.95 dBi − 2.14 dB
= 11.81 dBd
(b) Since the major lobe reaches 0 dB on the chart, all we have to do is look
at the decibel reading for the point 180° from the first point and change
the sign. Thus the front-to-back ratio is about 15 dB.
(c) The beamwidth is the angle between the two points at which the major
lobe is 3 dB down from its maximum. The 3 dB down point is shown as a
dashed circle. From this we can see that the 3 dB beamwidth is about 44°.
(d) The most important minor lobe is to the rear. We discovered in part (b)
that its gain is 15 dB less than for the major lobe. Therefore the gain for
this lobe is 13.95 − 15 = −1.05 dB. Its beamwidth is about 20°.
X
ANTENNAS ! 271
Effective Isotropic In a practical situation we are usually more interested in the power emitted
Radiated Power in a particular direction than in the total radiated power. Looking from a dis-
and Effective tance, it is impossible to tell the difference between a high-powered trans-
Radiated Power mitter using an isotropic antenna and a transmitter of lower power working
into an antenna with gain. In Chapter 7 we defined effective isotropic radi-
ated power (EIRP), which is simply the actual power going into the antenna
multiplied by its gain with respect to an isotropic radiator.
EIRP = P t G t (8.7)
Another similar term that is in common use is effective radiated power
(ERP), which represents the power input multiplied by the antenna gain
measured with respect to a half-wave dipole. Since an ideal half-wave dipole
has a gain of 2.14 dBi, the EIRP is 2.14 dB greater than the ERP for the same
antenna-transmitter combination. That is,
EIRP = ERP + 2.14 dB (8.8)

where
EIRP = effective isotropic radiated power for a given transmitter
and antenna
ERP = effective radiated power for the same transmitter and
antenna
The path loss equations in Chapter 7 require EIRP, but they can easily be
used with ERP values. Simply add 2.14 dB to any ERP value to convert it to
EIRP. Convert the power to dBm or dBW first, if it is not already expressed in
such units.
EXAMPLE 8.6 Y
The ERP of a transmitting station is specified as 17 W in a given direction.
Express this as an EIRP in dBm so that it can be used with the path loss equa-
tions in Chapter 7.
SOLUTION
First convert the ERP to dBm, then add 2.14 dB.
ERP
ERP(dBm) = 10 log
1 mW
= 10 log (17 × 103)
= 42.3 dBm
272 ! CHAPTER 8
EIRP(dBm) = ERP(dBm) + 2.14 dB

= 42.3 + 2.14
= 44.44 dBm
X
Impedance The radiation resistance of a half-wave dipole situated in free space and fed
at the center is approximately 70 Ω. The impedance is completely resistive
at resonance, which occurs when the length of the antenna is about 95% of
the calculated free-space half-wavelength value. The exact length depends
on the diameter of the antenna conductor relative to the wavelength. If the
frequency is above resonance, the feedpoint impedance has an inductive
component; if the frequency is lower than resonance, the antenna imped-
ance is capacitive. Another way of saying the same thing is that an antenna
that is too short appears capacitive, while one that is too long is inductive.
Figure 8.7 shows graphically how reactance varies with frequency.
FIGURE 8.7
Variation of dipole
reactance with
frequency
A center-fed dipole is a balanced device, and should be used with a bal-

anced feedline. If coaxial cable is used, a balun (balanced-to-unbalanced)
transformer should be connected between the cable and the antenna. Balun
transformers are described in Appendix A.
A half-wave dipole does not have to be fed at its midpoint. It is possible
to feed the antenna at some distance from the center in both directions, as
shown in Figure 8.8. This is called a delta match and allows the impedance
ANTENNAS ! 273
FIGURE 8.8
Delta match
to be adjusted to match a transmission line. The greater the separation be-

tween the connections to the antenna, the higher the impedance. Just as for
a center-fed antenna, the impedance is resistive only if the antenna length is
accurate.
Polarization Recall from Chapter 7 that the polarization of a radio wave is the orientation
of its electric field vector. The polarization of the radiation from a half-wave
dipole is easy to determine: it is the same as the axis of the wire. That is, a
horizontal antenna produces horizontally polarized waves, and a vertical
antenna gives vertical polarization.
It is important that the polarization be the same at both ends of a com-
munication path. Wireless communication systems usually use vertical po-
larization because this is more convenient for use with portable and mobile
antennas.
' 8.4 Other Simple Antennas

The half-wave dipole is simple, useful, and very common, but it is by no
means the only type of antenna in use. In this section some other simple
antennas will be introduced. Later, ways of combining antenna elements
into arrays with specific characteristics will be examined.
Folded Dipole Figure 8.9 shows a folded dipole. It is the same length as a standard
half-wave dipole, but it is made with two parallel conductors, joined at both
ends and separated by a distance that is short compared with the length of
the antenna. One of the conductors is broken in the center and connected to
a balanced feedline.
The folded dipole differs in two ways from the ordinary half-wave dipole
described above. It has a wider bandwidth, that is, the range of frequencies
within which its impedance remains approximately resistive is larger than
274 ! CHAPTER 8
FIGURE 8.9
Folded dipole
for the single-conductor dipole. For this reason, it is often used—alone or

with other elements—for television and FM broadcast receiving antennas. It
also has approximately four times the feedpoint impedance of an ordinary
dipole. This accounts for the extensive use of 300-Ω balanced line, known as
twinlead, in TV and FM receiving installations.
It is easy to see why the impedance of a folded dipole is higher than that
of a standard dipole. First, suppose that a voltage V and a current I are
applied to an ordinary half-wave dipole. Then, at resonance when the
feedpoint impedance is resistive, the power supplied is
P = VI
Now consider the folded dipole of Figure 8.9. Looking at the center of
the dipole, the length of the path from the center of the lower conductor to
the center of the upper conductor is one-half wavelength. Therefore, the
currents in the two conductors will be equal in magnitude. If the points were
one-half wavelength apart on a straight transmission line we would say that
they were equal in magnitude but out of phase. Here, however, because the
wire has been folded, the two currents, flowing in opposite directions with
respect to the wire, actually flow in the same direction in space and contrib-
ute equally to the radiation from the antenna.
If a folded dipole and a regular dipole radiate the same amount of power,
the current in each must be the same. However, the current at the feedpoint
of a folded dipole is only one-half the total current. If the feedpoint current
is reduced by one-half, yet the power remains the same, the feedpoint volt-
age must be doubled. That is,
P = VI
 I
= 2V  
 2
ANTENNAS ! 275
Assuming equal power is provided to both antennas, the feedpoint voltage

must be twice as great for the folded dipole. The resistance at the feedpoint
of the ordinary dipole is, by Ohm’s Law,
V
R =
I
For the folded dipole, it will be
2V
R′ =
I/2
4V
=
I
= 4R
Since the current has been divided by two and the voltage multiplied by
two, the folded dipole has four times the feedpoint impedance as the regular
version.
It is also possible to build folded dipoles with different-sized conductors,
and with more than two conductors. In this way, a wide variety of feedpoint
impedances can be produced.
Monopole Many wireless applications require antennas on vehicles. The directional ef-
Antenna fects of a horizontal dipole would be undesirable. A vertical dipole is possi-
ble, but awkward to feed in the center and rather long at some frequencies.
Similar results can be obtained by using a vertical quarter-wave monopole
antenna. It is mounted on a ground plane, which can be the actual ground or
an artificial ground such as the body of a vehicle. The monopole is fed at the
lower end with coaxial cable. The ground conductor of the feedline is con-
nected to the ground plane. See Figure 8.10(a).
The radiation pattern of a quarter-wave monopole in the vertical plane
has the same shape as that of a vertical half-wave dipole in free space. Only
half the pattern is present, however, since there is no underground radia-
tion. In the horizontal plane, of course, a vertical monopole is omni-
directional. Since, assuming no losses, all of the power is radiated into
one-half the pattern of a dipole, this antenna has a power gain of two (or
3 dB) over a dipole in free space. See Figures 8.10(b) and (c) for the radiation
patterns for a monopole.
The input impedance at the base of a quarter-wave monopole is one-half
that of a dipole. This can be explained as follows: with the same current, the
276 ! CHAPTER 8
antenna produces one-half the radiation pattern of a dipole, and therefore

one-half the radiated power. The radiated power is given by
Pr = I 2Rr
where
I = antenna current at the feedpoint
R r = radiation resistance measured at the feedpoint
If the radiated power decreases by a factor of two for a given current, then so
must the feedpoint radiation resistance.
In some mobile and portable applications a quarter wavelength is too
long to be convenient. In that case, the electrical length of the antenna can
be increased by adding inductance to the antenna. This can be done at the
FIGURE 8.10
Monopole
antenna
ANTENNAS ! 277
FIGURE 8.10
(continued)
base or at the center, or the whole antenna can be coiled. The “rubber
duckie” antennas on many handheld transceivers use this technique. Induc-
tors used to increase the effective length of antennas are called loading coils.
Figure 8.11 shows the three types of inductive loading just described.
FIGURE 8.11 Inductive loading

278 ! CHAPTER 8
FIGURE 8.11
(continued)
(c) Helical “rubber duckie” antenna
The Five-Eighths This antenna is often used vertically as either a mobile or base antenna
Wavelength in VHF and UHF systems. Like the quarter-wave monopole, it has omni-
Antenna directional response in the horizontal plane. However, the radiation
is concentrated at a lower angle, resulting in gain in the horizontal direc-
tion, which is often useful. In addition, it has a higher feedpoint imped-
ance and therefore does not require as good a ground, because the current
at the feedpoint is less. The impedance is typically lowered to match that
of a 50-Ω feedline by the use of an impedance-matching section. The circu-
lar section at the base of the antenna in the photograph in Figure 8.12(a)
is an impedance-matching device. Figure 8.12(a) shows a 5/8 wave-
length antenna, and Figure 8.12(b) shows its radiation pattern in the vertical
plane.
ANTENNAS ! 279
(a) Photograph of 5/8 wavelength

antenna
(Courtesy of Cushcraft Corporation)
FIGURE 8.12 Five-Eighths wavelength antenna
The Discone The rather unusual-looking antenna shown in Figure 8.13 is known appro-
Antenna priately as the discone. It is characterized by very wide bandwidth, covering
approximately a ten-to-one frequency range, and an omnidirectional pat-
tern in the horizontal plane. The signal is vertically polarized and the gain is
comparable to that of a dipole. The feedpoint impedance is approximately
50 Ω; the feedpoint is located at the intersection of the disk and the cone.
The disk-cone combination acts as a transformer to match the feedline im-
pedance to the impedance of free space, which is 377 Ω. Typically the length
measured along the surface of the cone is about one-quarter wavelength at
the lowest operating frequency.
The wide bandwidth of the discone makes it a very popular antenna
for general reception in the VHF and UHF ranges. It is a favorite for use
with scanners. These receivers can tune automatically to a large number of
channels in succession and are often used for monitoring emergency ser-
vices. The discone can be used for transmitting but seldom is. Most transmit-
ting stations operate at one frequency or over a narrow band of frequencies.
280 ! CHAPTER 8
FIGURE 8.13
Discone antenna
(Courtesy of Tandy
Corporation)
Simpler antennas with equivalent performance or equally elaborate anten-

nas with better performance, are available when wide bandwidth is not
required.
Helical Antennas A helical antenna is a spiral, usually several wavelengths long. Such an an-
tenna is shown in Figure 8.14. Typically the circumference of each turn is
about one wavelength and the turns are about a one-quarter wavelength
apart.
FIGURE 8.14
Helical antenna
ANTENNAS ! 281
Helical antennas produce circularly polarized waves whose sense is

the same as that of the helix. A helical antenna can be used to receive
circularly polarized waves with the same sense and can also receive plane-
polarized waves with the polarization in any direction. Helical antennas are
often used with VHF and UHF satellite transmissions. Since they respond to
any polarization angle, they avoid the problem of Faraday rotation, which
makes the polarization of waves received from a satellite impossible to
predict.
The gain of a helical antenna is proportional to the number of turns. An
approximate expression for the gain, as a power ratio with respect to an iso-
tropic radiator is:
15 NS ( πD )2
G = (8.8)
λ3
where
G = gain (as a ratio, not in dB), with respect to an isotropic
radiator
N = number of turns in the helix, N > 3
S = turn spacing in meters, S ≅ λ/4
D = diameter of the helix in meters, D ≅ λ/π
The radiation pattern for this antenna has one major lobe and sev-
eral minor lobes. For the major lobe, the 3 dB beamwidth (in degrees) is ap-
proximately
52 λ λ
θ = (8.9)
πD NS
EXAMPLE 8.7 Y
A helical antenna with 8 turns is to be constructed for a frequency of
1.2 GHz.
(a) Calculate the optimum diameter and spacing for the antenna and find
the total length of the antenna.
(b) Calculate the antenna gain in dBi.
(c) Calculate the beamwidth.
282 ! CHAPTER 8
SOLUTION
c
(a) λ =
ƒ
300 × 106
=
1200 × 106
= 0.25 m
λ
D =
π
0.25
=
π
= 0.08 m
= 80 mm
λ
S =
4
0.25
=
4
= .0625 m
= 62.5 mm
The length is just the number of turns multiplied by the turn spacing:
L = NS
= 8 × 62.5 mm
= 500 mm
15 NS ( πD )2
(b) G =
λ3
15 × 8 × 00625
. ( π × 008
. )2
= 3
0.25
= 30.3
= 14.8 dBi
ANTENNAS ! 283
52λ λ
(c) θ =
πD NS
52 × 0.25 0.25
=
π × 0.08 8 × 0.0625
= 36.6°
X
Slot Antenna Figure 8.15 shows a slot antenna, which is actually just a hole in a wave-
guide. The length of the slot is generally one-half wavelength. Its radiation
pattern and gain are similar to those of a dipole with a plane reflector behind
it. It therefore has much less gain than, for instance, a horn antenna. It is sel-
dom used alone but is usually combined with many other slots to make a
phased array (see the next section).
FIGURE 8.15
Slot antenna
Horn Antenna Horn antennas, like those shown in Figure 8.16, can be viewed as impedance
transformers that match waveguide impedances to that of free space. The
examples in the figure represent the most common types. The E- and
H-plane sectoral horns are named for the plane in which the horn flares; the
pyramidal horn flares in both planes. The conical horn is most appropriate
with circular waveguide.
The gain and directivity of horn antennas depend on the type of horn
and its dimensions. Let us examine the pyramidal horn, since that is the
most common type. Its gain is proportional to both of the aperture dimen-
sions, shown on the diagram as dE (for the E-plane) and dH (for the H-plane).
The flare angle is limited by impedance-matching considerations, so high
gain requires a long, unwieldy horn. The equation for gain is
7.5 d E d H
G = (8.10)
λ2
284 ! CHAPTER 8
FIGURE 8.16 Horn antennas
where
G = gain as a power ratio, with respect to an isotropic radiator
d E = E-plane aperture
d H = H-plane aperture
λ = wavelength
The beamwidth is different in the two directions. In the H-plane it is
70λ
θH = (8.11)
dH
where
θ H = H-plane beamwidth in degrees
λ = wavelength
d H = H-plane aperture
and in the E-plane it is
56λ
θE = (8.12)
dE
ANTENNAS ! 285
where
θ E = E-plane beamwidth in degrees
λ = wavelength
d E = E-plane aperture
For practical horns, the gain is often in the vicinity of 20 dBi with a
beamwidth of about 25°. The bandwidth is about the same as that of the as-
sociated waveguide. That is, it works over a frequency range of approxi-
mately 2:1.
EXAMPLE 8.8 Y
A pyramidal horn has an aperture (opening) of 58 mm in the E plane and
78 mm in the H plane. It operates at 10 GHz. Calculate:
(a) its gain in dBi
(b) the beamwidth in the H plane
(c) the beamwidth in the E plane
SOLUTION
(a) At 10 GHz the wavelength is:
c
λ =
ƒ
300 × 106
=
10 × 109
= 0.03 m
7.5 d E d H
G =
λ2
7.5 × 0.058 × 0.078

=
0.032
= 37.7
= 15.8 dBi
286 ! CHAPTER 8
(b) From Equation (8.9),

70λ
θH =
dH
70 × 0.03
=
0.078
= 26.9°
(c) From Equation (8.10),

56λ
θE =
dE
56 × 0.03
=
0.058
= 29°
X
As well as being used with a parabolic reflector, the horn antenna can be
and often is used alone as a simple, rugged antenna with moderate gain.
Patch Antenna A patch antenna consists of a thin metallic patch placed a small fraction of a
wavelength above a conducting ground plane. The patch and ground plane
are separated by a dielectric. Generally, a piece of double-sided circuit board
(with all the copper coating left on one side for the ground plane and the
patch photo-etched on the other) is used. For best results a low-loss circuit
board material should be used. Patch antennas are low in cost, compact at
UHF and microwave frequencies, and have gain on the order of 6 dBi.
The patch conductor can have any shape, but simple geometries are
most common; this simplifies the design and analysis of the antenna. Most
of the radiation is in the direction perpendicular to the plane of the antenna
and on the patch side of the antenna. The ground plane prevents the patch
from radiating very much in the opposite direction.
The half-wave rectangular patch is the most commonly used microstrip
antenna. It is characterized by its length L, width W and thickness d, as
shown in Figure 8.17. The length is approximately equal to one-half wave-
length in the dielectric; the width is not critical but the antenna is often
made square. Radiation is from fringing fields between the patch and the
ground plane. This works best with a relatively thick substrate (a few milli-
meters) having a fairly low relative permittivity, normally between 1 and 4.
ANTENNAS ! 287
FIGURE 8.17
Patch antenna
The patch antenna can be fed by a coaxial cable. The inner conductor of
the coaxial line (sometimes referred to as a probe) is connected to the radiat-
ing patch, while the outer conductor is connected to the ground-plane. An
impedance match for the cable used is obtained by moving the connection
point: the impedance is high (several hundred ohms) at the edge, declin-
ing to zero at the center. Another feed method is use of a microstrip line,
photo-etched on the same substrate as the antenna and connected to
its edge.
EXAMPLE 8.9 Y
Calculate the approximate dimensions for a square patch antenna, for a fre-
quency of 2 GHz, on a substrate with a relative permittivity of 2.
SOLUTION
Recall that wavelength is given by
vp
λ = (8.13)
ƒ
where
λ = wavelength in m
vp = propagation velocity in m/s
ƒ = frequency in Hz
In a dielectric, the propagation velocity is
c
vp = (8.14)
Er
where
c = 300 × 106 m/s
Er = relative permittivity
288 ! CHAPTER 8
Substituting Equation (8.14) into Equation (8.13) we find

c
λ =
ƒ Er
300 × 106
=
2 × 109 2
= 0.106 m
The antenna width and length are each approximately half of this, or
50 mm.
X
' 8.5 Antenna Arrays

The simple elements described above can be combined to build a more elab-
orate antenna. The radiation from the individual elements will combine, re-
sulting in reinforcement in some directions and cancellation in others to
give greater gain and better directional characteristics. For instance, it is of-
ten desirable to have high gain in only one direction, something that is not
possible with the simple antennas previously described. In mobile and por-
table applications, often an omnidirectional pattern in the vertical plane is
wanted, but radiation upward and downward from the antenna is of no use.
In both of these situations, a properly designed array could redirect the un-
wanted radiation in more useful directions.
Arrays can be classified as broadside or end-fire, according to their direc-
tion of maximum radiation. If the maximum radiation is along the main
axis of the antenna (which may or may not coincide with the axis of its indi-
vidual elements), the antenna is an end-fire array. If the maximum radiation
is at right angles to this axis, the array has a broadside configuration. Fig-
ure 8.18 shows the axes of each type.
Antenna arrays can also be classified according to the way in which the
elements are connected. A phased array has all its elements connected to the
feedline. There may be phase-shifting, power-splitting, and impedance-
matching arrangements for individual elements, but all receive power from
the feedline (assuming a transmitting antenna). Since the transmitter can be
said to drive each element by supplying power, these are also called driven
arrays. On the other hand, in some arrays only one element is connected to
the feedline. The others work by absorbing and reradiating power radiated
from the driven element. These are called parasitic elements, and the anten-
nas are known as parasitic arrays.
ANTENNAS ! 289
FIGURE 8.18
Broadside and
end-fire arrays
The following sections describe several common types of arrays that are
useful in wireless communication, and give some general characteristics of
each. Detailed analysis of antenna arrays is usually done with the aid of a
computer.
Phased Arrays Phased arrays can be made by connecting together any of the simple an-
tenna types already discussed. Depending on the geometry of the array and
the phase and current relationships between the elements, the array can be
either broadside or end-fire.
Collinear Array Figure 8.19 shows one type of broadside array using half-
wave dipoles. This is called a collinear array because the axes of the elements
are all along the same line.
Suppose that the collinear antenna in the figure is used for transmitting,
and imagine a receiving antenna placed along the main axis of the antenna.
FIGURE 8.19
Collinear array
290 ! CHAPTER 8
None of the individual elements radiate any energy in this direction, so of

course there will be no signal from the array in this direction. Now move the
hypothetical receiving antenna to a point straight out to one side of the an-
tenna. All of the individual elements will radiate a signal in this direction,
but it is necessary to determine whether these signals add constructively or
destructively. If the elements are in phase, the signals will add. A quick check
shows that this is indeed the case. The half-wave dipoles are linked by quar-
ter-wave transmission-line sections that provide a phase reversal between
adjacent ends. Therefore, all the dipoles will be in phase.
Collinear antennas are often mounted with the main axis vertical. They
are then omnidirectional in the horizontal plane but have a narrow angle of
radiation in the vertical plane. Thus, they make good base-station antennas
for mobile radio systems. Many cellular radio and PCS base stations use col-
linear antennas.
Broadside Array Although only one of many broadside arrays (the collin-
ear array above also qualifies, for example), the configuration of dipoles
shown in Figure 8.20 is often given that name.
FIGURE 8.20
Broadside array
This time the elements are not collinear, but they are still in phase. That
may not be obvious because of the crossed lines connecting them, but notice
that the separation between the elements is one-half wavelength. This
would cause a 180° phase shift as the signal travels along the feedline from
one element to the next. This phase shift is cancelled by the crossing of the
transmission line section that joins the dipoles.
Although this antenna, like the previous one, is a broadside array, its pat-
tern is not identical. There is no radiation off the end of any of the elements,
so of course there is no radiation in the equivalent direction from the array.
If this antenna were erected with its main axis vertical, it would not be
omnidirectional in the horizontal plane. In addition to having a narrow ver-
tical pattern, it would have a bidirectional pattern in the horizontal plane.
As for radiation off the end of the antenna, each dipole provides a signal in
this direction, but the signals from adjacent elements cancel due to the
ANTENNAS ! 291
one-half wavelength difference in path lengths between adjacent elements.

Figure 8.21 shows a comparison of the radiation patterns in the horizontal
plane for the two types of antennas.
FIGURE 8.21
Collinear-broadside
comparison
Still another difference between these two broadside arrays is in their po-
larization. The polarization for a dipole is in the plane of the antenna: that
is, a horizontal dipole has horizontal polarization, and a vertical dipole has
vertical polarization. If each of the two dipole combinations just mentioned
has its main axis vertical, then the collinear array has vertical polarization
and the broadside array has horizontal polarization. Either can be useful, but
remember that the polarization must be the same at both ends of a radio
link. If one end is a portable or mobile unit, then vertical polarization is the
usual choice.
End-fire Array Figure 8.22 shows an end-fire array using dipoles. It is iden-
tical to the broadside array in Figure 8.20 except that the feedline between
elements is not crossed this time. This causes alternate elements to be 180°
out of phase. Therefore, the radiation from one element cancels that from
the next in the broadside direction. Off the end of the antenna, however, the
radiation from all the elements adds, since the 180° phase shift between ad-
jacent elements is cancelled by the one-half wavelength physical separation
292 ! CHAPTER 8
FIGURE 8.22
End-fire array
between them. Once again the polarization of the array is determined by the
orientation of its constituent elements; this antenna is normally mounted
with its main axis horizontal for radiation in one direction. It can have
either horizontal or vertical polarization depending on whether it is set up
with the dipoles horizontal or vertical.
Turnstile Antenna arrays are not always designed to give directionality and
gain. The turnstile array illustrated in Figure 8.23(a) is a simple combination
of two dipoles designed to give omnidirectional performance in the horizon-
tal plane, with horizontal polarization. The dipoles are fed 90° out of phase.
FIGURE 8.23 Turnstile antenna

ANTENNAS ! 293
The gain of a turnstile antenna is actually about 3 dB less than that of a

single dipole in its direction of maximum radiation, because each of the
elements of the turnstile receives only one-half the transmitter power. Fig-
ure 8.23(b) shows the radiation pattern for a typical turnstile antenna.
Turnstile antennas are often used for FM broadcast reception, where
they give reasonable performance in all directions without the need for a
rotor.
Log-Periodic Dipole Array The log-periodic array derives its name from the
fact that the feedpoint impedance is a periodic function of the operating fre-
quency. Although log-periodic antennas take many forms, perhaps the sim-
plest is the dipole array, illustrated in Figure 8.24. The log-periodic dipole
array is the most popular antenna for television reception.
FIGURE 8.24
Log-periodic
dipole array
The elements are dipoles, with the longest at least one-half wavelength
in length at the lowest operating frequency, and the shortest less than
one-half wavelength at the highest. The ratio between the highest and low-
est frequencies can be 10:1 or more. A balanced feedline is connected to the
narrow end, and power is fed to the other dipoles via a network of crossed
connections as shown. The operation is quite complex, with the dipoles that
are closest to resonance at the operating frequency doing most of the radia-
tion. The gain is typically about 8 dBi.
The design of a log-periodic antenna is based on several equations. A
parameter τ is chosen, with a value that must be less than 1 and is typically
between 0.7 and 0.9. A value toward the larger end of the range gives an
294 ! CHAPTER 8
antenna with better performance but more elements. The variable τ is the ra-
tio between the lengths and spacing of adjacent elements. That is,
L1 L L
τ = = 2 = 3 = ⋅⋅⋅ (8.15)
L2 L3 L4
where
L1, L2 ⋅ ⋅ ⋅ = lengths of the elements, in order from shortest to
longest
and
D1 D D
τ = = 2 = 3 = ⋅⋅⋅ (8.16)
D2 D3 D4
where
D1, D2 ⋅ ⋅ ⋅ = spacings between the elements and the apex of the
angle enclosing them, in order from shortest to
longest.
The angle designated α in the figure is typically about 30°. From simple trig-
onometry it can be shown that
L1 α
= tan (8.17)
2 D1 2
EXAMPLE 8.10 Y
Design a log-periodic antenna to cover the frequency range from 100 to
300 MHz. Use τ = 0.7 and α = 30°.
SOLUTION
In order to get good performance across the frequency range of interest, it is
advisable to design the antenna for a slightly wider bandwidth. For the lon-
gest element, we can use a half-wave dipole cut for 90 MHz, and for the
shortest, one designed for 320 MHz.
142.5
L =
ƒ
For a frequency of 90 MHz

142.5
L =
90
= 158
. m
ANTENNAS ! 295
Similarly, for 320 MHz,
L = 0.445 m
Because of the way the antenna is designed, it is unlikely that elements of ex-
actly these two lengths will be present, but we can start with the shorter one
and simply make sure that the longest element has at least the length calcu-
lated above.
Starting with the first element and using Equation (8.17),
L1 α
= tan
2 D1 2
We have specified L1 = 0.445 m and α = 30°, so
L1
D1 =
α
2 tan
2
0.445
=
2 tan 15°
= 0.83 m
L1 L L
τ = = 2 = 3 = ⋅⋅⋅
L2 L3 L4
τ and L1 are known, so L2 can be calculated.
L1
L2 =
τ
0.445
=
0.7
= 0.636 m
We continue this process until we obtain an element length that is greater

than 1.58 m.
L3 = 0.909 m, L4 = 1.30 m, L5 = 1.85 m
L5 is longer than necessary, so the antenna will need five elements.

296 ! CHAPTER 8
The spacing between the elements can be found from Equation (8.16):
D1 D D
τ = = 2 = 3 = ⋅⋅⋅
D2 D3 D4
Since τ and D1 are known, it is easy to find D2 and then the rest of the spac-
ings, in the same way as the lengths were found above. We get D2 = 1.19 m,
D3 = 1.69 m, D4 = 2.42 m, and D5 = 3.46 m.
X
Phased arrays are not restricted to combinations of simple dipoles. Any

of the other elementary antennas described in the previous section, such as
monopoles or patch antennas, can be combined into an array. For that mat-
ter, an array can be used as one element of a more complex array.
Parasitic Arrays The Yagi-Uda array shown in Figure 8.25 is the most popular type of parasitic
array. It has one driven element, one reflector behind the driven element,
and one or more directors in front of the driven element. The driven element
is a half-wave dipole or folded dipole. The reflector is slightly longer than
one-half wavelength, and the directors are slightly shorter. The spacing be-
tween elements varies, but is typically about 0.2 wavelength. The Yagi-Uda
antenna is often called just the Yagi array.
The Yagi antenna is a unidirectional end-fire array with a single main
lobe in the direction shown in Figure 8.25. The antenna pattern for a typical
Yagi is shown in Figure 8.26. This one has eight elements: one driven, one re-
flector, and six directors. Yagis are often constructed with five or six directors
for a gain of about 10 to 12 dBi, but higher gains, up to about 16 dBi, can be
achieved by using more directors.
FIGURE 8.25
Yagi array
ANTENNAS ! 297
FIGURE 8.26
Radiation pattern for
8-element Yagi
The Yagi is a relatively narrow-band antenna. When optimized for gain,

its usable bandwidth is only about two percent of the operating frequency.
Wider bandwidth can be obtained by varying the length of the directors,
making them shorter as the distance from the driven element increases.
' 8.6 Reflectors

The antennas and arrays described in the preceding section can often be used
with reflecting surfaces to improve their performance. A reflector may consist
of one or more planes, or it may be parabolic in shape. In order to reduce wind
and snow loads, reflectors are often constructed of mesh or closely-spaced rods.
As long as the spacing is small compared with a wavelength, the effect on the
antenna pattern, compared with a solid reflector, is negligible.
Plane and Corner A plane reflector acts in a similar way to an ordinary mirror. Like a mirror, its
Reflectors effects can be predicted by supposing that there is an “image” of the antenna
on the opposite side of the reflecting surface at the same distance from it as
298 ! CHAPTER 8
the source. Reflection changes the phase angle of a signal by 180°. Whether
the image antenna’s signal aids or opposes the signal from the real antenna
depends on the spacing between the antenna and the reflector and on the lo-
cation of the receiver. In Figure 8.27 the antenna is one-quarter wavelength
from the reflector, and the signals aid in the direction shown. The reflected
signal experiences a 180° phase shift on reflection and another 180° shift be-
cause it must travel an additional one-half wavelength to reach the receiver.
The magnitude of the electric field in the direction shown is thus increased
by a factor of two. The power density in this direction is increased by a factor
of four, or 6 dB, because power is proportional to the square of voltage.
FIGURE 8.27
Plane reflector and
image
It is possible to use a plane reflector with almost any antenna. For example,
a reflector can be placed behind a collinear antenna, as shown in Figure 8.28.
The antenna becomes directional in both the horizontal and vertical planes.
Base antennas for cellular radio systems are often of this type.
The corner reflector creates two images, as shown in Figure 8.29, for a
somewhat sharper pattern. Corner reflectors are often combined with Yagi
arrays in UHF television antennas.
Parabolic Parabolic reflectors have the useful property that any ray originating at a
Reflector point called the focus and striking the reflecting surface will be reflected par-
allel to the axis of the parabola. That is, a collimated beam of radiation will be
produced. The parabolic dish antenna, familiar from backyard satellite-
receiver installations, consists of a small antenna at the focus of a large para-
bolic reflector, which focuses the signal in the same way as the reflector of a
searchlight focuses a light beam. Figure 8.30 shows a typical example. Of
course the antenna is reciprocal: radiation entering the dish along its axis
will be focused by the reflector.
ANTENNAS ! 299
FIGURE 8.29 Corner reflector and images
FIGURE 8.28 Collinear array with plane reflector
FIGURE 8.30 Parabolic antenna

(Courtesy of Andrew Corporation)
300 ! CHAPTER 8
Ideally the antenna at the feedpoint should illuminate the entire surface
of the dish with the same intensity of radiation and should not “spill” any
radiation off the edges of the dish or in other directions. If that were the case,
the gain and beamwidth of the antenna could easily be calculated. The equa-
tion for beamwidth is
70λ
θ = (8.18)
D
where
θ = beamwidth in degrees at the 3 dB points
λ = free-space wavelength in m
D = diameter of the dish in m
The radiation pattern for a typical parabolic antenna is shown in Figure

8.31. The width of the beam measured between the first nulls is approxi-
mately twice the 3 dB beamwidth.
For gain, the equation is
ηπ 2 D 2
G = (8.19)
λ2
FIGURE 8.31
Pattern for typical
parabolic antenna
ANTENNAS ! 301
where
G = gain with respect to an isotropic antenna, as a power ratio
(not in dB)
D = diameter of the dish in m
λ = the free-space wavelength in m
η = the antenna efficiency
The effect of uneven illumination of the antenna, of losses, or of any

radiation from the feed antenna spilling off at the edges, is to reduce the effi-
ciency, and therefore, the gain. To include these effects in gain calculations,
it is necessary to include a constant η, the efficiency of the antenna. η for a
typical antenna is between 0.5 and 0.7.
EXAMPLE 8.11 Y
A parabolic antenna has a diameter of 3 m, an efficiency of 60%, and oper-
ates at a frequency of 4 GHz. Calculate its gain and beamwidth.
SOLUTION
The free-space wavelength is
c
λ =
ƒ
300 × 106
=
4 × 109
= 0.075 m
Substituting this into Equation (8.18), the beamwidth is
70λ
θ =
D
70 × 0.075
=
3
= 1.75°
The gain is given by Equation (8.19):

302 ! CHAPTER 8
ηπ 2 D 2
G =
λ2
0.6π 2 D 2
=
0.0752
= 9475
= 39.8 dBi
X
Any type of antenna can be used with a parabolic reflector. In the micro-
wave portion of the spectrum, where parabolic reflectors are most useful be-
cause they can have a practical size, a horn antenna provides a simple and
efficient method to feed power to the antenna.
Besides the simple horn feed shown in Figure 8.30, there are several ways
to get power to the parabolic reflector. For example, the Gregorian feed
shown in Figure 8.32(a) uses a feed horn in the center of the dish itself,
which radiates to a reflector at the focus of the antenna. This reflects the sig-
nal to the main parabolic reflector. By removing the feedhorn from the fo-
cus, this system allows any waveguide or electronics associated with the
feedpoint to be placed in a more convenient location. The strange-looking
antenna shown in Figure 8.32(b) is a combination of horn and parabolic
antennas called a hog-horn; it is often used for terrestrial microwave links.
FIGURE 8.32
Parabolic antenna
variations
(Courtesy of Andrew
Corporation)
(a) Gregorian feed (b) Hog-horn antenna

ANTENNAS ! 303
' 8.7 Cellular and PCS Antennas

This section considers antennas for 800-MHz cellular radio and 1900-MHz
PCS systems. The requirements are similar, though of course the PCS anten-
nas are physically smaller. Dual-band antennas also exist, both for base and
portable units.
Cell-Site The antennas used for cellular radio systems have to fulfill the requirements
Antennas described in Chapter 7. There is a need for omnidirectional antennas and for
antennas with beamwidths of 120° and less for sectorized cells. Narrower
beamwidths are also useful for filling in dead spots. Typical cellular anten-
nas use variations of the collinear antenna described earlier for omni-
directional patterns and either collinear antennas backed by reflectors or
log-periodic antennas for directional patterns. Figure 8.33 shows patterns
and specifications for a typical omnidirectional antenna, and Figure 8.34
shows specifications for an antenna with a beamwidth of 120°. As expected,
the directional antenna has higher gain.
When cell sizes are small, in high-traffic areas, directional base-station
antennas are often tilted downwards in order to reduce the distance the sig-
nal travels and the interference to neighboring cells. This downtilt can be
done mechanically by mounting the antenna so that it aims downwards at a
slight angle, but it can also be done electrically. Figure 8.35 shows the verti-
cal pattern for a cellular antenna with a built-in downtilt of 9°.
Cellular and PCS base-station receiving antennas are usually mounted in
such a way as to obtain space diversity. For an omnidirectional pattern, typi-
cally three antennas are mounted at the corners of a tower with a triangular
cross section, as shown in Figure 8.36(a). When the cell is divided into three
120° sectors, it is usual to mount two antennas for each sector on the sides of
the tower, as shown in Figure 8.36(b). It is possible for a single antenna to be
used for both receiving and transmitting using a duplexer, but often the
transmitting antenna is located separately. Only one transmitting antenna
is needed for a cell with an omnidirectional pattern; otherwise, one is
needed per sector. Of course, other mounting locations, such as the walls of
buildings, require variations of these arrangements.
A recent development that can reduce the number of antennas required
for diversity is the use of dual-polarization antennas. We noted earlier that
vertical polarization is usual for portable and mobile systems because this
polarization is much easier to implement in the mobile units. However, in a
cluttered mobile environment, signal polarization may be randomized by
reflections. In that case, diversity can be achieved by using two polariza-
tions, typically at 45° angles to the vertical. Dual-polarization antennas can
considerably reduce the number of visible structures needed at a cell site.
304 ! CHAPTER 8
SPECIFICATIONS
Antenna Type: CT1D0F-0080-009

Description: omni, 9 dBi, no downtilt, 7-16 DIN Female
Frequency Band, MHz: 824-896
Gain (dBi) 9.0
Gain (dBd) 7.0
Azimuth Beamwidth (deg) 360
Polarization Type Single/Vertical
Intermodulation <−150 dBc
Downtilt (deg) 0
Impedance (ohms) 50
Return Loss, dB (VSWR) >14.0 (<1.5)
Maximum Input Power (watts) 500
Lightning Protection DC Ground
Input Connector Type 7-16 DIN Female
Dimensions, L × W × D, mm (inches) 2540 (100) × 76 (3.0)
Weight, kg (lb) 11.3 (25)
Radome Material FIBERGLASS
Radome Color Gray
Survival Wind Speed, km/h (mph) 200 (125)
Temperature Range, C, (F) −40/+70 (−40/+158)
Humidity Up to 100%
Traditional Weatherproofing Kit 221213
Omni Mount Part Number 600033
Connector type 1
Pattern Type 0
Polarization 0
Elevation Beamwidth (deg) 10.0
Flat Plate Area, sq cm (sq in) 1135 (176)
Omni Wind Load, N (lbf) 280 (63)
(a) Specifications
FIGURE 8.33 Omnidirectional cellular base antenna

ANTENNAS ! 305
FIGURE 8.33 (continued)

306 ! CHAPTER 8
SPECIFICATIONS
Antenna Type: CTS08-12010-0D

Description: 120 deg., 12 dBi, no downtilt, 7-16 DIN Female
Gain (dBi) 12.0
Gain (dBd) 10.0
Downtilt (deg) 0
Impedance (ohms) 50
Front to Back Ratio (dB) 20
Dimensions, L × W × D, mm (inches) 1140 (44.9) × 267 (10.5) × 127 (5)
Weight, kg (lb) 8 (18)
Radome Material PVC
Radome Color Gray
Wind Load, Frontal, N (lbf) 431 (97)
Wind Load, Lateral, N (lbf) 109 (24)
Wind Load, Back, N (lbf) 565 (127)
Humidity Up to 100%
Tilt Mount Part Number 600899A
Connector type 1
Pattern Type 1
Polarization 0
(a) Specifications
FIGURE 8.34 Directional cellular base antenna

ANTENNAS ! 307

308 ! CHAPTER 8
SPECIFICATIONS
Antenna Type: CTS08-06013-9D

Description: 60 deg., 15 dBi. 9 degree downtilt, 7-16 DIN Female
Gain (dBi) 15.0
Gain (dBd) 13.0
Downtilt (deg) 9
Impedance (ohms) 50
Front to Back Ratio (dB) 25
Dimensions, L × W × D, mm (inches) 1390 (54.7) × 267 (5) × 127 (5)
Weight, kg (lb) 9 (21)
Radome Material UV Protected PVC
Radome Color Gray
Wind Load, Frontal, N (lbf) 525 (118)
Wind Load, Lateral, N (lbf) 133 (30)
Wind Load, Back, N (lbf) 688 (155)
Humidity Up to 100%
Tilt Mount Part Number 600899
Connector type 1
Pattern Type 1
Polarization 0
(a) Specifications
FIGURE 8.35 Cellular base antenna with downtilt

ANTENNAS ! 309

310 ! CHAPTER 8
FIGURE 8.36
Cell-site antenna
mounting
Mobile and The portable and mobile antennas used with cellular and PCS systems have
Portable to be omnidirectional and small, especially in the case of portable phones.
Antennas The latter requirement is, of course, easier to achieve at 1900 MHz than at
800 MHz. Many PCS phones must double as 800-MHz cell phones, however,
so they need an antenna that works well at 800 MHz.
The simplest suitable antenna is a quarter-wave monopole, and these are
the usual antennas supplied with portable phones. For mobile phones,
where compact size is not quite as important, a very common configuration
consists of a quarter-wave antenna with a half-wave antenna mounted col-
linearly above it. The two are connected by a coil which matches imped-
ances. Figure 8.37 shows such an antenna. This combination has a gain of
about 3 dB compared with the quarter-wave monopole.
FIGURE 8.37
Collinear mobile
antenna
ANTENNAS ! 311
( Antennas are reciprocal passive devices that couple electrical energy

between transmission lines and free space.
( The isotropic radiator is convenient for calculations because it emits all
the energy supplied to it equally in all directions. Antenna gain is usually
calculated with reference to an isotropic radiator.
( Gain is obtained in an antenna when more power is emitted in certain
directions than in others.
( The beamwidth of a directional antenna is the angle between the half-
power points of the main lobe.
( The half-wave dipole is a simple practical antenna with a bidirectional
pattern, a small gain over an isotropic radiator, and a feedpoint imped-
ance of about 70 Ω. The dipole is sometimes used instead of an isotropic
antenna as a reference for antenna gain.
( The radiation resistance of an antenna is the equivalent resistance that
appears at its feedpoint due to the radiation of energy into space.
( The polarization of most simple antennas is the same as the axis of the
antenna.
( Simple antennas can be used as elements of arrays, in order to obtain
specified values of gain and directivity.
( Plane and parabolic reflectors can be used to increase the gain and
directivity of antennas.
( Cellular systems use multiple antennas for space diversity and cell
sectorization. Sometimes polarity diversity is used as well.
( Equation List
142.5
L = (8.1)
ƒ
Rr
η = (8.4)
Rt
15 NS ( πD )2
G = (8.8)
λ3
52 λ λ
θ = (8.9)
πD NS
312 ! CHAPTER 8
7.5 d E d H
G = (8.10)
λ2
70λ
θH = (8.11)
dH
56λ
θE = (8.12)
dE
L1 L L
τ = = 2 = 3 = ⋅⋅⋅ (8.15)
L2 L3 L4
D1 D D
τ = = 2 = 3 = ⋅⋅⋅ (8.16)
D2 D3 D4
L1 α
= tan (8.17)
2 D1 2
70λ
θ =
D
(8.18)
ηπ 2 D 2
G = (8.19)
λ2
( Key Terms
angle of elevation angle measured upward from the horizon. Used to
describe antenna patterns and directions
array combination of several antenna elements
beamwidth angle between points in an antenna pattern at which
radiation is 3 dB down from its maximum
dipole any antenna with two sections
directivity gain of an antenna with losses ignored
effective radiated power (ERP) transmitter power that would, if used
with a lossless dipole oriented for maximum gain, produce the same
power density in a given direction as a given transmitting installation
element an antenna used as part of an array
far-field region distance from an antenna great enough to avoid local
magnetic or electrical coupling, and great enough for the antenna to
resemble a point source
ANTENNAS ! 313
monopole an antenna with only one conductor, generally using ground

or a ground plane to represent a second conductor
near-field region the region of space close to an antenna, where the
radiation pattern is disturbed by induced, as well as radiated, electric
and magnetic fields
radiation resistance representation of energy lost from an antenna by
radiation as if it were dissipated in a resistance
( Questions
1. Explain how an antenna can have a property called gain, even though it
is a passive device.
2. Sketch the radiation patterns of an isotropic antenna and of a half-wave
dipole in free space, and explain why a dipole has gain over an isotropic
radiator.
3. What is the significance of radiation resistance, and what would be the
effect on an antenna’s efficiency of reducing the radiation resistance, all
other things being equal?
4. What is the difference between gain and directivity?
5. Suggest two reasons for using a directional antenna.
6. Distinguish between the near and far field of an antenna. Why is it nec-
essary to use the far field for all antenna measurements?
7. How can an antenna’s electrical length be increased without increasing
its physical length?
8. What advantages does a five-eighths wavelength antenna have over a
quarter-wave antenna for mobile use? Does it have any disadvantages?
9. Name two types of antennas that are noted for having particularly wide
bandwidth. Which of these is directional?
10. Name one type of antenna that produces circularly polarized waves.
Under what circumstances is circular polarization desirable?
11. What polarization is generally used with portable and mobile radio sys-
tems, and why?
12. Distinguish between end-fire and broadside arrays. Name and sketch
one example of each.
13. Distinguish between parasitic and phased arrays. Name and sketch one
example of each. What is another term for a phased array?
14. What is the purpose of using a plane reflector with an antenna array?
314 ! CHAPTER 8
15. What is meant by a collinear antenna? Sketch such an antenna and its
radiation pattern, and classify it as broadside or end-fire, phased or par-
asitic.
16. How is a patch antenna constructed?
17. Why are parabolic dish antennas impractical at frequencies below the
microwave region?
18. What is meant by space diversity? How is it implemented in a typical
cell site?
19. How is polarization diversity possible in cellular radio, and why is it useful?
20. Why are base-station antennas sometimes tilted downward in cellular
systems?
( Problems
1. Calculate the length of a practical half-wave dipole for a frequency of
150 MHz.
2. Calculate the efficiency of a dipole with a radiation resistance of 68 Ω
and a total feedpoint resistance of 75 Ω.
3. Given that a half-wave dipole has a gain of 2.14 dBi, calculate the elec-
tric field strength at a distance of 10 km in free space in the direction of
maximum radiation from a half-wave dipole that is fed, by means of
lossless, matched line, by a 15 W transmitter.
4. Refer to the plot in Figure 8.38 and find the gain and beamwidth for the
antenna shown.
FIGURE 8.38
ANTENNAS ! 315
5. Calculate the EIRP in dBW for a 25 W transmitter operating into a

dipole with 90% efficiency.
6. Calculate the length of a quarter-wave monopole antenna for a fre-
quency of 900 MHz.
7. Calculate the optimum length of an automobile FM broadcast antenna,
for operation at 100 MHz.
8. Draw a dimensioned sketch of a discone antenna that will cover the
VHF range from 30 to 300 MHz.
9. A helical antenna consists of 10 turns with a spacing of 10 cm and a
diameter of 12.7 cm.
(a) Calculate the frequency at which this antenna should operate.
(b) Calculate the gain in dBi at the frequency found in part (a).
(c) Calculate the beamwidth at the frequency found in part (a).
10. Assuming the aperture of a pyramidal horn is square, how large does it
have to be to have a gain of 18 dBi at 12 GHz?
11. Calculate the beamwidth in each plane of the antenna referred to in
problem 10.
12. Calculate the size of a square patch antenna for the ISM band at 2.4
GHz, if the circuit board used has a relative permittivity of 3.
13. Characterize each of the arrays sketched in Figure 8.39 as phased or par-
asitic, and as broadside or end-fire. Show with an arrow the direction(s)
of maximum output for each antenna.
FIGURE 8.39
316 ! CHAPTER 8
14. Design a log-periodic antenna to cover the low-VHF television band,

from 54 to 88 MHz. Use α = 30°, τ = 0.7, L1 = 1.5 m.
15. Two horizontal Yagis are stacked with one of them one-half wavelength
above the other and fed in phase. What will be the effect of the stacking
on the radiation
(a) in the horizontal plane?
(b) in the vertical plane?
16. Calculate the minimum diameter for a parabolic antenna with a beam-
width of 2° at a frequency of:
(a) 4 GHz
(b) 12 GHz
17. Calculate the gain of each of the antennas in the previous question, as-
suming an efficiency of 65%.
18. Calculate the effective area of a 3 m dish, with an efficiency of 0.7, at 3
GHz and at 12 GHz. Explain your results.
19. A Yagi antenna has a gain of 10 dBi and a front-to-back ratio of 15 dB. It
is located 15 km from a transmitter with an ERP of 100 kW at a fre-
quency of 100 MHz. The antenna is connected to a receiver via a
matched feedline with a loss of 2 dB. Calculate the signal power sup-
plied to the receiver if the antenna is
(a) pointed directly toward the transmitting antenna.
(b) pointed directly away from the transmitting antenna.
20. A lossless half-wave dipole is located in a region with a field strength of
150 µV/m. Calculate the power that this antenna can deliver to a re-
ceiver if the frequency is:
(a) 100 MHz
(b) 500 MHz
21. Calculate the gain, with respect to an isotropic antenna, of a half-wave
dipole with an efficiency of 90% that is placed one-quarter wavelength
from a plane reflector that reflects 100% of the signal striking it.
Transmitter and
Receiver Circuitry
9
Objectives
( Discuss the requirements and specifications for transmitters and determine
whether a given transmitter is suitable for a particular application.
( Draw block diagrams for several types of transmitters and explain their operation.
( Analyze the operation of transmitter circuits.
( Explain the basic operation of frequency synthesizers and mixers.
( Describe the characteristics of Class A, B, and C amplifiers and decide which
type is the most suitable for a given application.
( Describe the ways in which FM transmitters and receivers differ from those
for AM and explain why they differ.
( Analyze the operation of analog and digital modulators.
( Describe the basic superheterodyne system, and explain why it is the
preferred design for most receivers.
( Explain the requirements for each stage in a receiver.
( Analyze the operation of demodulators for analog and digital signals.
( Analyze specifications for receivers and use them to determine suitability for
a given application.
( Explain how transceivers differ from separate transmitters and receivers.
318 ! CHAPTER 9
' 9.1 Introduction

The transmitter and receiver are very important parts of any radio communi-
cation link. Sometimes, as at most base stations, they are separate compo-
nents; sometimes they are combined into a transceiver. In the latter case, it
may be necessary for the unit to transmit and receive simultaneously, as in a
cellular phone; or it may be sufficient to do one at a time with manual or
voice-actuated switchover, as with most marine and aircraft radios. We begin
by looking separately at the requirements for transmitters and receivers,
then we examine transceivers.
We will find that all transmitters and all receivers have certain similari-
ties. A thorough familiarity with the basic topologies for each will help the
reader understand the many variations that appear. New types of transmit-
ters and receivers appear almost daily, but the underlying structures change
much more slowly.
' 9.2 Transmitters

The function of a transmitter is to generate a modulated signal with sufficient
power, at the right frequency, and to couple that signal into an antenna
feedline. The modulation must be done in such a way that the demodula-
tion process at the receiver can yield a faithful copy of the original modulating
signal.
All of the foregoing applies to any transmitter. Differences among trans-
mitters result from variations in the required power level, carrier frequency,
and modulation type, as well as from special requirements such as portabil-
ity and the ability to be controlled remotely.
Basic Topologies Figure 9.1 shows the block diagrams of some typical transmitters. Real trans-
mitters have almost infinite variety, but most are variations on these struc-
tures. We will look at some of the variations and the reasons for them, as we
proceed through this chapter and in other chapters that examine particular
systems in more detail.
An overview of the figure shows that in all cases a modulated RF signal
is generated and transmitted. In Figure 9.1(a), which represents a typical
transmitter for full-carrier AM, the carrier is generated by a frequency syn-
thesizer and amplified to its full output power before modulation takes
place. The optional frequency multiplier would be used if the required
carrier frequency were higher than could be conveniently generated by the
synthesizer.
TRANSMITTER AND RECEIVER CIRCUITRY ! 319
FIGURE 9.1 Typical transmitter topologies

320 ! CHAPTER 9
By delaying modulation as long as possible, this topology allows all the

RF amplifier stages in the transmitter to operate in a nonlinear mode for
greater efficiency. High-power AM modulators are relatively easy to build, so
this is the preferred topology for AM.
When the modulation involves changing the transmitted frequency as
in FM and FSK, it is usual to modulate the carrier oscillator. In the example
shown in Figure 9.1(b), that oscillator is a frequency synthesizer. The op-
tional frequency multiplier would multiply the frequency deviation as well
as the carrier frequency, which makes it useful in case the modulated
oscillator can not achieve the required deviation.
Figure 9.1(c) shows a slightly more complex design, generally used
where it is more convenient to modulate the signal at a fixed frequency. The
modulated signal is then moved to the required output frequency by an
oscillator-mixer combination, with the oscillator usually taking the form of
a frequency synthesizer. A bandpass filter eliminates undesired mixing com-
ponents and sends the signal to an amplifier chain.
All three transmitter designs end with a multistage amplifier. Two stages
are shown: the driver and power amplifier. More may be required, especially
for high output power levels. The matching circuit matches the power
amplifier to the antenna feedline impedance, which is usually 50 Ω, and also

removes harmonics and other spurious signals from the transmitter output.
The power amplifier must be linear for any signal that has variable ampli-
tude unless, as with Figure 9.1(a), the modulation is done at the transmitter
output.
Next, let us examine the requirements for each stage of our generic
transmitters in more detail.
Frequency Crystal oscillators are very stable but work at only one frequency unless
Synthesizer crystals are changed. Variable-frequency oscillators (VFOs), where the
operating frequency is controlled by LC resonant circuits, are available but
generally quite unstable in frequency. A frequency synthesizer uses a phase-
locked loop to set the frequency of an oscillator to a multiple of a crystal-
controlled reference frequency, combining most of the advantages of both
crystal-controlled and LC oscillators.
Figure 9.2 shows how a simple frequency synthesizer works. A phase-
locked loop controls the frequency of a voltage-controlled oscillator so that
it is always a multiple of a crystal-controlled reference frequency.
ƒ o = Nƒ ref (9.1)
where
ƒ o = output frequency
ƒ ref = reference frequency
N = divider modulus (must be an integer)
By changing the modulus N of the programmable divider, the output fre-
quency can be varied in steps equal to the reference frequency ƒref. Since ƒref
must often be quite small (a few kilohertz), usually a fixed divider is used to
reduce the frequency of a crystal oscillator to the required value.
FIGURE 9.2
Basic frequency
synthesizer
322 ! CHAPTER 9
EXAMPLE 9.1 Y
Using a 25.6-MHz crystal for the reference oscillator, configure a frequency
synthesizer to generate frequencies from 100 to 200 MHz at 100-kHz
intervals.
SOLUTION
The step size is 100 kHz so let us use that value for ƒref. Then
. × 106
256
Q =
100 × 103
= 256
To find the range of values of N, look at the extremes of the frequency

range. At the low end,
100 × 106
N =
100 × 103
= 1000
Similarly, at the high end, N = 2000.

X
Carrier Oscillator The main requirements for the carrier oscillator are frequency stability
and spectral purity. Almost always this means crystal control; that is, the
frequency of the oscillator is determined by the mechanical resonant fre-
quency of a small slab of quartz. Crystal oscillators are available in frequen-
cies from about 100 kHz to tens of megahertz. Their only real drawbacks are
that the frequency is fixed (unless the crystal is changed) and they are lim-
ited in power output. Neither is a significant problem here, as the frequency
can be moved by the synthesizer-mixer combination and the required power
level is achieved by amplification, which takes place after modulation and
mixing.
Analog The type of modulator required depends on the modulation scheme in use.
Modulation This can be either analog or digital, involving modulation of amplitude,
frequency or phase, or a combination of amplitude and phase. The vari-
ous types of modulation were described in Chapters 2 and 4, and digital
baseband processing was described in Chapter 3. Let us look at suitable
modulators for both analog and digital systems, beginning with analog.
While we do this, we should note for each system whether there are ampli-
tude variations in the modulated signal; that is, whether the modulated
signal has an envelope. Signals with envelopes require linear amplification
after the modulator, while those without amplitude variation can be ampli-
fied more efficiently using Class C or switching amplifiers.
Analog systems use full-carrier AM, its variation SSBSC AM, or FM. Full-
carrier AM is a “mature,” perhaps obsolescent, technology, but it is still
in use for CB radio, VHF aircraft radio, and of course standard AM-band and
short-wave broadcasting. SSBSC AM is used in military, commercial,
and amateur HF communication and in some point-to-point microwave sys-
tems. A variation of it is also used for the video portion of terrestrial televi-
sion broadcasting; DSBSC AM is used in FM and television stereo and
television color signals. FM is very widely used in cellular radio, cordless
phones, VHF marine radio, commercial and public-service mobile radio, and
broadcasting.
Amplitude The usual way to achieve full-carrier AM is to use the baseband signal to vary
Modulation the power-supply voltage to a nonlinear (usually Class C) amplifier, as
shown in Figure 9.3. Since the output voltage of such an amplifier is propor-
tional to its supply voltage rather than to the input signal voltage, this
achieves the desired result. If the amplifier is the final stage before the
antenna, then all RF stages in the transmitter can be nonlinear for greater
efficiency.
FIGURE 9.3
AM modulator
324 ! CHAPTER 9
Single-Sideband Suppressed-Carrier AM Generating an SSB signal is much

more complicated than for AM, and is nearly always done at low power
and at only one frequency. The frequency is then shifted by mixing, as
in Figure 9.1(c). The usual way to generate SSB is to first develop a double-
sideband suppressed-carrier signal (DSBSC) using a balanced modulator
(otherwise known as a multiplier).
It can be shown that if the carrier and baseband signals are multiplied
together the result will contain upper and lower sidebands but no carrier.
Suppose that a carrier and a modulation signal, each sine waves with 1 V
peak amplitude, are applied to a multiplier as shown in Figure 9.4.
FIGURE 9.4 DSBSC generation
The output of the multiplier will be
v o = sin ω m t sin ω c t
= 0.5[cos(ω c − ω m )t − cos(ω c + ω m )t] (9.2)
which is a DSBSC signal.

The DSBSC signal can be converted into SSB by using a bandpass filter to
pass the desired sideband while rejecting the other sideband. Figure 9.5 illus-
trates the idea.
FIGURE 9.5 SSBSC generation
EXAMPLE 9.2 Y
An SSBSC generator of the type shown in Figure 9.5 has the following specifi-
cations:
Filter center frequency: 5.000 MHz
Filter bandwidth: 3 kHz
Carrier oscillator frequency: 4.9985 MHz
(a) Which sideband will be passed by the filter?

(b) What frequency should the carrier oscillator have if it is required to gen-
erate the other sideband?
SOLUTION
(a) Since the carrier frequency is at the low end of the filter passband, the
upper sideband will be passed.
326 ! CHAPTER 9
(b) To generate the lower sideband, the carrier frequency should be moved
to the high end of the filter passband, at 5.0015 MHz.
X
Frequency Modulation Frequency modulation generally involves modu-

lating an oscillator, since that is the only way to directly change the trans-
mitted frequency. (There is an indirect way, via phase modulation, but that
is seldom used nowadays.)
Any crystal or LC oscillator can be turned into a voltage-controlled
oscillator (VCO) by using a varactor diode as part of the frequency-
determining circuit. A varactor is a reverse-biased diode whose junction
capacitance varies with the applied bias voltage. The baseband signal modu-
lates the bias voltage to generate FM.
Unfortunately, VCOs using crystal oscillators have a very small maxi-
mum frequency deviation and are able to operate at only one frequency (a
significant drawback). On the other hand, LC oscillators tend to be unstable.
However, by making the modulated VCO part of a frequency synthesizer, it
is possible to build a stable signal source whose frequency can be varied over
a wide range. Figure 9.6 shows how it can be done.
FIGURE 9.6
PLL FM
modulator
As in the frequency synthesizer described earlier, the carrier frequency is

equal to Nƒref. The baseband signal is added to the VCO control voltage from
the phase detector. The change in VCO frequency due to modulation is de-
tected by the phase detector, but the low-pass filter prevents the correction
signal generated by the phase detector from reaching the VCO and canceling
out the modulation. We thus have a combined frequency synthesizer and

FM modulator. Most FM transmitters used for wireless communication use
this scheme or a variation of it.
Digital Digital modulation in wireless communication usually uses either FSK or

Modulation QPSK. QAM can be used in fixed links but tends not to be robust enough for
mobile and portable communication paths.
FSK modulation is identical to analog FM except that there are usually
only two transmitted frequencies rather than an infinite number. A special
case of FSK, mentioned earlier, is AFSK (audio frequency-shift keying). In
that system, an analog FM modulator is supplied with two audio tones, one
for mark and one for space. AFSK is a good choice for sending data over exist-
ing analog voice FM systems.
PSK Modulators The easiest way to understand QPSK modulation is to be-

gin with binary phase-shift keying (BPSK). This system uses two phase angles
180° apart. It can easily be generated with a multiplier, also known as a bal-
anced modulator. All that is necessary to create a working modulator is to
apply the carrier to one input and a voltage to the other input, which is posi-
tive for one binary state and negative for the other. For example, let the data
input be +1V for logic 1, and −1V for logic 0, as shown in Figure 9.7(a). For
simplicity, assume the carrier amplitude is also 1V (peak).
The balanced modulator output is the product of the two inputs. If the
carrier signal is
e c = sin ω c t
then the output for a binary 1 will be the same,
v(t) = sin ω c t
and the output for a binary 0 is
v(t) = −sin ω c t
which is the same signal with a 180° phase shift.

Now let’s look at QPSK. First, of course, we note that this is a dibit sys-
tem; that is, it transmits two bits simultaneously so the serial data has to
be converted to parallel using a shift register. Assume this has been done.
Now we can apply each of the two bits to a separate balanced modulator,
as shown in Figure 9.7(b). We could do it either way, but let us supply the
first bit to the top balanced modulator. We call this the in-phase (I) bit be-
cause the carrier has not been shifted. The second bit we call the quadrature
328 ! CHAPTER 9
FIGURE 9.7
PSK modulators
(Q) bit and apply it to the lower modulator, along with the carrier, shifted
90° in phase, as shown. The BPSK output signals from the two balanced mod-
ulators are summed to give the output. We can show mathematically that
the output is indeed QPSK by looking at each of the four possible inputs, as
follows.
The first of our dibits is applied to the upper balanced modulator. For
the first bit equal to 1, the output is sin ωct, and for the first bit equal to 0,
the output is −sin ωct, as before. Similarly when the second bit is 1, the lower
balanced modulator outputs cos ωct, and for −1, −cos ωct. Now all we have to
do is add the two outputs for each bit and see what happens. A little trigo-
nometry lets us simplify the resulting expressions. A truth table (Table 9.1)
might help.
If we plot these four possibilities in a vector diagram, we see that we have
indeed achieved QPSK. See Figure 9.8.
TABLE 9.1 Truth Table for QPSK Modulator
Input Top BM Out Bottom BM Out Sum Simplified Sum
00 −sin ωct −cos ωct −sin ωct − cos ωct sin (ωct − 135°)
01 −sin ωct cos ωct −sin ωct + cos ωct sin (ωct + 135°)
10 sin ωct −cos ωct sin ωct − cos ωct sin (ωct − 45°)
11 sin ωct cos ωct sin ωct + cos ωct sin (ωct + 45°)
FIGURE 9.8
QPSK vector diagram
Mixing The frequency of a signal, whether modulated or not, can be changed by

combining it with another signal in a circuit called a mixer. The multiplier
described in the previous section is an example of a mixer, as is the ampli-
tude modulator described earlier. In general, any nonlinear circuit will com-
bine two input frequencies to produce the sum and the difference of the
two input frequencies. In the case of the multiplier (also called a balanced
mixer), those are the only frequencies produced. In the more general case,
the two input frequencies will appear along with their harmonics and other
spurious frequencies. Usually, either the sum or the difference frequency sig-
nal is the desired one, and all other output frequencies must be removed by
filtering.
A little math will show how this works. In general a nonlinear device
produces a signal at its output that can be represented by a power series, for
example
v o = Av i + Bv i 2 + Cv i 3 + ⋅ ⋅ ⋅ (9.3)
330 ! CHAPTER 9
where
v o = instantaneous output voltage
v i = instantaneous input voltage
A, B, C, ⋅ ⋅ ⋅ = constants
With a single input frequency, the output contains all the harmonics as
well as the fundamental. As the order of the harmonics increases, the magni-
tude decreases. This is the principle of the frequency multiplier and also the
cause of harmonic distortion.
When the input contains two different frequencies we get cross products
that represent sum and difference frequencies. If the two frequencies are ƒ1
and ƒ2 we get mƒ1 ± nƒ2, where m and n are integers. Usually the most impor-
tant are ƒ1 + ƒ2 and ƒ1 − ƒ2, where ƒ1 is assumed to be higher than ƒ2. When the
cross products are not wanted they are called intermodulation distortion.
Two important special cases of mixers are the balanced mixer already
described, and the square-law mixer. The operation of the latter is described
by a truncated version of Equation (9.3):
v o = Av i + Bv i 2 (9.4)
To see how a square-law mixer works, let us apply two signals to this cir-
cuit, summing them at the input. The signals will be sine waves of different
frequencies. For convenience we can let each signal have an amplitude of 1 V
peak. Then
v i = sin ω 1 t + sin ω 2 t
and
v o = Av i + Bv i 2 (9.5)
= A(sin ω 1 t + sin ω 2 t) + B(sin ω 1 t + sin ω 2 t) 2
= A sin ω 1 t + A sin ω 2 t + B sin 2 ω 1 t + B sin 2 ω 2 t

+ 2B sin ω 1 t sin ω 2 t
The first two terms in Equation (9.5) are simply the input signals multi-
plied by A, which is a gain factor. The next two terms involve the squares of
the input signals, which are signals at twice the input frequency (plus a dc
component). This can be seen from the trigonometric identity
sin 2 A = ½ − ½ cos 2A
The final term is the interesting one. We can use the trigonometric identity
sin A sin B = ½ {cos (A − B) − cos (A + B)}
to expand this term.

2B sin ω 1 t sin ω 2 t = 2 (B/2){cos (ω 1 − ω 2 )t − cos (ω 1 + ω 2 )t}

= B {cos (ω 1 − ω 2 )t − cos (ω 1 + ω 2 )t}
As predicted, there are output signals at the sum and difference of the
two input frequencies, in addition to the input frequencies themselves
and their second harmonics. In a practical application, either the sum or the
difference frequency would be used and the others would be removed by
filtering.
EXAMPLE 9.3 Y
Sine-wave signals with frequencies of 10 MHz and 11 MHz are applied to a
square-law mixer. What frequencies appear at the output?
SOLUTION
Let ƒ1 = 11 MHz and ƒ2 = 10 MHz. Then the output frequencies are as follows.
ƒ1 = 11 MHz
ƒ2 = 10 MKz
2ƒ 1 = 22 MHz
2ƒ 2 = 20 MHz
ƒ1 + ƒ2 = 21 MHz
ƒ1 − ƒ2 = 1 MHz
X
Power The type of power amplification used depends on whether the signal has an
Amplification envelope. If it does, a linear amplifier (Class A or AB) is required. These are
discussed in any analog electronics text. As we have seen, all varieties of AM
including SSBSC and QAM, have envelope variations and require linear
amplification. Perhaps surprisingly, QPSK in its classic form does require lin-
earity. A second look at Figure 9.8 will show why. Suppose that the signal
changes from the 11 state to the 00 state. This requires that the signal pass
through the origin, that is, its amplitude goes through zero.
If the signal has a constant amplitude, as with frequency modulation
whether analog or digital (FSK), nonlinear amplifiers like Class C or Class D
(switching) amplifiers are more efficient and quite satisfactory. If a transmit-
ter is designed for more than one modulation scheme, it needs a linear
power amplifier unless all of the modulation types used are free from linear-
ity requirements.
In order to understand nonlinear amplifiers, it would be useful to look at
a very simple Class C amplifier using a bipolar transistor. The transistor is
332 ! CHAPTER 9
biased beyond cutoff, that is, it is off for most of the input cycle. Yet for max-
imum efficiency, the transistor must almost saturate at peaks of the input cy-
cle. This type of operation minimizes power dissipation in the transistor,
which is zero when the transistor is cut off, low when it is saturated, and
much higher when it is in the normal, linear operating range. Since Class C
amplifiers are biased beyond cutoff, they would be expected to have zero
power dissipation with no input. This is indeed the case, provided the bias is
independent of the input signal.
There remains the problem of distortion. Figure 9.9(a), a simplified cir-
cuit for a Class C amplifier, shows how the distortion is kept to a reasonable
level. The process can be explained using either the time or the frequency
domain. In the time domain, the output tuned circuit is excited once per cy-
cle by a pulse of collector current. This keeps oscillations going in the reso-
nant circuit. They are damped oscillations, of course, but the Q of the circuit
will be high enough to ensure that the amount of damping that takes place
in one cycle is negligible, and the output is a reasonably accurate sine wave.
The frequency-domain explanation is even simpler: the resonant circuit
constitutes a bandpass filter that passes the fundamental frequency and at-
tenuates harmonics and other spurious signals. In many cases, especially
where the amplifier is the final stage of a transmitter, additional filtering
after the amplifier further reduces the output of harmonics.
FIGURE 9.9
Class C amplifier
From the foregoing description, it might seem that Class C amplifiers

would have to have narrow bandwidth. Many of them do, but this is not a
requirement as long as the output is connected to a low-pass filter that will
attenuate all the harmonics that are generated.
Let us look at the circuit of Figure 9.9(a) a little more closely. We begin
with the bias circuit, which is unconventional. Far from being biased in
the middle of the linear operating range, as for Class A, a Class C ampli-
fier must be biased beyond cutoff. For a bipolar transistor, that can mean no
base bias at all, since a base-to-emitter voltage of about 0.7 V is needed for
334 ! CHAPTER 9
conduction. With no signal at its input, the transistor will have no base
current and no collector current and will dissipate no power.
If a signal with a peak voltage of at least 0.7 V is applied to the input of
the amplifier, the transistor will turn on during positive peaks. It might seem at
first glance that an input signal larger than this would turn on the transistor
for more and more of the cycle, so that with large signals the amplifier would
approach Class B operation, but this is not the case. The base-emitter junction
rectifies the input signal, charging CB to a negative dc level that increases with
the amplitude of the input signal. This self-biasing circuit will allow the ampli-
fier to operate in Class C with a fairly large range of input signals.
The collector circuit can be considered next. At the peak of each input
cycle, the transistor turns on almost completely. This effectively connects
the bottom end of the tuned circuit to ground. Current flows through the
coil L1.
When the input voltage decreases a little, the transistor turns off. For the
rest of the cycle, the transistor represents an open circuit between the lower
end of the tuned circuit and ground. The current will continue to flow in the
coil, decreasing gradually until the stored energy in the inductor has been
transferred to capacitor C 1 , which becomes charged. The process then re-
verses, and oscillation takes place. Energy is lost in the resistance of the
inductor and capacitor, and of course energy is transferred to the load, so the
amplitude of the oscillations would gradually be reduced to zero, except
for one thing: Once each cycle, the transistor turns on, another current pulse
is injected, and, because of this, oscillations continue indefinitely at the
same level.
Figure 9.9(b) shows some of the waveforms associated with the Class C
amplifier. Part (i) shows the input signal, and part (ii) shows the actual signal
applied to the base. Note the bias level that is generated by the signal itself.
Part (iii) shows the pulses of collector current, and part (iv), the collector
voltage. It should not be surprising that this reaches a peak of almost 2VCC.
The peak voltage across the inductor must be nearly V CC , since, when the
transistor is conducting, the top end of the coil is connected to VCC and the
bottom end is almost at ground potential (“almost” because, even when
saturated, there will be a small voltage across the transistor). At the other
peak of the cycle, the inductor voltage will have the same magnitude but
opposite polarity and will add to V CC to make the peak collector voltage
nearly equal to 2VCC.
This description implies that the output tuned circuit must be tuned
fairly closely to the operating frequency of the amplifier, and that is indeed
the case. Since the transistor must swing between cutoff and something
close to saturation for Class C operation, it is also implicit in the design that
this amplifier will be nonlinear; that is, doubling the amplitude of the input
signal will not double the output.
Class C amplification can be used with either field-effect or bipolar tran-

sistors. It is also quite commonly used with vacuum tubes in large transmit-
ters. Class D is similar but even more extreme, with the active device acting
strictly as a switch.
Transmitter Transmitters for different functions have some different types of specifica-
Specifications tions, but there are a number of types of ratings that are common to all or
most transmitters. Some of the most common are described in the remainder
of this section.
Frequency Accuracy and Stability The exact requirements for accuracy

and stability of the transmitter frequency vary with the use to which the
transmitter is put and are set by government regulatory bodies: the Federal
Communications Commission in the United States and the Communica-
tions Branch of Industry Canada in Canada, for example. Depending on the
application, frequency accuracy and stability may be specified in hertz or as
a percentage of the operating frequency. It is easy to convert between the
two methods, as the following example shows.
EXAMPLE 9.4 Y
A crystal oscillator is accurate within ±0.005%. How far off frequency could
its output be at 270 MHz?
SOLUTION
The frequency could be out by 0.005% of 270 MHz which is
0.005
∆ƒ = × 270 × 106
100
= 13.5 × 103 Hz
= 13.5 kHz
X
Frequency Agility Frequency agility refers to the ability to change operating

frequency rapidly, without extensive retuning. In a broadcast transmitter
this is not a requirement, since such stations rarely change frequency. When
they do it is a major, time-consuming operation involving extensive
changes to the antenna system as well as the retuning of the transmitter.
With other services, for example citizens’ band radio, the situation is dif-
ferent. Rapid retuning to any of the 40 available channels is essential to any
336 ! CHAPTER 9
modern CB transceiver. Cellular and PCS phones use many more frequencies
and must change frequencies quickly under microprocessor control. In addi-
tion to a frequency synthesizer for setting the actual transmitting frequency,
such transmitters are required to use broadband techniques throughout so
that frequency changes can be made instantly with no retuning. The cellular
and PCS base stations, on the other hand, can use a separate transmitter for
each channel.
Spectral Purity All transmitters produce spurious signals. That is, they
emit signals at frequencies other than those of the carrier and the sidebands
required for the modulation scheme in use. Spurious signals are often har-
monics of the operating frequency or of the carrier oscillator if it operates at
a different frequency.
All amplifiers produce harmonic distortion. Class C and switching am-
plifiers, especially, produce a large amount of harmonic energy. All frequen-
cies except the assigned transmitting frequency must be filtered out to avoid
interference with other transmissions.
The filtering of harmonics can never be perfect, of course, but in mod-
ern, well-designed transmitters it is very effective. Usually harmonics
and other spurious signals are specified in decibels with respect to the un-
modulated carrier level. The abbreviation for this is dBc.
EXAMPLE 9.5 Y
An FM transmitter has a power output of 10 W and a specified maximum
level for its second harmonic of −60 dBc. A reading of −40 dBm for the sec-
ond harmonic is obtained using a spectrum analyzer. Is this reading within
specifications?
SOLUTION
FM is a constant-power modulation scheme so the unmodulated carrier
power is 10 W, which corresponds to 40 dBm. The maximum allowable
second-harmonic power is 60 dB less than this, or −20 dBm. The measured
reading of −40 dBm is less than this, so the transmitter is within specifi-
cations.
X
Power Output Wireless communication systems often require careful cali-

bration of transmitter power. The old “more is better” idea has been com-
pletely discarded in cellular radio systems.
Different modulation schemes require different power measurements.
AM and FM transmitters for analog communication and all digital systems
are generally rated in terms of carrier power output with no modulation.

This specification would be useless for single-sideband suppressed-carrier
systems, so their power is measured at the peak of the transmitted envelope
and known as peak envelope power (PEP).
Measurement of transmitter power is easy when the signal is
unmodulated. Many instruments will even measure power while the trans-
mitter is connected to an antenna. The Thruline™ wattmeter shown in Fig-
ure 9.10 is an example. With modulated signals that have an envelope,
however, many power meters (including the one shown in the figure) be-
come inaccurate. True-average power reading meters are available: the clas-
sic method is to use a thermistor to measure the rise in temperature in a
resistive load and then calculate the power from that measurement.
FIGURE 9.10
Thruline wattmeter
(Courtesy of Bird
Electronics)
When testing communication transmitters, the technologist should be

aware of the rated duty cycle of the transmitter. Many transmitters designed
for two-way voice communication are not rated for continuous operation at
full power, since it is assumed that the operator will normally be talking for
less than half the time, and probably only for a few seconds to a minute at a
time. This is easy to forget when testing, and the result can be overheated
transistors or other components. Any full-duplex system, as used in cellular
telephone systems, for instance, must be rated for continuous duty.
Efficiency Transmitter efficiency is important for two reasons. The most

obvious one is energy conservation. This is especially important where very
large power levels are involved as in broadcasting, or at the other extreme of
the power-level range, where hand-held operation from batteries is required.
338 ! CHAPTER 9
Another reason for achieving high efficiency becomes apparent when we

consider what happens to the power that enters the transmitter from the
power supply but does not exit via the antenna. It is converted into heat in
the transmitter, and this heat must be dissipated. Large amounts of heat
require large components, heat sinks, fans, and in the case of some high-
powered transmitters, even water cooling. All of these add to the cost of the
equipment. Miniaturized transmitters like those in cell phones have diffi-
culty removing heat quickly enough because of their small surface area, so
high efficiency is very important even at quite low power levels.
When discussing efficiency it is important to distinguish between the ef-
ficiency of an individual stage and that of the entire transmitter. Knowing
the efficiency of an amplifier stage is useful in designing cooling systems
and sizing power supplies. In calculating energy costs, on the other hand,
what is important is the overall efficiency of the system. This is the ratio of
output power to power input from the primary power source, whether it be
the ac power line or a battery. Overall efficiency is reduced by factors such as
losses in the power supply.
Efficiency is easy to measure. The output power must be measured accu-
rately (see the previous section). The input power can be found from the
power supply current and voltage. The efficiency is simply
Po
η = (9.6)
Ps
where
η = efficiency
P o = output power
P s = power from power supply
EXAMPLE 9.6 Y
A mobile transmitter produces 24 W as measured at the antenna. It draws 3.4
A of current when connected to a 13.8 V power source. (Mobile equipment is
usually rated for approximately this voltage, which corresponds to the bat-
tery voltage of a vehicle with a 12 V battery when the engine is running.)
Calculate its efficiency.
SOLUTION
First calculate the supply power
Ps = 13.8 V × 3.4 A
= 46.92 W
Now calculate the efficiency.

Po
η =
Ps
24
=
46.92
= 0.511 or 51.1%
X
Modulation Fidelity As mentioned in Chapter 1, an ideal communication
system allows the original information signal to be recovered exactly, except
for a time delay. Any distortion introduced at the transmitter is likely to re-
main; in most cases it will not be possible to remove it at the receiver. It
might be expected then, that a transmitter would be capable of handling any
baseband frequency to preserve the information signal as much as possible.
In practice, however, the baseband spectrum must be restricted in order to
keep the transmitted bandwidth within legal limits.
In addition, some form of compression, where low-level baseband sig-
nals are amplified more than high-level signals, is often used to keep the
modulation index high. This distorts the original signal by reducing the dy-
namic range, which is the ratio between the levels of the loudest and the qui-
etest passages in the audio signal. The result is an improved signal-to-noise
ratio at the receiver, at the expense of some distortion. The effects of com-
pression, while unpleasant when used with music, are not usually objection-
able in voice communication.
The effects of compression on dynamic range can be removed by apply-
ing an equal and opposite expansion at the receiver. Such expansion would
involve giving more gain to signals at higher levels. The combination of
compression and expansion, called companding, is quite common in com-
munication systems.
Other kinds of distortion, such as harmonic and intermodulation distor-
tion, also have to be kept within reasonable limits. As would be expected,
low distortion levels are more important in the broadcast service than in the
mobile-radio service.
' 9.3 Receivers

The receiver performs an inverse function to that of the transmitter. It must
separate the desired signal from others present at the antenna, amplify it
greatly (often 100 dB or more), and demodulate it to recover the original
340 ! CHAPTER 9
baseband signal. There are many ways to do this, but almost all of them use
the superheterodyne principle, invented by Edwin Armstrong in 1918.
The Basic Figure 9.11 shows the basic topology of a superheterodyne receiver. As for
Superheterodyne transmitters, there are infinite variations of this basic layout.
Receiver The signal chain begins with one or more stages of RF amplification.
Low-cost receivers sometimes omit the RF amplifier, but they do include
some sort of input filter, such as a tuned circuit. The input filter and RF am-
plifier are sometimes referred to as the front end of a receiver.
The next stage is a mixer. The signal frequency is mixed with a sine-wave
signal generated by an associated stage called the local oscillator. A difference
frequency is created, which is called the intermediate frequency (IF). The local
oscillator is tunable, so the IF is fixed regardless of the signal frequency. The
combination of mixer and local oscillator is known as a converter.
The mixer is followed by the IF amplifier, which provides most of the re-
ceiver’s gain and selectivity. Selectivity is the ability to separate signals
from interference and noise. Generally there are two or more IF stages, with
selectivity provided either by resonant circuits or, in recent designs, by a
crystal or ceramic filter. The use of a fixed IF greatly simplifies the problem of
achieving adequate gain and selectivity.
The remainder of the receiver is straightforward. There is a demodulator
followed by whatever baseband processing is required to restore the original
signal.
The automatic gain control (AGC) adjusts the gain of the IF—and some-
times the RF—stages in response to the strength of the received signal,
FIGURE 9.11 Superheterodyne receiver

providing more gain for weak signals. This allows the receiver to cope with
the very large variation in signal levels found in practice. This range may be
more than 100 dB for a communications receiver.
Now that we have looked at the general structure of a typical superhet-
erodyne receiver, it is time to consider the design of the various stages in a
little more detail, starting at the input and working our way through to the
output.
The Front End Inexpensive receivers often omit the RF amplifier stage, especially if they are
designed for operation at low to medium frequencies where atmospheric
noise entering the receiver with the signal is likely to be more significant
than noise generated within the receiver itself. Nearly all other receivers use
a single RF stage to provide gain before the relatively noisy mixer.
It is not necessary for the front end to have a small enough bandwidth to
separate the desired signal from others on nearby frequencies. This selectiv-
ity is more easily provided in the IF amplifier. On the other hand, it is neces-
sary for this stage to reject signals that, even though they are not close in
frequency to the desired signal, may cause problems in later stages. Some-
times the RF stage is tuned along with the local oscillator, but more often
in current designs an RF stage uses relatively broadband filters so that the
frequency band of interest is covered but spurious responses are excluded.
The RF stage is a Class A amplifier. It should have a good noise figure and
a wide dynamic range. AGC can be used, but designers often prefer not to
apply it to the RF stage because any alteration of the stage gain from the
optimum level will degrade the noise figure.
A switch is sometimes provided to remove the RF stage from the signal
path for strong signals to prevent overloading; such signals are applied
directly to the mixer after going through the input filter. Another position
on the same switch may add a few decibels of attenuation to prevent very
strong signals from overloading the mixer. Of course, both removing the RF
amplifier and adding attenuation have adverse effects on the receiver’s noise
performance, but for very strong signals it is less important to maintain the
noise figure than to avoid overloading.
The Mixer and Any type of mixer circuit will work. Diode mixers are generally rejected as
Local Oscillator too noisy and too lossy, except in the simplest receivers. Either bipolar tran-
sistors or FETs can be used as mixers, but the latter are preferred because they
create fewer intermodulation distortion components. The problem here, as
in the RF amplifier, is not so much distortion of the modulating signal as the
creation of spurious responses due to interactions between desired and inter-
fering signals.
342 ! CHAPTER 9
It is extremely important that the local oscillator be stable, as any fre-

quency change will result in the receiver drifting away from the station to
which it is tuned. LC oscillators must be carefully designed for stability. This
is expensive, of course, and some designs have used various types of auto-
matic frequency control (AFC), which reduce drift by feeding an error signal
back to the local oscillator.
Another approach to local-oscillator stability is to use crystal control.
The L.O. can be a simple crystal oscillator with a switch to change crystals
for different channels. This becomes unwieldy with more than a few chan-
nels, so crystal-controlled frequency synthesizers have become very popular
in newer designs. The use of a synthesizer also allows remote control, com-
puter control, direct frequency or channel number entry via a keypad, and
other similar conveniences. AFC is not necessary with a properly designed
synthesizer.
Oscillator spectral purity is also important. Any frequency components
at the oscillator output can mix with incoming signals and creating undesir-
able spurious responses. Noise generated by the local oscillator degrades the
noise performance of the receiver and should be minimized. Some designs
use a bandpass filter to remove spurious signal components and noise from
the local oscillator signal before it is applied to the mixer.
Generally the intermediate frequency is chosen to be lower than the
signal frequency to make it easier to provide gain and selectivity in the IF
amplifier. The IF is the difference between the incoming signal frequency
and the local oscillator frequency. The local-oscillator frequency can be
higher than the signal frequency, in which case the receiver is said to use
high-side injection, or lower than the signal frequency (low-side injection).
For high-side injection, we need
ƒ IF = ƒ LO − ƒ SIG (9.7)
so
ƒ LO = ƒ SIG + ƒ IF (9.8)
where
ƒ IF = intermediate frequency
ƒ LO = local-oscillator frequency
ƒ SIG = signal frequency
Similarly, for low-side injection we need
ƒ IF = ƒ SIG − ƒ LO (9.9)
ƒ LO = ƒ SIG − ƒ IF (9.10)
An example will show how this works.
EXAMPLE 9.7 Y
A receiver tunes from 500 MHz to 600 MHz with an IF of 20 MHz. Calculate
the range of local-oscillator frequencies required if the receiver uses:
(a) high-side injection
(b) low-side injection
SOLUTION
(a) For a signal frequency of 500 MHz we require
ƒLO = ƒSIG + ƒIF
= 500 MHz + 20 MHz
= 520 MHz
Similarly, a signal frequency of 600 MHz requires a local-oscillator
frequency of 620 MHz.
(b) For a signal frequency of 500 MHz we need
ƒLO = ƒSIG − ƒIF
= 500 MHz − 20 MHz
= 480 MHz
Similarly, for a signal frequency of 600 MHz, the local oscillator must op-
erate at 580 MHz.
X
The above example shows that if either Equation (9.7) or (9.9) is satis-
fied, a signal will be produced at the intermediate frequency. For any given
local-oscillator and IF frequencies then, there ought to be two signal fre-
quencies that will mix with the local oscillator to produce a signal at the IF.
An example should make this clear.
EXAMPLE 9.8 Y
A receiver has its local oscillator set to 550 MHz and its IF amplifier designed
to work at 20 MHz. Find two signal frequencies that can be received.
SOLUTION
First we use Equation (9.7):
ƒIF = ƒLO − ƒSIG
344 ! CHAPTER 9
ƒSIG = ƒLO − ƒIF

= 550 MHz − 20 MHz
= 530 MHz
Now try Equation (9.9):
ƒIF = ƒSIG − ƒLO
ƒSIG = ƒLO + ƒIF
= 550 MHz + 20 MHz
= 570 MHz
X
Obviously, it is not a good idea to have the receiver tune to two frequen-
cies at once. Which one is received depends on the bandpass filter in the
front end; it must be designed to accept one and reject the other. The unde-
sired signal is called the image frequency. You can see why by looking at
Figure 9.12. Think of the local oscillator frequency as a mirror; the two possi-
ble signals are equidistant from it, on opposite sides. One of these is the
desired frequency; the other is the image.
FIGURE 9.12 Images
IF Amplifier The IF amplifier accounts for most of the receiver’s gain and selectivity. The
classical method to provide this is to use several stages, to give sufficient
gain, coupled by tuned transformers that provide the selectivity. In modern
receivers, other types of bandpass filters (for example, crystal and ceramic
filters) are more popular. The filter is often placed at the beginning of the IF
chain, right after the mixer.
The IF amplifier of any receiver where the signal has an envelope must
be linear to avoid distorting the signal envelope. This requires that the gain
of the amplifier be reduced with strong signals to prevent overloading. The
circuitry for this is called automatic gain control (AGC).
Receivers for constant-amplitude signals, such as FM, need not be linear;

in fact, it is better if they incorporate limiting to avoid any effect on the de-
modulator due to amplitude variations.
It is important to understand the different functions of the bandpass
filters in the front end and in the IF amplifier. Filtering before the mixer is
required to remove image frequencies and any other signals that can mix
with the local oscillator or with each other to produce a signal at the IF. Such
signals cannot be removed after the mixer; they are at the intermediate fre-
quency so they will be passed by the filters in the IF amplifier. The IF filters,
on the other hand, should restrict the receiver bandwidth to that of the sig-
nal, thereby reducing noise and eliminating interfering signals that are close
in frequency to the desired signal.
Demodulation Demodulation, also called detection, is the inverse of modulation. Its purpose
is to restore the original baseband signal. The type of circuit used depends on
the modulation scheme. A few representative types will be examined here.
Analog Demodulators Full-carrier amplitude modulation can be demodu-

lated with a very simple circuit, as shown in Figure 9.13 on page 346. This
accounts for its early popularity. All that is necessary to rectify the signal
is to remove half of the envelope and then filter out the intermediate-
frequency component. A simple RC filter is sufficient for this.
Demodulation of SSBSC AM is more complex. The carrier must be rein-
serted at the receiver. Next a product detector is used. This is the same as the
balanced modulator described earlier, since both are actually modulators. It
can be shown that multiplying the SSB signal with the carrier results in the
original baseband signal with some additional high-frequency components
that are easily filtered out.
Analog FM requires a detector that converts frequency variations to volt-
age variations. Many circuits have been used for this, but two types account
for the overwhelming majority of current FM detectors.
PLL Method The use of a phase-locked loop to demodulate FM signals is

very straightforward. See Figure 9.14 on page 347, for a typical PLL detector.
The incoming FM signal is used to control the frequency of the VCO. As
the incoming frequency varies, the PLL generates a control voltage to
change the VCO frequency, which will follow that of the incoming signal.
This control voltage varies at the same rate as the frequency of the incoming
signal, thus it can be used directly as the output of the circuit. Unlike the
PLLs used in transmitter modulator circuits, this PLL must have a short time
constant so that it can follow the modulation. The capture range of the PLL
346 ! CHAPTER 9
FIGURE 9.13
AM detector
is not important, since the free-running frequency of the VCO will be set
equal to the signal’s carrier frequency at the detector (that is, to the center of
the IF passband). The lock range must be at least twice the maximum devia-
tion of the signal. If it is deliberately made wider, the detector will be able to
FIGURE 9.14
PLL FM detector
function in spite of a small amount of receiver mistuning or local oscillator

drift. Amplitude variations of the input signal will not affect the operation
of this detector, unless they are so great that it stops working altogether.
Quadrature Detector Like the PLL detector, the quadrature detector is adapted
to integrated circuitry. In the quadrature detector (see Figure 9.15), the incom-
ing signal is applied to one input of a phase detector. The signal is also applied
to a phase-shift network. This consists of a capacitor (C1 in the figure) with high
reactance at the carrier frequency, which causes a 90° phase shift (this is the
origin of the term quadrature). The tuned circuit consisting of L1 and C2 is reso-
nant at the carrier frequency. Therefore, it causes no phase shift at the carrier
frequency but does provide a phase shift at other frequencies that will add to or
subtract from the basic 90° shift caused by C1.
The output of the phase-shift network is applied to the second input of
the phase detector. When the input frequency changes, the angle of phase
shift in the quadrature circuit varies, as the resonant circuit becomes induc-
tive or capacitive. The output from the phase detector varies at the signal fre-
quency but has an average value proportional to the amount the phase angle
differs from 90°. Low-pass filtering the output will recover the modulation.
Figure 9.15 shows this function, which is accomplished by a simple first-
order filter consisting of R 2 and C 3 . The cutoff frequency should be well
FIGURE 9.15
Quadrature FM
detector
348 ! CHAPTER 9
above the highest modulating frequency and well below the receiver inter-
mediate frequency.
The phase detector is the same as is used for phase-locked loops. It can be
an analog multiplier (product detector) or a digital gate (either an AND or an
exclusive-OR gate).
Digital Demodulators FSK, being very similar to analog FM, can be de-
tected using similar means. QPSK, which is more common in wireless
use, can be demodulated by the circuit shown in block-diagram form in
Figure 9.16. Note that it is quite similar to the QPSK modulator shown in
Figure 9.7.
FIGURE 9.16 QPSK demodulator
In order to demodulate the QPSK signal correctly, the carrier must be re-
covered from the original signal. This can be done using a phase-locked
loop. The incoming signal is then applied to two multipliers, which are also
fed the carrier signal but with a 90° phase difference between them. The re-
sulting I and Q signals must be applied to a parallel-to-serial converter to
restore the original serial bit stream.
Receiver This section defines and describes some of the specifications commonly
Specifications used to describe and compare receivers. Receivers for specialized applica-
tions often have additional specifications, but those listed below apply to
most kinds of receiver.
Sensitivity The transmitted signal may have a power level, at the transmit-
ting antenna, ranging from milliwatts to hundreds of kilowatts. However,
the losses in the path from transmitter to receiver are so great that the power
of the received signal is often measured in dBf, that is, decibels relative to
one femtowatt (1 fW = 1 × 10−15 W). A great deal of amplification is needed
to achieve a useful power output. To operate an ordinary loudspeaker, for
instance, requires a power on the order of 1 W, or 150 dBf.
Because received signals are often quite weak, noise added by the re-
ceiver itself can be a problem. Most demodulators are inherently noisy and
also operate best with fairly large signals (hundreds of millivolts), so some of
the amplification must take place before demodulation.
The ability to receive weak signals with an acceptable signal-to-noise
ratio is called sensitivity. It is expressed in terms of the voltage or power at
the antenna terminals necessary to achieve a specified signal-to-noise ratio
or some more easily measured equivalent. When digital modulation is used,
the criterion may be bit-error rate rather than signal-to-noise ratio.
Selectivity In addition to noise generated within the receiver, there will be

noise coming in with the signal, as well as interfering signals with frequen-
cies different from that of the desired signal. All of these problems can be
reduced by limiting the receiver bandwidth to that of the signal (including
all its sidebands). The ability to discriminate against interfering signals is
known as selectivity.
Selectivity can be expressed in various ways. The bandwidth of the re-
ceiver at two different levels of attenuation can be specified. The bandwidth
at the points where the signal is 3 or 6 dB down is helpful in determining
whether all the sidebands of the desired signal will be passed without attenu-
ation. To indicate the receiver’s effectiveness in rejecting interference, a
bandwidth for much greater attenuation, for example 60 dB, should also be
given.
The frequency-response curve for an ideal IF filter would have a square
shape with no difference between its bandwidth at 6 dB and 60 dB down.
The closer the two bandwidths are, the better the design. The ratio between
these bandwidths is called the shape factor. That is,
B−60
SF = (9.11)
B−6
where
SF = shape factor
B −60 = the bandwidth at 60 dB down from maximum
B −6 = the bandwidth at 6 dB down from maximum
350 ! CHAPTER 9
The shape factor should be as close to one as possible. The following ex-
ample shows the effect of changing the shape factor on the rejection of inter-
fering signals.
EXAMPLE 9.9 Y
Calculate the shape factors for the two IF response curves shown in Fig-
ure 9.17, and calculate the amount by which the interfering signal shown
would be attenuated in each case.
FIGURE 9.17
SOLUTION
Figure 9.17(a) shows an ideal filter. Since the −6 dB and −60 dB bandwidths
are equal, the shape factor is 1. The interfering signal is attenuated by 60 dB.
In Figure 9.17(b), the −6 dB bandwidth is 8 kHz and the −60 dB band-
width is 16 kHz, so the shape factor is 2. The interfering signal is attenuated
approximately 35 dB compared to the desired signal.
X
Adjacent channel rejection is another way of specifying selectivity that is

commonly used with channelized systems. It is defined as the number of
decibels by which an adjacent channel signal must be stronger than the de-
sired signal for the same receiver output.
Alternate channel rejection is also used in systems, such as FM broadcast-
ing, where stations in the same locality are not assigned to adjacent chan-
nels. The alternate channel is two channels removed from the desired one.
It is also known as the second adjacent channel. Figure 9.18 shows the rela-
tion between adjacent and alternate channels. For example, if 94.1 MHz
were assigned to a station, the adjacent channel at 94.3 MHz would not be
assigned in the same community, but the alternate channel at 94.5 MHz
might be. Any signal at 94.3 MHz, then, will be relatively weak, and a re-
ceiver’s ability to reject strong adjacent channel signals will be less impor-
tant than its ability to reject the alternate channel (in this case, the one at
94.5 MHz).
FIGURE 9.18
Adjacent and
alternate channels
Distortion In addition to good sensitivity and selectivity, an ideal receiver

would reproduce the original modulation exactly. A real receiver, however,
subjects the signal to several types of distortion. They are the same types
352 ! CHAPTER 9
encountered in other systems: harmonic and intermodulation distortion,

uneven frequency response, and phase distortion.
Harmonic distortion occurs when frequencies are generated which are
multiples of those in the original modulating signal.
Intermodulation takes place when frequency components in the original
signal mix in a nonlinear device, creating sum and difference frequencies. A
type of intermodulation peculiar to receivers consists of mixing between the
desired signal and an interfering one that is outside the IF passband of
the receiver but within the passband of the RF stage and is therefore present
in the mixer. This can result in interference from a strong local station that is
not at all close in frequency to the desired signal. In addition, any non-
linearities in the detector or any of the amplification stages can cause
intermodulation between components of the original baseband signal.
EXAMPLE 9.10 Y
A receiver is tuned to a station with a frequency of 100 MHz. A strong signal
with a frequency of 200 MHz is also present at the amplifier. Explain how
intermodulation between these two signals could cause interference.
SOLUTION
See Figure 9.19. Assume that the tuned circuit at the mixer input has insuffi-
cient attenuation at 200 MHz to block the interfering signal completely. (Re-
member that the interfering signal may be much stronger than the desired
signal and therefore may still have similar strength to the desired signal,
even after being attenuated by the input tuned circuit.) The two signals
could mix in the mixer stage to give a difference frequency of 100 MHz, at
the same frequency as the original signal. This interfering signal will then
pass through the rest of the receiver in the same way as the desired signal.
FIGURE 9.19
X
The frequency response of an AM receiver is related to its IF band-

width. Restricting the bandwidth to less than the full width of the signal
reduces the upper limit of response, since the sidebands corresponding to
higher modulating frequencies are farther away from the carrier. For
example, a receiver bandwidth of 10 kHz will allow a modulation frequency
of 5 kHz to be reproduced. Reducing the bandwidth to 6 kHz reduces the
maximum audio frequency that can be reproduced to 3 kHz. In addition,
any unevenness within the IF passband will affect the audio frequency
response.
The relation between frequency response and bandwidth is not so sim-
ple for FM receivers, since a single baseband frequency produces multiple
sets of sidebands. To reduce distortion it is necessary for an FM receiver to
have enough bandwidth to encompass all the sidebands with significant
amplitude.
Phase distortion is slightly more difficult to understand than frequency
response. Of course, the signal at the receiver output will not be in phase
with the input to the transmitter: there will be some time delay, which can
be translated into a phase shift that increases linearly with frequency. Phase
distortion consists of irregular shifts in phase and is quite a common occur-
rence when signals pass through filters. Phase distortion is unimportant for
analog voice communication, but it is crucial that it be minimized when
digital PSK is used.
Dynamic Range As mentioned above, a receiver must operate over a con-

siderable range of signal strengths. The response to weak signals is usually
limited by noise generated within the receiver. On the other hand, signals
that are too strong will overload one or more stages, causing unacceptable
levels of distortion. The ratio between these two signal levels, expressed in
decibels, is the dynamic range of the receiver.
The above description is actually a bit simplistic. A well-designed AGC
system can easily vary the gain of a receiver by 100 dB or so. Thus, almost
any receiver can cope with very wide variations of signal strength, provided
only one signal is present at a time. This range of signal strengths is some-
times referred to as dynamic range, but should properly be called AGC con-
trol range.
The hardest test of receiver dynamic range occurs when two signals with
slightly different frequencies and very different power levels are applied
to the antenna input simultaneously. This can cause overloading of the
receiver input stage by the stronger signal, even though the receiver is
not tuned to its frequency. The result can be blocking (also called desensitiza-
tion or desense), which is a reduction in sensitivity to the desired signal.
Intermodulation between the two signals is also possible.
354 ! CHAPTER 9
EXAMPLE 9.11 Y
A receiver has a sensitivity of 0.5 µV and a blocking dynamic range of 70 dB.
What is the strongest signal that can be present along with a 0.5 µV signal
without blocking taking place?
SOLUTION
Since both signal voltages are across the same impedance, the input imped-
ance of the receiver, the general equation
P1 V
10 log = 20 log 1
P2 V2
can be used. Here, let signal 1 be the stronger. Then

P1
R = 10 log
P2
where
R = dynamic range in dB
So
V1
R = 20 log
V2
Given that R = 70 dB, V2 = 0.5 µV, we find

V1 R
= antilog
V2 20
R
V1 = V2 antilog
20
70
= 05
. µV antilog
20
= 1581 µV
= 1.581 mV
X
Spurious Responses The superheterodyne has many important advantages
over simpler receivers, but it is not without its problems. In particular, it has
a tendency to receive signals at frequencies to which it is not tuned and
sometimes to generate signals internally, interfering with reception. Careful
design can reduce these spurious responses almost to insignificance, but

they will still be present.
Image frequencies have already been discussed. Image rejection is de-
fined as the ratio of voltage gain at the input frequency to which the receiver
is tuned to gain at the image frequency. Image rejection is usually expressed
in decibels.
An image must be rejected prior to mixing. Once it has entered the IF
chain, the image is indistinguishable from the desired signal and impossible
to filter out. Image rejection is accomplished by filters before the mixer.
Using a higher intermediate frequency can improve image rejection by plac-
ing the image farther away in frequency from the desired signal, where it can
more easily be removed by tuned circuits before the mixer.
As well as images, superheterodyne receivers are subject to other prob-
lems. The local oscillator will have harmonics, for instance, and these can
mix with incoming signals to produce spurious responses. The incoming
signal may also have harmonics created by distortion in the RF stage. In
fact, it is possible for a receiver to respond to any frequency given by the
equation:
 m ƒ
ƒ s =   ƒ LO ± IF (9.12)
n n
where
ƒ s = frequency of the spurious response
ƒ LO = local oscillator frequency
ƒ IF = intermediate frequency
m, n = any integers
EXAMPLE 9.12 Y
A receiver has an IF of 1.8 MHz using high-side injection. If it is tuned to
a frequency of 10 MHz, calculate the frequencies which can cause an IF
response, for values of m and n ranging up to 2.
SOLUTION
First, we find ƒLO:
ƒLO = ƒsig + ƒIF
= 10 MHz + 1.8 MHz
= 11.8 MHz
Now the problem is easily solved using Equation (9.12) and a table of
values. All frequencies in the table are in MHz.
356 ! CHAPTER 9
m n (m/n)ƒLO ƒIF/n (m/n)ƒLO + (ƒIF/n) (m/n)ƒLO − (ƒIF/n)
1 1 11.8 1.8 13.6 10
1 2 5.9 0.9 6.8 5.0
2 1 23.6 1.8 25.4 21.8
2 2 11.8 0.9 12.7 10.9
Rearranging the results in the last two columns in order of ascending fre-
quency gives us the frequencies to which the receiver may respond. In MHz
they are:
5.0, 6.8, 10.0, 10.9, 12.7, 13.6, 21.8, 25.4
X
In spite of these problems, however, the superheterodyne remains the

preferred arrangement for almost all receiving applications. It proves to be
easier to improve the design to reduce its problems than to go to a different
system.
' 9.4 Transceivers

In wireless communication systems, base stations often use separate trans-
mitters and receivers, but portable and mobile units generally combine the
transmitter and receiver into one unit called a transceiver. This provides
convenience, reduced size and weight, and sometimes allows costs to be
reduced by using some components for both transmitting and receiving.
Half-Duplex Many wireless communication systems use transceivers that switch from
Transceivers transmit to receive, either using a push-to-talk switch or voice-activated
switching. In this case, it is often possible to use many components for both
transmit and receive functions. Consider the VHF-FM transceiver whose
block diagram is shown in Figure 9.20, for example.
Here, the CPU and the frequency synthesizer that it controls are used for
both transmit and receive. When transmitting, the synthesizer doubles as
carrier oscillator and FM modulator. When receiving, the synthesizer func-
tions as the first local oscillator. Some transceivers also share audio circuitry.
FIGURE 9.20 VHF-FM transceiver
Note that the receiver section of this transceiver uses double conversion;
that is, there is a second local oscillator/mixer combination. The second
local oscillator operates at a fixed frequency; the purpose of the second IF is
to allow there to be a high first IF for good image rejection followed by a low
second IF to simplify the design of a narrowband filter for good selectivity.
Half-duplex operation simplifies the connection of the transceiver to its
antenna. A mechanical switch can be used, but as in the example, more of-
ten the switching is electronic. Since the antenna is never connected to both
transmitter and receiver at the same time, there is no need for filters to sepa-
rate transmitted and received signals.
Mobile and When full-duplex operation is required as it is with cordless and cellular tele-
Portable phones, it is necessary to use two separate channels, one each for transmit
Telephones and receive. They must be far enough apart in frequency that a filter can sep-
arate them, so that the transmitter signal does not overload, and possibly
damage, the receiver. Figure 9.21 shows how this device, called a duplexer,
358 ! CHAPTER 9
fits into the system. Duplexers are also used at base stations to allow trans-
mitters and receivers to use the same antenna simultaneously.
FIGURE 9.21
Use of duplexer
for full-duplex
communication
( Transmitters are required to produce a modulated signal at the correct
power level and frequency and with acceptable levels of spurious signals.
( Receivers perform the inverse operation to transmitters, restoring the origi-
nal baseband signal with as little added noise and distortion as possible.
( The type of modulation and the point in the transmitter where modula-
tion is introduced determine the type of amplifiers needed.
( Transmitter efficiency is an important consideration because of energy,
size, and heat dissipation requirements.
( Most receivers use the superheterodyne principle, in which the signal is
moved to a fixed intermediate frequency.
( Receiver design emphasizes sensitivity, which is the ability to receive
weak signals, and selectivity, which is the ability to reject noise and inter-
ference.
( For a receiver, dynamic range is the ability to receive a weak signal in the
presence of stronger signals.
( Transceivers often use some of the same circuitry for both transmitting
and receiving.
( When a transmitter and receiver share the same antenna simultaneously,
they must use different frequencies, and a filter called a duplexer must be
used to separate the signals.
( Equation List
ƒ0 = Nƒref (9.1)
Po
η = (9.6)
Ps
ƒIF = ƒLO − ƒSIG (9.7)
ƒLO = ƒSIG + ƒIF (9.8)
ƒIF = ƒSIG − ƒLO (9.9)
ƒLO = ƒSIG − ƒIF (9.10)
 m ƒ
ƒ s =   ƒ LO ± IF (9.12)
n n
( Key Terms
driver amplifier immediately preceding the power amplifier stage in a
transmitter
dynamic range ratio, usually expressed in decibels, between the
strongest and weakest signals that can be present in a system
frequency multiplier circuit whose output frequency is an integer
multiple of its input frequency
frequency synthesizer device to produce programmable output
frequencies that are accurate and stable
image frequency reception of a spurious frequency by a superheterodyne
receiver, resulting from mixing of the unwanted signal with the local
oscillator signal to give the intermediate frequency
mixer nonlinear device designed to produce sum and difference
frequencies when provided with two input signals
selectivity ability of a receiver to discriminate against unwanted signals
and noise
sensitivity ability of a receiver to detect weak signals with a satisfactory
spurious responses in a receiver, reception of frequencies other than that
to which it is tuned
360 ! CHAPTER 9
spurious signals unwanted signals accidentally produced by a transmitter

superheterodyne principle use of a mixer and local-oscillator
combination to change the frequency of a signal
transceiver a combination transmitter and receiver
variable-frequency oscillator (VFO) an oscillator whose frequency can be
changed easily, usually by means of a variable capacitor or inductor
voltage-controlled oscillator (VCO) oscillator whose frequency can be
changed by adjusting a control voltage
( Questions
1. What is frequency agility, and under what circumstances is it desirable?
2. Why is it necessary to suppress the emission of harmonics and other
spurious signals by a transmitter?
3. What is meant by the overall efficiency of a transmitter?
4. Why is audio compression used with many transmitters?
5. What is meant by the duty cycle of a transmitter?
6. Is it possible to use Class C amplifiers to amplify an AM signal? Explain.
7. What advantages does the use of a frequency synthesizer for the oscilla-
tor stage of a transmitter have over:
(a) a crystal-controlled oscillator?
(b) an LC oscillator?
8. How is the power output of an SSB transmitter specified?
9. Why do FM transmitters use low-level modulation followed by Class C
amplification?
10. How can the deviation of an FM signal be increased?
11. Explain the operation of a QPSK modulator.
12. Distinguish between low- and high-side injection of the local-oscillator
signal in a receiver.
13. Explain how image-frequency signals are received in a superheterodyne
receiver. How may these signals be rejected?
14. What are the main characteristics of a well-designed receiver RF ampli-
fier stage?
15. Why is the stability of the local oscillator important?
16. In addition to images, what spurious responses are possible with super-
heterodyne receivers, and how are they caused?
17. State what is meant by the shape factor of a filter, and explain why a
small value for the shape factor is better for the IF filter of a receiver.
18. What advantage is gained by using double conversion in a receiver?
19. Describe two modern types of FM detectors, and explain how each
works. Use diagrams to help with the explanations.
20. Explain the operation of a QPSK demodulator.
( Problems
1. A frequency synthesizer of the type shown in Figure 9.2 is to generate
signals at 5-kHz intervals from 10 to 12 MHz using a reference crystal
oscillator operating at 10 MHz.
(a) Find a suitable value for Q.
(b) What is the range over which N has to be varied?
2. A frequency synthesizer of the type shown in Figure 9.2 has a crystal
oscillator operating at 5 MHz, Q = 100, and N ranges from 1000 to 1200.
Find the range of frequencies that can be generated and the minimum
amount by which the output frequency can be varied.
3. Draw a block diagram for an SSBSC generator that will generate an USB sig-
nal with a (suppressed) carrier frequency of 9 MHz. Show the carrier-
oscillator frequency and the filter center frequency and bandwidth.
4. Using the QPSK modulator shown in Figure 9.7, draw a series of vector
diagrams showing the output for the following data input: 00100111.
5. Frequencies of 120 MHz and 100 MHz are applied to a mixer. Calculate
the frequencies that appear at the output if the mixer is:
(a) a balanced mixer.
(b) a square-law mixer.
6. The frequency of a transmitter is guaranteed accurate to ± 0.0005%.
What are the maximum and minimum frequencies at which it could
actually be transmitting if it is set to transmit on a nominal carrier fre-
quency of 472.05 MHz?
7. A transmitter has an output carrier power of 25 W. If spurious signals
must be at a level of −70 dBc or less, what is the maximum power any of
the spurious signals can have?
8. A transmitter is rated to supply 4 W of power to a 50-Ω load, while oper-
ating from a power supply that provides 13.8 V. The nominal supply
current is 1 A. Calculate the overall efficiency of this transmitter.
362 ! CHAPTER 9
9. A superheterodyne receiver is tuned to a frequency of 5 MHz when the

local-oscillator frequency is 6.65 MHz.
(a) What is the intermediate frequency?
(b) Which type of injection is in use?
10. One receiver has a sensitivity of 1 µV and another a sensitivity of 10 dBf
under the same measurement conditions. Both receivers have an input
impedance of 50 Ω. Which receiver is more sensitive?
11. A receiver has a sensitivity of 0.3 µV, and the same receiver can handle a
signal level of 75 mV without overloading. What is its AGC control
range in dB?
12. The receiver of problem 11 has a blocking dynamic range of 80 dB. If the
desired signal has a level of 10 µV, what is the maximum signal level that
can be tolerated within the receiver passband?
13. A receiver uses low-side injection for the local oscillator and an IF of
1750 kHz. If the local oscillator is operating at 15.750 MHz,
(a) to what frequency is the receiver tuned?
(b) what is the image frequency?
14. An FM broadcast receiver with high-side injection and an IF of
10.7 MHz is tuned to a station at 91.5 MHz.
(a) What is the local-oscillator frequency?
(b) What is the image frequency?
(c) Find four other spurious frequencies that could be picked up by this
receiver.
15. A receiver’s IF filter has a shape factor of 2.5 and a bandwidth, at the 6
dB down points, of 6 kHz. What is its bandwidth at 60 dB down?
16. What is the shape factor of the filter sketched in Figure 9.22?
FIGURE 9.22
17. The block diagram of Figure 9.23 shows a double conversion receiver. It
has two mixers, two local oscillators, and two intermediate frequencies.
The idea is to have a high first IF for image rejection and a low second IF
for gain and selectivity.
(a) Does the first mixer use low-side or high-side injection?
(b) What is the second IF frequency?
(c) Suppose that the input signal frequency were changed to 17.000 MHz.
What then would the frequencies of the two local oscillators be?
FIGURE 9.23
18. Suppose a receiver has the following gain structure (all stages operating
at maximum gain):
RF amp: 12 dB gain
Mixer: 1 dB loss
IF amp: 76 dB gain
Detector: 3 dB loss
Audio amp: 35 dB gain
(a) What is the total power gain of the receiver from antenna to
speaker?
(b) What would be the minimum signal required at the antenna in
order to get a power of 0.5 W into the speaker? Express your answer
in both watts and microvolts, assuming a 50-Ω input impedance.
364 ! CHAPTER 9
(c) Now suppose a 100 mV signal is applied to the antenna. Calculate

the output power. Is this a reasonable answer? Explain. What would
actually happen in a real receiver?
19. (a) Draw the block diagram for an FM broadcast receiver. It is to have
one RF stage and an IF frequency of 10.7 MHz. The local oscillator
will operate above the signal frequency. Indicate on the diagram
the frequency or frequencies at which each stage operates when the
receiver is receiving a station at 94.5 MHz.
(b) What is the image frequency of the receiver described above?
(c) Give two ways in which the image rejection of the receiver could be
improved.
20. An FM receiver is double conversion with a first IF of 10.7 MHz and a
second IF of 455 kHz. High-side injection is used in both mixers. The
first local oscillator is a VFO, while the second is crystal-controlled.
There is one RF stage, one stage of IF amplification at the first IF, and
three stages of combined IF amplification and limiting at the second IF.
A quadrature detector is used as the demodulator. The receiver is tuned
to a signal with a carrier frequency of 160 MHz. Draw a block diagram
for this receiver.
21. Consider the transceiver whose block diagram is in Figure 9.20. Let the
first I.F. be 16.9 MHz, with low-side injection of the first local-oscillator
signal.
(a) What frequencies would the synthesizer produce while transmitting
and receiving, respectively, on marine channel 14 (156.7 MHz)?
(b) Describe and explain the technique used for modulation in the
transmitter section.
(c) The transmitter section uses a maximum deviation of 5 kHz and op-
erates with modulating frequencies in the range of approximately
300 Hz to 3 kHz. Suggest an appropriate receiver bandwidth.
Cellular Radio 10
Objectives
( Describe the history of personal communication up to the beginning of
digital cellular radio.
( Explain the operation and limitations of CB radio and cordless telephones.
( Explain the operation of and perform relevant calculations for North
American analog cellular telephone systems.
( Explain the operation of and perform relevant calculations for North
American digital cellular telephone systems.
366 ! CHAPTER 10
' 10.1 Introduction

With this chapter we begin our discussion of several systems that are some-
times grouped together as personal communication systems (PCS). Like
many technical terms, this one has several meanings. Specifically, it is used
for a particular variant of cellular radio which will be described in the next
chapter. But more generally, it can be applied to any form of radio communi-
cation between individuals.
In this chapter, after some historical introduction, we look at the com-
mon North American cellular telephone system, known as the Advanced
Mobile Phone Service (AMPS). The network, the cell sites, and the portable
and mobile telephones are described.
AMPS is still the most common cellular radio technology in North Amer-
ica, but a digital variation has become increasingly popular. We look at this
system in this chapter as well. Newer Personal Communication Systems (PCS),
which use cellular techniques but at a different frequency range and with
an assortment of digital modulation schemes, are the subject of the next
chapter.
' 10.2 Historical Overview

The cellular radiotelephone system has its origins in much earlier systems.
There has long been a need for portable and mobile communication. Three
early concepts, two of which are still in wide use today, show aspects of what
is needed. A brief look at each will show why cellular radio was created.
Citizens’ Band This is probably the earliest true personal communication system. Intro-
Radio duced in the United States in the 1960s, citizens’ band (CB) radio enjoyed
great popularity in the 1970s, followed by an almost equally steep decline as
its limitations became better known.
CB radio was intended to do some of the same things that are envisaged
by more recent personal communication systems. The relatively low fre-
quency of 27 MHz made transceivers affordable when CB radio was intro-
duced, and the absence of any test for a license made it easy for anyone to get
involved. The transmitter power limit of four watts for full-carrier AM, or
twelve watts peak envelope power for SSB, is designed to reduce interference
by restricting the communication range to a few kilometers. The fact that
most CB operation is between mobile units with low antenna heights and no
repeaters also limits the effective range. The restricted range is necessary to
limit interference since there are only 40 channels. This should not be a
problem since CB radio is intended for local communication.
CELLULAR RADIO ! 367
We can estimate the range of CB communication by making some as-

sumptions based on typical equipment and propagation paths, using the
techniques introduced in Chapter 7.
EXAMPLE 10.1 Y
Two handheld CB transceivers are held 1 m above flat, level terrain. The
transmitter power output is 4 W and the receiver sensitivity is 0.5 µV into
50 Ω. The transmitting and receiving antennas are both loaded vertical
monopoles with a gain of 1 dBi. Determine whether the maximum commu-
nication range is limited by power or distance. Assume there is no interfer-
ence and that free-space attenuation applies.
SOLUTION
First we should recognize that our answer is likely to be optimistic. Usually
the terrain is not flat and there are reflections from buildings, vehicles, and
so forth. Our results might be fairly accurate for transmission over water.
First let us find the possible line-of-sight range, which is limited by the
distance to the radio horizon. In Chapter 7 we noted that
d = 17ht + 17hr (10.1)
where
d = maximum distance in km
ht = transmitting antenna height in m
hr = receiving antenna height in m
Applying this equation to our situation gives
d = 17ht + 17hr
= 17 × 1 + 17 × 1
= 8.2 km
This is the maximum distance regardless of power level. Now let us see
what free-space distance would be possible with the given power level,
antenna gains, and receiver sensitivity, ignoring the horizon. We’ll ignore
transmission line losses, which are probably negligible anyway since the
antenna is mounted directly on the transceiver.
First we need to convert the required voltage at the receiver to a power
level in dBm.
368 ! CHAPTER 10
V2
P =
R
(0.5 × 10−6 )2
=
50
= 5 fW
= −113 dBm
We also need to express the 4 W transmitter power in dBm:
PT = 4 W
= 36 dBm
The antenna gains increase the effective power by 2 dB. Our allowable path
loss is then
Lƒs = 36 dBm − (−113 dBm) + 2 dB = 151 dB
The path loss is given by
Lƒs = 32.44 + 20 log d + 20 log ƒ (10.2)
where
Lƒs = free-space loss in decibels
d = path length in km
Here we know Lƒs and ƒ and we need to calculate d. Rearrange Equation (10.2):
20 log d = Lƒs − 32.44 − 20 log ƒ
= 151 − 32.44 − 20 log 27
= 89.9
d = 31378 km
Therefore, this system is quite obviously limited by the distance to the

radio horizon (and possibly by interference from other transmitters nearer
to the receiver) and not by transmitter power. In this situation the power
level could easily be reduced considerably with no effect on communication
quality.
In a practical mobile situation the attenuation might actually be propor-
tional to the fourth power of distance and would vary greatly depending on
reflections. Still, range tends to be limited by the horizon, reflections and
shadows, and interference, rather than power level.
X
Use of half-duplex (push-to-talk) operation for CB radio means only

one channel is needed per conversation, and using AM (including its nar-
rower bandwidth variant, SSB) keeps the required bandwidth less than
would be needed for FM. The channels are spaced at 10-kHz intervals.
Selection among the 40 available channels is done by the operator, who
is supposed to listen and manually switch to a clear channel before transmit-
ting. Since anyone can listen in on any channel, there is no privacy.
The main disadvantages of CB radio are directly related to its simplicity
and informality. Lack of privacy and co-channel interference are major prob-
lems. So is the lack of any connection to the wireline phone system or any
access to repeaters for reliable, long-distance communication. The relatively
high power level needed for communication without repeaters causes porta-
ble transceivers to be large and heavy. See Figure 10.1 for a typical example.
The low frequency, which requires antennas to be large if they are to be effi-
cient, is also a problem for portable transceivers.
FIGURE 10.1
Handheld CB transceiver
(Courtesy of Tandy Corporation)
Some of these problems have been addressed with unlicensed FM trans-

ceivers currently being sold for the 46/49-MHz and 460-MHz bands. The
latter frequency range is called the Family Radio Service (FRS). These trans-
ceivers are more compact but they still suffer from the other limitations
mentioned earlier.
370 ! CHAPTER 10
Cordless Most cordless phones are intended as simple wireless extensions to ordinary
Telephones telephone service. For best results, a telephone, cordless or otherwise,
should operate in full-duplex mode; that is, it should be capable of transmit-
ting and receiving at the same time. Thus a cordless phone needs two radio
channels, separated widely enough in frequency to avoid interference be-
tween them. Early designs had only a single channel for each direction, so
the transmitter power levels and range had to be kept very small to minimize
interference. Nonetheless, consumers found that a telephone that could be
carried freely throughout a house and its grounds was very useful and cord-
less phones have been extremely popular. A typical modern example is
shown in Figure 10.2.
FIGURE 10.2
Cordless phone
(Courtesy of Tandy Corporation)
Most current designs use analog FM in either the 43–49-MHz or

900-MHz bands. Older phones use AM at about 1.7 MHz for the handset,
with FM at 49 MHz for the base unit. Some 900-MHz models use analog FM
while others use digital spread-spectrum techniques, and there is at least one
model (the Panasonic Gigarange™) that uses a 2.4-GHz spread-spectrum sig-
nal from base to handset, and 900 MHz from handset to base. Most cordless
phones can automatically scan 10 or 25 channels, choosing a clear channel

(assuming one exists) without any user involvement.
Table 10.1 shows the frequencies currently in use for 43–49-MHz cord-
less phones. Note that the base and portable frequencies must be widely
separated, since both are in use simultaneously for full-duplex communica-
tion. Both the phone and base units require duplexers to separate the trans-
mit and receive frequencies. Different manufacturers of 900-MHz phones
use different channel frequencies and spacings, but the base and handset
TABLE 10.1 Cordless Telephone Frequencies: 43–49 MHz band
Channel Base Handset

1 43.720 48.760
2 43.740 48.840
3 43.820 48.860
4 43.840 48.920
5 43.920 49.000
6 43.960 49.080
7 44.120 49.100
8 44.160 49.160
9 44.180 49.200
10 44.200 49.240
11 44.320 49.280
12 44.360 49.360
13 44.400 49.400
14 44.460 49.480
15 44.480 49.500
16 46.610 49.670
17 (B) 46.630 49.845
18 (C) 46.670 49.860
19 46.710 49.770
20 (D) 46.730 49.875
21 (A) 46.770 49.830
22 (F) 46.830 49.890
23 46.870 49.930
24 46.930 49.990
25 46.970 49.970
The letters A through F denote channels that are also used for baby monitors.
372 ! CHAPTER 10
frequencies are typically separated by about 20 MHz. The power level is de-
liberately set very low so that range is power-limited rather than extending
to the radio horizon. It is surprising how low the power level for a cordless
phone can be and still give reasonable results.
EXAMPLE 10.2 Y
A cordless phone operating at 49 MHz is to have a range of 50 m. Assuming
0 dBi gain for the antennas and the same receiver sensitivity as in Exam-
ple 10.1, what transmitter power is required?
SOLUTION
Obviously the distance to the horizon is not the limiting factor here. We can
use Equation (10.2) to calculate the loss for a path length of 50 m:
Lƒs = 32.44 + 20 log d + 20 log ƒ

= 32.44 + 20 log 0.05 + 20 log 49
= 40.2 dB
If the required signal strength at the receiver is −113 dBm as before, then the
transmitter power must be at least
Pt = −113 dBm + 40.2 dB = −72.8 dBm = 52.7 pW

X
In practice, the power levels are much higher to cope with fading due to
reflections and absorption. Cordless phones in the 46/49-MHz band are re-
stricted to an EIRP of about 30 µW, while 900 MHz digital phones can use
about 16 mW EIRP.
Cordless telephones share much of the simplicity of CB radio. There are
no license requirements, and there is no official coordination of frequencies.
Users, or in most cases the phones themselves, simply try to choose a chan-
nel that is not in use. The newer cordless phones use digital access codes to
prevent unauthorized persons from dialing the phone and possibly making
unauthorized toll calls, but it is still not possible to use two nearby phones
on the same channel at the same time. The use of FM does provide some pro-
tection from interference: due to the capture effect, the desired signal has
only to be a few decibels stronger than the interfering signal in order to re-
duce interference to a reasonable level.
Privacy is not quite as nonexistent as with CB radio, since the newer
phones automatically avoid occupied channels. This reduces accidental
privacy violations, but anyone who wants to eavesdrop on an analog phone
can do so by using a scanner, for instance. The digital phones offer much
better privacy.
At this time, listening to others’ cordless or cellular phone calls is illegal
in the U.S.A., though not in Canada. (In Canada, listening to these conversa-
tions is legal, but divulging what you heard to another person is not.)
Because of the limited number of channels, cordless phones rely on ex-
tremely low transmitter power (microwatts to a few milliwatts depending on
the band and the phone) to limit interference. Of course, this also limits
their range. These phones certainly provide access to the wireline phone net-
work, but in general, only from the customer’s own premises or very nearby.
The newer spread-spectrum phones do have more range—up to a kilometer
or so under ideal conditions.
Despite interference problems and severely limited range, cordless
phones have been and remain very popular with consumers. Various at-
tempts have been made, particularly in Europe, to devise systems called
telepoints that would enable users to take their cordless handsets to public
places like malls and office buildings and use them there. However, recent
developments in cellular and PCS systems have caused these ideas to lose fa-
vor. The cordless phone seems likely to remain popular in its current niche
as a low-cost, wireless extension phone. Its low power allows it to have long
battery life (weeks of standby, hours of talk time), and especially at 900 MHz,
antennas are reasonably small and unobtrusive. Comparing the cordless
phone shown in Figure 10.2 with the portable CB transceiver displayed ear-
lier shows that cordless phones are a step in the right direction in terms of
convenience.
Improved Mobile The familiar cellular radiotelephone system has its origins in much earlier
Telephone systems that used a few widely spaced repeaters. Wide coverage was obtained
Service (IMTS) by using powerful base-station transmitters with antennas mounted as high
as possible. The mobile transceivers likewise used relatively high power, on
the order of 30 watts. Very similar systems are still widely used in dispatch-
ing systems, such as those for taxicabs and ambulances, for example.
The most common type of mobile telephone, from its introduction
in the mid-1960s until the coming of cellular radio in the early 1980s (the
first commercial cellular system became operational in Chicago in 1983),
was known as the Improved Mobile Telephone Service (IMTS). IMTS is a
trunked system; that is, radio channels are assigned by the system to mobile
users as needed, rather than having one channel, or pair of channels, perma-
nently associated with each user. Narrowband FM technology is used. Two
frequency ranges, at about 150 and 450 MHz were used for IMTS, with an
earlier system called MTS operating at around 40 MHz. The three systems
374 ! CHAPTER 10
combined had only 33 available channels. A few IMTS systems are still in
use, mainly in remote locations.
IMTS is capable of assigning channels automatically, by the rather sim-
ple means of transmitting a tone from the base station on unoccupied chan-
nels. The receiver in the mobile unit scans channels until it detects the tone.
IMTS is capable of full-duplex operation using two channels per tele-
phone call. Direct dialing is also possible, so using a mobile phone is almost
as simple as using an ordinary telephone at home.
The main problem with IMTS and similar systems is that whatever band-
width is made available to a single repeater, is tied up for a radius of perhaps
50 km or even more, depending on the height of the antenna and the power
of the transmitter at the base station. Any attempt to reuse frequencies
within this radius is likely to result in harmful interference. Simple systems
like this also suffer from fading and interference near the edges of their cov-
erage areas. For instance, suppose two similar trunked systems with identical
repeaters are located 50 km apart. Then, at a location midway between the
two, a receiver would receive equally strong signals from each. Communica-
tion would be impossible if both repeaters used the same frequencies.
' 10.3 Introduction to the Advanced Mobile Phone

System (AMPS)
As the demand for mobile telephony grew, it became obvious that another
way had to be found to accommodate more users. More spectrum could be
found, but only by going up in frequency, since all the spectrum near the ex-
isting mobile telephone allocations was already occupied. The 800-MHz re-
gion was already assigned to UHF television broadcasting, but more
spectrum had been assigned to this service than was actually being used,
probably because of the popularity of cable television for specialty channels.
A band approximately 40-MHz wide (increased to about 50 MHz in 1986)
was assigned to the new system.
To make the system more efficient, cellular radio technology, based on
many repeaters with their range deliberately restricted, was introduced at
the same time. See Chapter 7 for a description of the cellular principle. Cel-
lular radio had to wait until enough computing power could be introduced
into the system, at reasonable cost, to allow the system to keep track of mo-
biles as they moved from one cell site to the next and to allow mobile phones
to change frequency and power level by remote control from the cell site.
When cellular telephony was introduced in North America, it was
decided to allow two different companies, called carriers, to operate in any
given region. One of these would be the local wireline telephone company
(telco). The other would be an independent company called a radio com-
mon carrier (RCC). Each carrier was assigned half the channels in each area
in an attempt to encourage competition. (It would actually be more efficient
in terms of spectrum usage to have only one provider: with the current sys-
tem it is possible for one provider’s channels to be fully loaded while the
other still has available frequencies.)
Cellular radio goes a long way toward relieving the congestion described
above by essentially reversing the conventional wisdom about radio systems
using repeaters. Instead of trying to achieve long range by using high power,
cellular repeaters are deliberately restricted in range by using low power. As dis-
cussed in Chapter 7, the high path loss associated with mobile propagation ac-
tually makes it easier to reduce interference in a cellular system. A reasonable
elevation for the base-station antennas is still required to minimize radio shad-
ows (behind buildings, for instance). Similarly, the mobile radios use low power
(no more than 4 W ERP for mobiles and 600 mW or less for portable phones),
and in fact, the mobile transmitter power is automatically limited by the sys-
tem to the minimum required for reliable communication.
Instead of one repeater, there are many, located in a grid pattern like that
shown in Figure 10.3. Each repeater is responsible for coverage in a small
cell. As shown, the cells are hexagons, but of course in a real situation the an-
tenna patterns will not achieve this precision—the cells are more likely to be
approximately circular, with some overlap. All the cell sites in a region are
connected by copper cable, fiber optics, or microwave link to a central office
called a mobile switching center (MSC) or mobile telephone switching
office (MTSO), and the MSCs are themselves interconnected so that the sys-
tem can keep track of its mobile phones. The cellular system is connected at
a point of presence to the wireline network, so that cellular customers can
speak to wireline customers.
Note that there is no provision for direct mobile-to-mobile radio com-
munication. Even if two cell phones are in the same room, a call from one to
FIGURE 10.3
Cell boundaries (seven-
cell repeating pattern)
376 ! CHAPTER 10
the other must go through a cell site and an MSC. Provided both portable
phones are connected to the same network (A or B), there would be no need
to go through the PSTN.
Since each transmitter operates at low power, it is possible to reuse fre-
quencies over a relatively short distance. As we saw in Chapter 7, typical mo-
bile propagation conditions allow for a repeating pattern of either seven or
twelve cells; the available bandwidth is divided among these cells. The fre-
quencies can then be reused in the next twelve or seven cells, with the lower
number possible when directional antennas are used with three sectors us-
ing different frequencies per cell.
Cellular Carriers In the current North American system, there are 395 duplex voice channels,
and Frequencies each consisting of one channel in each direction for each of the two carriers.
There are also 21 control channels for each carrier used to set up calls and ad-
minister the system. AMPS uses narrowband analog FM, with a maximum
frequency deviation of 12 kHz and a channel spacing of 30 kHz.
Table 10.2 shows how these channels are divided between the two car-
riers: A represents the non-wireline carrier and B represents the wireline
carrier. Note that the frequencies assigned to each carrier are not all contigu-
ous because of the extra frequencies added to the system in 1986. Note
also the rather large separation (45 MHz) between base and mobile trans-
mit frequencies. This allows for simple duplexers to separate transmit and
TABLE 10.2 North American Cellular Radio Frequencies
Base Frequencies Mobile Frequencies Type of

(forward channels) (reverse channels) Channel Carrier
869.040–879.360 824.040–834.360 Voice A
879.390–879.990 834.390–834.990 Control A
880.020–880.620 835.020–835.620 Control B
880.650–889.980 835.650–844.980 Voice B
890.010–891.480 845.010–846.480 Voice A*
891.510–893.970 846.510–848.970 Voice B*
Table denotes transmit carrier frequencies. Mobile transmits 45 MHz below base.
A = non-wireline carrier (RCC) B = wireline carrier (telco)
* = frequencies added in 1986
receive signals in the phones. The base transmits to the mobile on a for-
ward channel, while transmissions from mobile to base use a reverse
channel.
An individual cell doesn’t use all these channels, of course. Each cell has
only one-seventh or one-twelfth of the total number of channels assigned
to a carrier, depending on the system. Contiguous frequencies are not used
in order to reduce interference. With a seven-cell repeating pattern, trans-
mitters in the same cell are generally separated by about seven channels or
210 kHz. Each cell in a seven-cell pattern also has three of the 21 control
channels.
To further reduce receiver selectivity requirements, adjacent channels
are not used in adjoining cells. Therefore, transmitters in adjacent cells are
separated in frequency by at least 60 kHz.
Channel The control channels are used, among other things, to allocate voice chan-
Allocation nels to phones. When a user dials a phone number on a mobile phone and
presses the Send button, the phone scans all the control channel frequencies
to find the strongest. This control channel should be associated with the
closest cell site. The cell phone transmits on its corresponding control chan-
nel, and once the call has been set up, the cell site assigns it a clear voice
channel, assuming one is available.
While the conversation continues, the cell sites adjacent to the one in
use monitor the signal strength from the mobile. When the strength is
greater in one of the adjacent cells, the system transfers the call to that cell.
This procedure is called a handoff. Handoffs, of course, require a change in
frequency for the mobile phone, under control of the system.
A similar procedure takes place for incoming calls. The mobile periodi-
cally identifies itself to the system whenever it is turned on, so the system
usually has a good idea of its location. Paging signals are sent out on control
channels and the mobile responds, enabling the system to locate it more pre-
cisely. In the early days of cell phones it often took some time, a minute or
more, to find a mobile, but improved communication within the system has
reduced this time to a few seconds in most cases. The phone is instructed to
ring, and once it is answered, the system assigns it a voice channel. After that
the system follows the phone as it moves from one cell to the next, as
explained earlier.
Frequency Reuse The reason for the complexity of the cellular system is, of course, frequency
reuse. Once a mobile has moved out of a cell, the frequency pair it occupied
is available for another conversation. By making cells smaller, frequencies
378 ! CHAPTER 10
can be reused at shorter distances. There is no theoretical limit to this, but

there are practical limits. As cells become smaller, more cell sites are needed
and handoffs occur more frequently, requiring more computing power and
faster response both at the system level and in the individual mobile phone.
Once the radius drops below about 0.5 km the handoffs occur so frequently
that it is difficult to cope with a mobile moving at high speed. The flexibility
of cell sizes allows for larger cells in less-developed areas and smaller cells in
areas of the greatest traffic.
EXAMPLE 10.3 Y
A vehicle travels through a cellular system at 100 kilometers per hour. Ap-
proximately how often will handoffs occur if the cell radius is:
(a) 10 km?
(b) 500 m?
SOLUTION
The reason for the word “approximately” in the problem statement is that
we are not sure how the road crosses the cell boundaries. Let us assume for
simplicity that the vehicle drives along a road leading directly from one cell
site to the next. Thus, the vehicle will change cells each time it travels a dis-
tance equal to the diameter of a cell (twice the radius).
First convert the speed into meters per second.
100 km/hr × 1000 m/km

v =
3600 s/hr
= 27.8 m/s
Now we can find the time between handoffs.

(a) Let the diameter be d = 20 km
d
t =
v
20 × 103 m
=
27.8 m/s
= 719 s
= 12 min
(b) This time, d = 1 km
d
t =
v
1 × 103 m
=
27.8 m/s
= 36 s
X
' 10.4 AMPS Control System

In this section we study in more detail the process by which the AMPS cellu-
lar system keeps track of phones and calls. We need to look at the functions
of the control channels and also at the control information that is sent over
the voice channels.
An effective control system has to do several things. It needs to keep
track of mobile phones, knowing which ones are turned on and ready to re-
ceive a call and where they are. It needs to keep track of telephone numbers
for authentication and billing, and it should have some way to detect and
prevent fraudulent use. It must be able to set up calls, both from and to mo-
bile phones and transfer those calls from cell to cell as required. It would be
best if all this were transparent to the user, who should only have to dial the
phone number or answer the phone, just as with a wireline phone at home.
A more advanced system might also be able to send faxes and e-mail, surf the
internet, and so forth, but let’s look at the basics first!
First we need to understand the functions of the voice and control chan-
nels. You might assume that the voice channels are for talking and the con-
trol channels are for control signals, and that is mostly correct. However,
there is a problem: cell phones contain only one receiver and one transmit-
ter, so they can’t receive both a voice channel and a control channel at the
same time. Therefore, any control messages that have to be sent during a
conversation must use the voice channel. Some of this is done using
in-channel, out-of-band signaling (consisting of tones above the voice fre-
quency range), and the rest is done with blank-and-burst signaling, during
which the voice signal is muted for a short time (100 ms) while data is sent
over the voice channel.
Digital signals on the control channel and those sent during blank-
and-burst signaling on the voice channel are sent in a relatively simple way.
They use FSK with 8 kHz deviation (16 kHz total frequency shift) and a chan-
nel bit rate of 10 kb/s. The data is transmitted using Manchester code. In
380 ! CHAPTER 10
order to reduce the likelihood of errors, the control channel sends each mes-
sage five times and also uses Hamming error-correction codes. This increases
the robustness of the control system but reduces the actual data throughput
to 1200 b/s. There is no encryption in the AMPS system: all the data coding
information is publicly available. This is a serious oversight that has been
remedied in the newer PCS systems to be described in the next chapter.
Mobile and Base Each mobile unit has two unique numbers. The mobile identification
Identification number (MIN) is stored in the number assignment module (NAM) in the
phone. The MIN is simply the 10-digit phone number for the mobile phone
(area code plus 7-digit local number), translated according to a simple algo-
rithm into a 34-bit binary number. The NAM has to be programmable, since
it may be necessary to assign a different telephone number to the phone, but
it is not supposed to be changeable by the user. In most cases, however, it can
be changed from the keypad if the user knows the right procedure. (Check
the internet—it took the author less than ten minutes to find the procedure
for his own cell phone.)
Usually a cell phone is registered on either the A or B system and has one
MIN. It can operate on the other system as a roamer, if necessary and if there
is an agreement between the two systems to allow it. It is also possible for a
phone to have two MINs so that it can be used on both A and B systems with-
out roaming. In that case the user of the phone has two phone numbers (and
two bills to pay).
The other identification number is an electronic serial number (ESN),
which is a unique 32-bit number assigned to the phone at the factory. It is
not supposed to be changeable without rendering the phone inoperable, but
in practice it is often stored in an EPROM (erasable programmable read-only
memory chip) that can be reprogrammed or replaced by persons with the
right equipment and knowledge. The combination of the MIN and the ESN
enables the system to ensure proper billing and to check for fraudulent use
(for instance, if a registered MIN appears with the wrong ESN the system will
not allow the call to go through).
The mobile phone also has a number called the station class mark
(SCM), which identifies its maximum transmitter power level. There
are three power classes corresponding to phones permanently installed in a
vehicle, transportable “bag phones,” and handheld phones. The maximum
power levels, specified as ERP (effective radiated power with respect to a
half-wave dipole) are as follows:
Class I (mobile): +6 dBW (4 W)

Class II (transportable): +2 dBW (1.6 W)
Class III (portable): −2 dBW (600 mW)
Mobile transmitter power is controlled by the land station in 4 dB incre-

ments, with the lowest power level being −22 dBW (6.3 mW) ERP. The idea is
to reduce interference by using as little power as possible. Mobile and trans-
portable phones thus have better performance than portable phones only
when propagation conditions are bad enough, or cells large enough, that the
system needs to increase mobile power past the maximum for a portable
phone. Using a portable phone inside a vehicle attenuates the signal consid-
erably, so communication from a portable phone can sometimes be estab-
lished in marginal areas by simply getting out of a car.
The cellular system has an identifying number called the system identifi-
cation number (SID). This enables the mobile phone to determine whether it
is communicating with its home system or roaming. (Using a “foreign” system
usually costs more and the user may disable this ability if desired.) In addition
each cell site has a digital color code (DCC). When the mobile detects a
change in DCC without a change in frequency, it is an indication that co-chan-
nel interference is being received from another base station.
Turning on When a cell phone is turned on, it identifies itself to the network. First it
a Phone scans all the control channels for its designated system (A or B) and finds the
strongest. It looks for the SID from the system to determine whether or not
it is roaming. If it does not receive this information within three seconds, it
tries the next strongest control channel. After receiving the system informa-
tion, the mobile tunes to the strongest paging channel. Paging channels are
control channels that carry information about calls that the system is trying
to place to mobiles. If someone is calling the mobile, its number will be
transmitted by the paging channel.
The control channel constantly updates the status of its associated
reverse control channel (from mobile to system). Only the system transmits
on the forward channel, but any mobile can transmit on the reverse chan-
nel. The system tells the mobiles when this channel is busy to reduce the
chance of a collision, which occurs when two or more mobiles try to use
the control channel at the same time. After checking that the reverse chan-
nel is free, the newly activated phone transmits its ESN and MIN to the land
station so that the system knows the phone is ready for calls and in which
cell the phone is located. If the mobile loses the signal and reacquires it or
detects that it has moved to a different cell, it identifies itself again. In addi-
tion, the system may periodically poll its mobiles to see which are still
active.
While turned on but otherwise idle, the mobile phone continues to peri-
odically (at least once every 46.3 ms) check the control channel signal from
the cell site. It has to verify that a signal is still available, that it is from the
same system, and that there are no calls for the mobile phone.
382 ! CHAPTER 10
Originating a Call When the user of a mobile phone keys in a phone number and presses Send,
the mobile unit transmits an origination message on the reverse control
channel (after first checking that this channel is available). This message in-
cludes the mobile unit’s MIN and ESN and the number it is calling. The cell
site passes the information on to the mobile switching center for processing.
Once authorization is complete, the cell site sends a message to the mobile
on the forward control channel, telling it which voice channel to use for the
call. It also sends the digital color code which identifies the cell site, and a
Control Mobile Attenuation Code (CMAC), which sets the power level to
be used. This power level can be changed by the land station as needed dur-
ing the call by means of a control message on the forward voice channel.
Now both stations switch from the control channel to a voice channel,
but the audio is still muted on the phone. The cell site sends a control mes-
sage on the forward voice channel confirming the channel. It then sends a
supervisory audio tone (SAT) on the voice channel to the mobile phone.
This is a continuous sine wave, with a frequency above the voice band. There
are three possible frequencies: 5970, 6000, and 6030 Hz. The mobile relays
the tone back to the cell site. Reception of this tone by the cell site confirms
that the correct cell site and mobile are connected. The mobile sends a con-
firmation message on the reverse voice channel. After this handshaking, the
call can begin.
During the call the SAT continues (it is filtered out before the audio
reaches the speaker in the phone, of course). Reception of the wrong fre-
quency tone by the base station indicates an interfering signal and interrup-
tion of the tone indicates that the connection has been lost, perhaps due
to severe fading. If the tone is not resumed within five seconds, the call is
terminated.
A 10-kHz signaling tone (ST) may also be transmitted on the voice chan-
nel during a call. It is used to signal handoffs to another cell and the termina-
tion of the call.
Receiving a Call An incoming call is routed by the network to the cell where the mobile last
identified itself. (If it has not identified itself to the network, it is assumed to
be turned off and a recorded message to that effect is given to the caller.) The
land station sends the MIN on the paging channel along with the voice
channel number and power level to use. The mobile confirms this message
and sends its ESN on the reverse control channel to be matched by the net-
work with the MIN. This is to avoid fraudulent use. The base sends its infor-
mation again on the forward voice channel along with the digital color code
information, then the mobile confirms the information on the reverse voice
channel. After this handshaking, the supervisory audio tone is transmitted
on the voice channel and the conversation can begin.
Handoffs The network monitors the received power from the mobile at adjacent cell
sites during a call. When it detects that its strength is greater at an adjacent
cell site than at the site with which it is communicating, it orders a handoff
from one cell to the next. This always involves a change in channel, since to
avoid co-channel interference the same channels are never used in adjacent
cells. The order to do this is sent by the first cell site to the mobile on the for-
ward voice channel using blank-and-burst signaling. The resulting 100 ms
interruption in the conversation is barely perceptible. The voice channel
must be used, because during a conversation the mobile is not monitoring
any of the control channels. The mobile is given the new channel number,
new attenuation code, and new SAT frequency. After confirmation on the
reverse voice channel, the mobile switches to the new channel, which
connects to the new cell site, and the conversation continues. There will
probably be an audible disturbance while this occurs.
' 10.5 Security and Privacy

The AMPS system is not very private. Voices are transmitted using ordinary
FM and conversations can be picked up with any FM receiver that will tune
to the correct frequency. Base stations often repeat mobile transmissions, so
quite often both sides of the conversation can be picked up with one re-
ceiver, just as with a cordless phone but from much greater distances (typi-
cally a few kilometers). The change in channels as a mobile is handed off
does make it hard to follow conversations when the cell phone user is talk-
ing from a moving vehicle.
In 1988, in an attempt to increase cell phone privacy, the United States
government banned the import or sale of scanners or other receivers that
can tune to cellular frequencies. However, these are still legal in many coun-
tries (including Canada); there are millions of old scanners around and fre-
quency converters are easy to build. It should therefore be assumed that
AMPS voice transmissions are public. The transmission of confidential infor-
mation, such as credit card numbers for instance, is not advisable with ana-
log cell phones.
Stolen cell phones work only until the owner has the service cancelled.
(There is a code to lock the phone, but most people don’t bother, or they
leave the password at the factory setting, such as “1234.” Generally they do
this because the password is supposed to be changeable only by the dealer,
though in fact it can often be done from the keypad if one knows the correct
method.) Even if the phone number is changed by a knowledgeable thief,
the Electronic Serial Number will give away the fact that the phone is “hot.”
However, it is not impossible, at least on some phones, to change the ESN. It
384 ! CHAPTER 10
is also possible to acquire valid pairs of MIN and ESN numbers by monitor-
ing the reverse control channels. There is no encryption and the exact speci-
fications for the data fields are publicly available. It is just a matter of
acquiring the hardware and software to decode a 10 kb/s FSK data stream—
not a very formidable task. Once the numbers are available it is possible to
“clone” a cell phone to emulate a valid phone. Calls made on the cloned
phone are billed to the unfortunate legitimate subscriber.
Service providers do have some protection. For instance, if the network
detects the “same” phone trying to make two calls at once or two calls in
quick succession from widely separated locations, it will flag the occurrence
and someone will investigate. As the networks become larger and better inte-
grated, this type of fraud becomes a little more difficult.
Another fraud is to use a cloned or stolen phone on another network as a
roamer. If the foreign network is not capable of checking the phone’s home
network in real time, it may accept the call. This is becoming less likely as
networks become better connected with each other. In the meantime, net-
works are becoming less trusting (especially in the United States) and less
likely to allow roaming without identification.
' 10.6 Cellular Telephone Specifications and Operation

In this section we look at the requirements for the mobile or portable phone
itself and consider some examples of phone construction. See Figure 10.4 for
a block diagram of a typical analog cell phone. Because the system is full-
duplex, the transmitter and receiver must operate simultaneously with a sin-
gle antenna. A duplexer is used to separate the two signals. The wide 45-MHz
frequency separation between transmit and receive frequencies makes this
relatively easy. The constant frequency separation also simplifies frequency
synthesizer design.
Microprocessor control is necessary to allow the phone to switch chan-
nels and power levels by remote control from the base station. The processor
and its associated memory are also useful for timing calls, storing passwords
to unlock the phone, storing lists of frequently-called numbers, and so on.
Some cell phones can enter a sleep state between calls to conserve battery
life; they must emerge from this state to check the control channel at least
once in each 46.3 ms time period.
Transmitter Power In the previous section we noted that cell phones come in three station
and Frequency classes. This term refers to the maximum power level produced. The actual
transmitted power level is adjusted in 4 dB steps by signals from the cell site.
The mobile transmitter must transmit at within 3 dB of the correct power
FIGURE 10.4 Block diagram of analog cell phone
level within 2 ms of turning on and must reduce its output to −60 dBm ERP
or less within 2 ms of being turned off. The transmitted frequency must be
within 1 kHz of the specified channel frequency.
The power levels for mobile, transportable, and portable phones are
shown in Table 10.3. The abbreviation MAC refers to the mobile attenuation
TABLE 10.3 Power Levels for Mobile Phones (EIRP in dBW)
MAC Class I Class II Class III

000 +6 +2 −2
001 +2 +2 −2
010 −2 −2 −2
011 −6 −6 −6
100 −10 −10 −10
101 −14 −14 −14
110 −18 −18 −18
111 −22 −22 −22
There is a range of 28 dB between maximum and minimum mobile

power levels with a Class I phone.
386 ! CHAPTER 10
code, which is transmitted from the base station to adjust the power of the
mobile according to propagation conditions. Because FM and FSK are used,
there is no need for linearity in the transmitter power amplifier, and Class C
operation can be used for greater efficiency.
Transmitter As mentioned earlier, voice transmission uses FM with a maximum devia-

Modulation tion of 12 kHz each way from the carrier frequency. Data transmission uses
FSK with 8-kHz deviation each way.
Companding with a ratio of 2:1 is used in voice transmission. That is, in
the transmitter, a 2 dB change in audio level from the microphone causes
only a 1 dB change in modulation level. The process is reversed in the
receiver. The result is an improvement in signal-to-noise ratio for low-level
audio signals.
As with almost all FM systems, pre-emphasis is used in the transmitter
and de-emphasis in the receiver. This means, the higher audio signals are
given more gain in the transmitter and correspondingly less gain in the
receiver. In the cell phone system, all frequencies above 300 Hz are boosted
at the transmitter, with a slope of 6 dB per octave. Figure 10.5 shows the pre-
emphasis curve with the corresponding de-emphasis curve for the receiver.
FIGURE 10.5
Cellular radio
pre-emphasis and
de-emphasis
Mobile and Since the transmitted power of cellular phones is specified in terms of ERP,
Portable Antennas the use of more efficient antennas allows transmitter power to be reduced.
This is especially important in the case of portable phones, since lower trans-
mitter power leads to longer battery life. On the other hand, more efficient
antennas tend to be larger.
Most portable cell phones use a quarter-wave monopole antenna. At
800 MHz, the length of this antenna is about 9.5 cm. The options are wider
for mobile antennas. Many of these use a quarter-wave and a half-wave
section, separated by an impedance-transforming coil. See Figure 10.6 for
examples of typical portable and mobile antennas.
FIGURE 10.6
Portable and
mobile antennas
' 10.7 Cell-Site Equipment

The radio transmitting equipment at the cell site operates at considerably
higher power than do the mobile phones, but this power is shared among all
the channels that are used at the site. Similarly, there must be receivers for
each voice and control channel in use at the site, as well as extra receivers for
monitoring the signal strength of mobiles in adjacent cells. Consequently,
the cell site equipment is much more complex, bulky, and expensive than
the individual cell phones. In addition, as already noted, cell sites often
need directional antennas to facilitate the division of each cell into sectors.
Transmit and receive antennas may be separate or combined at the cell site.
Often two receive antennas and one transmit antenna are used per cell, or
388 ! CHAPTER 10
per sector, in a sectorized configuration. This allows for space diversity in

the receiver; to counteract the effects of fading, the receiver monitors the
signal from both antennas and chooses the stronger signal. The effects of
fading are likely to be greater for one antenna than the other at any given
moment. Figure 10.7 is a block diagram of the equipment in a typical cellular
base station.
FIGURE 10.7 Cellular radio base station
The combination of the mobile cellular phone and the cell-site radio
equipment is known as the air interface. There is much more to cellular tele-
phony than radio, however. The network must be organized and adminis-
tered as a whole. This administration includes keeping track of phone
locations, billing, setting up and handing off calls, and so on. The substan-
tial computing resources required to do all this are, at least in part, responsi-
ble for the delay in the introduction of cellular telephony for many years
after the idea was first proposed.
Figure 10.8 shows a typical cellular telephone system. Each cell has sev-
eral radio transceivers (one per channel); usually one wideband power am-
plifier is used to provide the transmit power for all channels in a site (or
sector, for sectorized systems). The site’s radio equipment is operated by
a base station controller (BSC). The base station controller takes care of
the air interface: assigning channels and power levels, transmitting signal-
ing tones, and so on. The mobile switching centers (MSCs), also called mo-
bile telephone switching offices (MTSOs), route calls along a private copper,
fiber optic, or microwave network operated by the cellular service provider.
FIGURE 10.8
Cellular radio
system
390 ! CHAPTER 10
Their action is also required in authorizing calls, billing, initiating handoffs,

and so on. Sometimes the BSC and MSC are combined. Associated with
the MSCs are data banks where the locations of local and roaming mobiles
are stored.
At certain points in the system, the cellular network is connected to the
public switched telephone network (PSTN). These gateways allow calls to be
made between landline and cellular phones, and between cell phones using
different service providers. The cellular system communicates with the
PSTN using Signaling System Seven, which was described in Chapter 5.
Traffic and The optimum size of a cell depends on the amount of traffic. Ideally, most of
Cell Splitting the available radio channels should be in use at peak periods, but situations
where all channels are in use should be rare. If all channels in a cell are busy,
it is impossible for anyone to place a call to or from that cell. The user has
to hang up and try again later. This situation is called call blocking and is
obviously undesirable. It causes revenue loss, and if it is frequent, unhappy
customers may switch to the competing system. (This is always a possibility
with the North American AMPS system since there are two competing
systems in each area.) Call blocking takes place on the wireline network as
well. For instance, long-distance trunks are sometimes unavailable during
peak calling periods. This means that customers are used to blocking and
will put up with a small percentage, perhaps one or two percent, of calls
being blocked.
A more unpleasant situation occurs when a mobile phone moves into a
cell that has all its channels busy. The attempt by the system to hand off the
call to the new cell is frustrated by the lack of free channels, and the call
must be terminated, or dropped. Dropped calls are very inconvenient and
sometimes embarrassing, as the effect is the same as if one of the parties had
hung up on the other. Call dropping is very rare on the wireline system, so
customers are not as tolerant of this problem.
Since call blocking also occurs on the wireline network, which has been
in operation for about a century, you might guess that someone has already
studied the problem. You’d be correct: A. K. Erlang, a Swedish engineer, stud-
ied the problem using statistical analysis early in the twentieth century. He
found, not surprisingly, that the more channels there were, the smaller the
possibility of blocking for a given amount of traffic. Perhaps less obviously,
he found that with more channels, the amount of possible traffic per chan-
nel increases for a given blocking probability. This phenomenon is called
trunking gain, and it is the reason a two-provider system is theoretically less
efficient than one using a single provider.
Trunking gain can perhaps be better understood by looking at an every-
day situation: customers lining up to use tellers at a bank. Suppose there are
two tellers and a separate line for each. Further suppose that the lines are as-
signed to customers on the basis of the type of accounts they have. Those
with checking accounts use the first line, those with savings accounts use
the second line.
Now suppose I arrive at the bank. My account is a checking account, but
there are several people in line at that window. There is no line at the savings
window, but I can’t use that one. I am blocked and decide to try again later. Of
course, after several frustrating attempts, I may notice that the competing
bank across the street has shorter lines and change service providers. Now
my bank changes its policy. There is only one line, and anyone can use the
next available teller. The next time I arrive at the bank I have a much lower
probability of being blocked. A similar logic applies to many situations: it is
always more efficient to combine channels, and the gains are greater with
more channels.
Phone traffic is defined in erlangs (E). One erlang is equivalent to one
continuous phone conversation. Thus if 1000 customers use the phone
ten percent of the time each, they generate 100 E of traffic on average. Math-
ematically,
T = NP (10.3)
where
T = traffic in erlangs
N = number of customers
P = probability that a given customer is using the phone
We see immediately that traffic analysis does not coordinate channels

directly with the number of customers. Real customers may use the phone
continuously for an hour, then not at all for the rest of the day. If enough of
them use the phone at the same time (afternoon rush hour, for example, is a
peak in cell phone usage in cities), the peak traffic will be much greater than
the average. Some cell phone owners use the phone continually for business
during the working day; others use it mainly for emergencies and generate
very little traffic. Also, the usage patterns may vary, in response to changing
rates, for instance (lower rates generate more traffic). Evening and weekend
use tends to be light for cell phones, just as it is for wireline long distance,
and both wireline and cellular providers commonly provide monetary in-
centives to customers to shift some of their usage to those periods.
The most obvious way to avoid call blocking and call dropping is, of
course, to provide more channels. However, the number of channels for the
system is fixed, and so is the number per cell once the repeating pattern, usu-
ally seven or twelve cells, has been established. In a 12-cell repeating pat-
tern, each provider has 395/12 ≅ 33 voice channel pairs per cell. According to
392 ! CHAPTER 10
Erlang’s theory, this number of channels will accommodate about 23 E of

traffic for one percent blocking.
The amount of traffic can be increased at the expense of a larger blocking
probability. For instance, with 33 channels, a traffic level of 24.6 E can be ac-
commodated with a two percent blocking probability. It might seem at first
glance that a 7-cell repeating pattern can allow more traffic, but this is illu-
sory. The cells are each divided into three sectors, using different channels.
Therefore, each sector has only 395/21 ≅ 19 duplex voice channels. Using
the 7-cell pattern saves money by reducing the number of cell sites needed,
not by increasing the number of channels.
Table 10.4 shows traffic levels in erlangs for 19 and 33 channels, with
various blocking probabilities. It also shows traffic levels for larger numbers
of channels. We’ll use these with digital cell phones shortly.
TABLE 10.4 Cellphone Traffic in Erlangs per Cell or Sector
Blocking Probability
Number of
Channels 1% 2% 5%
19 11.2 12.3 14.3
33 22.9 24.6 27.7
55 42.4 44.9 49.5
97 81.2 85.1 92.2
EXAMPLE 10.4 Y
A cellular telephone system uses a 12-cell repeating pattern. There are
120 cells in the system and 20,000 subscribers. Each subscriber uses the
phone on average 30 minutes per day, but on average 10 of those minutes are
used during the peak hour.
Calculate:
(a) the average and peak traffic in erlangs for the whole system
(b) the average and peak traffic in erlangs for one cell, assuming callers are
evenly distributed over the system
(c) the approximate average call-blocking probability
(d) the approximate call-blocking probability during the peak hour
SOLUTION
(a) The average traffic is
0.5
T = 20,000 ×
24
= 416 E
The peak traffic is
10
T = 20,000 ×
60
= 3333 E
(b) The average traffic per cell is
416
t =
120
= 3.47 E
The peak traffic per cell is
3333
t =
120
= 27.8 E
(c) Use the line from Table 10.4 corresponding to 33 channels, since this is
the number available in a 12-cell repeating system. For average traffic at
3.47 E, the blocking probability is much less than 1% (in fact it is negligi-
ble, since the average number of calls is much less than the number of
channels). At peak periods, however, the blocking probability increases
to just over 5%.
X
The other way to increase capacity is to increase the number of cells. The
number of channels per cell remains the same as before, but since each cell
covers a smaller area, with less potential traffic, the probabilities of call
blocking and call dropping are reduced. The downside of this, of course, is
that the expense of the system increases with the number of cell sites, and
more frequent handoffs occur, increasing the system overhead.
This reduction of cell size to increase traffic is called cell-splitting.
Cell-splitting allows the network to begin with large cells throughout, with
394 ! CHAPTER 10
the cell size decreasing in high-traffic areas as the traffic increases. Cellular
telephone infrastructure is expensive, but it does not all have to be built
at once.
Cell-splitting allows spectrum space to be used more efficiently, but it is
not particularly cost-effective in terms of equipment. Doubling the number
of cells doubles the system capacity with the same bandwidth allocation, but
it also doubles the number of cell sites, roughly doubling the system cost.
Microcells, Cell-splitting, as described in the preceding paragraphs, can be used to in-

Picocells and crease the capacity of a cellular system. At a certain point diminishing
Repeaters returns set in, however. Cell sites are expensive, and the increase in capacity
does not justify the increase in cost for very small cells. Another problem is
that real estate costs are highest in the areas where demand is greatest, and
the use of small cells means less choice in cell-site location and thus higher
costs for access to the sites. Finally, there is increased load on the switching
system due to the increase in the number of handoffs with small cells.
In high-demand areas, microcells are often used to help relieve the con-
gestion at relatively low cost. A microcell site is a very small unit that can
mount on a streetlight pole. The microcell antenna is deliberately mounted
lower than the tops of nearby buildings to limit its range. A typical microcell
covers about 500 meters of a busy street, but has very little coverage on side
streets. See Figure 10.9 for a typical unit.
FIGURE 10.9
Microcell site
(Courtesy of Bell
Mobility)
Because microcells have such small, narrow patterns, it is difficult to

obtain general coverage this way. Consequently the original larger cells
(macrocells) are left in place, so that calls can be handed off between micro-
cells and conventional cells as required. The microcells must use different
frequencies than the overarching conventional cells, of course; this is ac-
complished by assigning to the microcells some of the channels that were
formerly used by the macrocells.
A microcell is often under the control of a conventional cell site, with
which it usually communicates by microwave radio. The microcell may
itself be divided into several zones. Figure 10.10 illustrates how microcells
and conventional cells (sometimes called macrocells) can work together.
FIGURE 10.10
Overlay of microcells
and macrocells
In order to save costs, many microcell sites are not true transceivers but
are only amplifiers and frequency translators. Figure 10.11 on page 396
shows the idea. The main cell site up-converts the whole transmitted spec-
trum to microwave frequencies. The transmitter at the microcell site simply
down-converts the block of frequencies to the cellular-radio band and am-
plifies it. No modulation or channel switching is required at the microcell
site. Similarly, the microcell’s receiver consists of a low-noise amplifier that
amplifies the entire frequency range in use at that site plus an up-converter
to convert the signal to microwave frequencies. All demodulation is handled
at the main cell site.
Sometimes cellular radio signals are too weak for reliable use indoors.
This is especially true in well-shielded areas like underground concourses.
When reliable indoor reception is needed, sometimes very small cells called
picocells are used. These are more common with the PCS systems to be
described in the next chapter but are sometimes used with AMPS.
Indoor picocells can use the same frequencies as the outdoor cells in the
same area if the attenuation of the structure is sufficient. This is the case in
underground malls, for instance.
396 ! CHAPTER 10
FIGURE 10.11
Microcell block
diagram
Sometimes the problem is not excessive traffic, but a “hole” in the sys-
tem coverage caused by propagation difficulties, such as a tall building or
hill that casts a radio shadow. Obviously cell sites should be chosen to mini-
mize these problems, but it is not always possible to eliminate them com-
pletely. In that case a repeater can be used, as shown in Figure 10.12. The
repeater simply amplifies signals from the cell site and from the mobiles.
FIGURE 10.12
Cellular repeater
It can be connected to the main site by a microwave link, but often the
repeater simply receives and transmits at the same frequencies, avoiding
feedback by careful location of directional transmit and receive antennas.
' 10.8 Fax and Data Communication Using

Cellular Phones
As can be seen from the foregoing, AMPS is a circuit-switched analog system
designed for voice communication, as was the wireline system in its original
form. Consequently, the most straightforward way to send data, as with the
conventional wireline system, is to use modems at each end of the link. It is
also possible to send packet-switched data using AMPS, as will be described
shortly.
Cellular Modems The main differences between wireline phone service and analog cellular
phones, for modem use, is that cell phone connections tend to be noisier
and are subject to interruption during handoffs and fading. These interrup-
tions, though brief in human terms (usually on the order of 100 ms to about
2 s), result in the loss of a considerable amount of data and possibly in a
dropped connection. Consequently, the error-correction schemes on
modems for cell phone use should be more robust than is necessary for wire-
line operation. Of course, this greater robustness results in slower data trans-
mission. In addition, the modem must be set up not to require a dial tone
before dialing. The dialing connections must be made separately to the
phone: the situation is more complex than just plugging in a phone jack, as
for a landline modem.
Many (but not all) modem cards for notebook and laptop computers
will work with cell phones. Similarly, many, but not all, cell phones can be
used with modems. A proprietary cable is required to connect modem
to phone. It is also possible to find cell phone cards which plug into a note-
book computer and enable it to send data via cellular radio.
Cellular modems are advertised as having speeds of up to 28.8 kb/s but
actual speeds are usually 9600 b/s or less. Performance is improved by oper-
ating from a stationary vehicle, as this eliminates handoffs and reduces fad-
ing.
An error-correcting protocol called MNP10 is usually used with cellu-
lar connections. It must be used at both ends of the connection. MNP10 in-
corporates some special cellular enhancements. For instance, rather than
398 ! CHAPTER 10
trying to connect at the maximum speed and reducing speed if necessary, an

MNP10 modem starts at 1200 b/s and gradually increases the speed. Multiple
attempts are made to connect, and the packet size starts small and increases
provided error rates remain low.
Another protocol, developed by AT&T, is called Enhanced throughput
cellular (ETC). It has the advantage of being able to work with a conventional
wireline modem at the other end of the connection while retaining at least
some of its benefits.
In order to allow communication between cellular modems and conven-
tional wireline modems which may lack these more robust protocols, some
cellular providers maintain a modem pool. This allows for conversion be-
tween cellular and wireline standards. Generally, the cellular user accesses
this service by dialing a special prefix before dialing the telephone number
of the landline modem. See Figure 10.13 for the idea.
FIGURE 10.13
Data transmission
by cellular radio
Facsimile transmission is also possible with cell phones. A fax modem

and notebook computer can be used, or a conventional fax machine can be
used with a special adapter. The adapter allows the cell phone to simulate a
conventional two-wire telephone line with dial tone. Fax performance is
much better from a stationary vehicle, since no special protocol is used.
Cellular Digital The previous section describes how data can be sent over a cellular voice
Packet Data channel in a manner similar to that used with landline telephony. That pro-
(CDPD) cedure is relatively simple and requires no prior arrangement with the cellu-
lar service provider, but it tends to be expensive, as full airtime costs, plus
long-distance costs if applicable, must be paid for the entire time the call is
connected.
Another way exists to send data over the AMPS cellular radio system. The
Cellular Digital Packet Data (CDPD) system uses packet-switched data and
tends to be less expensive than using a cellular modem, especially when data
needs to be transmitted in short bursts. On the other hand, a separate ac-
count is required, and the cellular system has to be specially configured to
use CDPD.
The principle behind CDPD is that at any given moment there are usu-
ally some voice channels in an AMPS system that are not in use. The CDPD
system monitors the voice channels, using those that are idle to transmit
data. When traffic is detected on the voice channel, the data transmissions
cease within 40 ms. Since this is less than the setup time for a voice call, the
voice customer is not affected. Users, once registered with the CDPD system,
can transmit data as required without maintaining a continuous connection
and tying up an expensive pair of voice channels.
The bit rate in the RF channel for CDPD is 19.2 kb/s, achieved using
Gaussian minimum-shift keying (GMSK), a form of FSK. When overhead
is taken into account, the maximum data rate is comparable to that obtained
with a 14.4 kb/s modem—slow by current wireline standards, but not too
bad for wireless. When the network is busy, the throughput is lower, as pack-
ets are stored and forwarded when a channel becomes available.
' 10.9 Digital Cellular Systems

Until recently, digital voice communication used more bandwidth than
analog. The conventional wisdom was that digital was preferred where
bandwidth was not an issue, because of its flexibility and immunity to the
accumulation of noise, but that analog was better when bandwidth was
constrained. For instance, you will recall that wireline telephony uses
64 kb/s for each one-way voice channel. In order to transmit this data rate in
a 30-kHz channel, an elaborate modulation scheme would be needed. This
would not be robust enough for the mobile radio environment with its noise
and fading. Consequently, all first-generation cellular radio systems use
analog modulation schemes. FM needs more bandwidth than AM and its
variants, but it was found that FM’s resistance to noise and interference
more than made up for the additional required bandwidth.
400 ! CHAPTER 10
Recent advances in data compression and voice coding, as discussed in

Chapter 3, have reversed the conventional wisdom about bandwidth. It is
now possible to transmit a digitized voice signal in less bandwidth than is re-
quired for an analog FM signal. This removes the last big obstacle to digital
communication. Digital systems still tend to be more complex than analog,
but large-scale integrated circuits have made it possible to build complex
systems at low cost and in small packages. In fact, it appears likely that the
analog cellular radio system just described will be the last of the great analog
communication systems. Soon there will be no reason to use analog except
for compatibility with legacy systems and perhaps for very low-cost commu-
nication devices.
Analog cellular radio can be seen as the first generation of wireless com-
munication. Digital systems, which we begin to examine here and continue
in the next chapter, are the second. The third generation will undoubtedly
be digital. Its exact makeup is being debated at this time. A synopsis of the
progress thus far can be found in Chapter 14.
Advantages of The main incentive for converting cellular radio to a digital system was, as
Digital Cellular suggested earlier, to reduce the bandwidth requirements, allowing more
Radio voice channels in a given spectrum allotment. Other reasons also exist. Digi-
tal systems have more inherent privacy than analog, being harder to decode
with common equipment. They also lend themselves to encryption, if
required. Note that analog AMPS already uses digital signaling data. The fact
that it is not encrypted is due to oversight: the system designers underesti-
mated the ingenuity of hackers.
Digital communication systems can use error correction to make them
less susceptible to noise and signal dropouts. They lend themselves to time-
and code-division multiplexing schemes, which can be more flexible than
the frequency-division multiplexing used in analog systems. Digital signals
are easier to switch: in fact, most of the switching of analog telephone
signals, including AMPS cellular telephony, is done digitally after analog-
to-digital conversion.
The gradual conversion of the world’s cellular radio systems to digital
form has been accomplished differently in various parts of the world. In
North America, which is the focus of this book, there existed one analog sys-
tem with a very large installed base. Most of this infrastructure was not yet
fully paid for, so operators were understandably reluctant to replace it. The
requirement was for a digital system that would allow as much as possible of
the analog equipment to be retained.
There were also millions of analog phones in consumers’ hands. Many of
these also were not yet paid for, especially since operators had given many
analog phones to customers in exchange for service contracts. Perhaps you
still have one of these (the author does). Making the analog phones obsolete
would probably obligate the operators to provide new digital phones to their
customers at no charge.
For these reasons, the system that evolved in North America uses the
same radio frequencies, power levels, and channel bandwidths as AMPS.
Some of the existing analog channels were converted to digital, leaving
some analog channels in place in every cell for coexistence with analog
equipment. The reduction of bandwidth requirements for digital communi-
cation allowed those analog channels that were converted to digital to com-
bine three communication channels on one RF channel using TDMA, as
described in the following section. The new system is so similar to AMPS that
sometimes it is referred to as digital AMPS (D-AMPS).
The situation in Europe was completely different. Almost every country
seemed to have a different analog cell phone scheme. Some had more than
one! There were even several different frequency bands in use in different
countries. The European Community made the radical decision to scrap the
entire analog infrastructure and begin again with a common digital system.
This system is called GSM (Global System for Mobile). It is not used in the cel-
lular radio bands in North America but it is one of three systems in use for
North American PCS, so GSM will be described in the next chapter.
Other countries varied in the type and extent of analog cell phone pene-
tration. The result is a worldwide mixture of incompatible analog and digital
formats, which will probably continue at least until the third generation of
wireless telephony.
Conversion of The compatibility requirements outlined in the previous section con-

AMPS to TDMA strained the development of a digital cellular radio system for North Amer-
ica. It was decided to combine three digital voice channels into one 30-kHz
radio channel using TDMA. This is known as a full-rate TDMA system. The
first such system went into operation in 1990. The specifications also
encompass a half-rate system with six voice channels in one 30-kHz slot to
be implemented at a later date when vocoder technology has improved.
The digital system would seem to be able to carry three times as much
traffic as the analog system, but, due to trunking gain, the actual increase for
a given level of blocking is greater. In the last two lines of Table 10.4, the traf-
fic for various blocking probabilities is shown for both 7-cell sectorized and
12-cell repeating systems. The numbers of voice channels are calculated on
the basis that one analog channel, for backward compatibility, is available in
each cell or sector. Note that the traffic capacity is more than tripled.
At first, the AMPS control channels were left alone, and only the voice
channels were digitized. This digital specification is known as IS-54B. A later
402 ! CHAPTER 10
modification (1991), called IS-136, added high-speed digital control chan-

nels for better security and additional features. IS-136 is used both in the
800-MHz cellular radio band and the 1900-MHz PCS band.
TDMA Voice Figure 10.14 shows how the RF voice channel is divided in the TDMA
Channel system. There are 25 frames per second so each frame is 1/25 s = 40 ms in
length. Each frame has 1944 bits so the total bit rate for the RF signal is 1944
× 25 = 48.6 kb/s. Phase-shift keying with four levels (π/4 QPSK) is used, so
there are two bits per symbol and the baud rate is 24.3 kbaud. Note that this
is a data rate of 48.6/30 = 1.6 b/s per hertz of bandwidth. This is quite conser-
vative, necessarily so because of the radio environment, which is subject to
noise, interference, and very deep fading.
FIGURE 10.14
TDMA frame
Each frame has six time slots lasting 40 ms/6 = 6.67 ms and containing
1944/6 = 324 bits each. For full-rate TDMA, each voice signal is assigned to
two time slots as shown. Six voice signals, occupying one slot each, can be
accommodated with half-rate TDMA. For the full-rate system, speech data
corresponding to 40 ms of real time is transmitted in 2 × 6.67 ms = 13.3 ms.
As with analog AMPS, the TDMA system uses separate RF channels for trans-
mit and receive.
Speech encoding (see Chapter 3) is used to limit the bit rate to approxi-
mately 8 kb/s for each speech channel for the full-rate system. The full-rate
system allocates two noncontiguous time slots to each voice channel: slots 1
and 4 for the first, 2 and 5 for the second, and 3 and 6 for the third. In addi-
tion, the bits corresponding to each 20 ms of speech are divided among
two time slots. Interleaving the data bits in this way reduces the effect of
burst errors.
Overhead reduces the number of data bits available per time slot to 260.
The data rate available for each voice channel is 260 bits/20 ms = 13 kb/s.
The voice is actually encoded at 7.95 kb/s and the remaining bits are used for
error correction. The half-rate system will use 4 kb/s for voice coding.
The frames as just shown are similar for both forward (base to mobile)
and reverse (mobile to base) channels, but the composition of the time slots
differs. The frames are synchronized for the forward and reverse channels,
but the timing is offset so that a frame starts 90 bits (1.85 ms) earlier at the
mobile. A mobile transmits during two of the six time slots and receives on a
different two slots. The remaining two time slots are idle: the phone may use
these to check the signal strength in adjacent cells to assist in initiating a
handoff. This technique allows the digital cell phone to have only one trans-
mitter and one receiver, just as for AMPS. In fact the RF situation is a little
simpler, since there is no need for the mobile to transmit and receive simul-
taneously. (Since digital cell phones have to work with the analog AMPS
system as well, this unfortunately does not really simplify the design, as a
duplexer is still needed for analog operation.)
Each slot contains 324 bits for both forward and reverse channels. The
allocation of these bits is different for the forward and reverse links, how-
ever. See Figure 10.15 for an illustration of the differences.
In particular, the mobile needs time to turn its transmitter on for each
transmit time slot, since to avoid interference it must be off when the mobile
is not scheduled to transmit. Six bit periods (123 µs) are allocated to this. The
base station transmitter is on all the time, since it uses all six time slots to
FIGURE 10.15 TDMA voice time slots

404 ! CHAPTER 10
transmit to three mobiles on the same RF channel. The mobile also waits for
a guard time (another six bits) to pass before transmitting. This is necessary
because different amounts of propagation delay could cause mobile trans-
missions (on the same RF channel but in different time slots) to overlap
when received at the base station.
EXAMPLE 10.5 Y
Calculate the maximum distance between base and mobile that can be
accommodated with a guard time of 123 µs.
SOLUTION
The signal from base to mobile is delayed by the propagation time between
the two. This causes the mobile’s synchronization to be off by that much
time, and it will be late in starting its transmission. In addition, the signal is
further delayed by the propagation time from mobile to base. Therefore, the
123 µs allowance must include the round-trip propagation time. Since radio
waves propagate at the speed of light, the maximum total round-trip
distance is:
d = ct
= 300 × 106 m/s × 123 × 10−6 s
= 36.9 km
The maximum one-way distance is half that, or about 18.5 km. For even
larger cells, the mobile can instruct the base to advance its transmission by
up to 30 bit periods (617 µs).
X
In addition to voice samples and their associated error correction, the

digital traffic channels contain synchronizing, equalizer training, and con-
trol information.
The Coded Digital Verification Color Code (CDVCC) provides essentially
the same information as the Supervisory Audio Tone (SAT) in AMPS. The
Slow Associated Control Channel (SACCH) provides for control signal ex-
changes during calls and essentially replaces the blank-and-burst signaling
in AMPS, though there are provisions to “steal” voice data bits for additional
control information as required. These stolen bits form the Fast Associated
Control Channel (FACCH), which is used for urgent information such as
handoff commands. The coded digital locator (CDL) field tells the mobile
where to find digital control channels, if available.
TDMA Control As mentioned, there are two “flavors” of TDMA cellular radio in use. The ear-
Channels lier specification, called IS-54B, uses the same control channels and formats
as AMPS. These are called Analog Control Channels (ACCH) because of their
association with the analog system, but as noted earlier, they are actually
digital, using FSK and a channel data rate of 10 kb/s.
The IS-136 specification incorporates separate control channels for the
digital system. These are called Digital Control Channels (DCCH) to distin-
guish them from the older type. Digital control channels consist of pairs of
slots on the same RF channels that are used for voice. The DCCH can be as-
signed to any RF channel; it does not have to be one of the 21 control chan-
nels used in the analog system. As with the voice channels, separate forward
and reverse channels are needed. Normally there is one DCCH pair per cell,
or per sector in a sectorized system. See Figure 10.16.
FIGURE 10.16
TDMA digital
control channel
The total bit rate for a DCCH is one-third of the RF channel bit rate, or
44.6/3 = 14.9 kb/s, compared with 10 kb/s for an ACCH. This extra capacity
makes the digital control channels useful for many added features, such as
call display and short text messages. As with the analog system, digital con-
trol channels are used in setting up calls. They cannot be used during a call,
since the single receiver in the mobile unit is otherwise occupied.
Privacy and Privacy is considerably improved in digital cellular radio compared to the
Security in Digital analog system. Ordinary analog scanners can make no sense of the digitized
Cellular Radio voice signal. Even decoding it from digital to analog is not straightforward,
due to the need for a vocoder. However, obviously vocoders are present in all
digital cell phones, so a modified cell phone could do the job.
There is some encryption of the authorization information in the TDMA
system, enough to make cell phone cloning and impersonation difficult. In
general the level of security of the TDMA system is considerably better than
with AMPS and is probably adequate for general use.
406 ! CHAPTER 10
Dual-Mode One of the most important features of the TDMA digital cellular radio
Systems and system is its backward compatibility with AMPS. Cell-site radio equipment
Phones does not need to be replaced, and provided that all cells keep some analog
channels, neither do analog cell phones. Cell sites can incorporate digital
channels as needed to cope with increased traffic.
Since not all cells are expected to have digital channels in the near
future, it is necessary for digital cell phones to work with the analog system
as well. While analog-only phones will continue to work throughout the sys-
tem, digital-only phones would work only in major metropolitan areas.
Therefore, all current TDMA digital cell phones are dual-mode: they attempt
to make a digital connection first, then if that fails, revert to analog.
One difference in the RF component of TDMA cell phones is the addi-
tion of a new power class, Class IV. This is the same as the analog Class III, ex-
cept for the addition of three new power levels at the bottom of the range.
These levels of −26, −30, and −34 dBW ERP allow for better operation with
microcells. In addition, lower power levels, coupled with the fact that the
transmitter operates only one-third of the time, help to improve battery life
in digital mode.
Figure 10.17 is a block diagram for a typical dual-mode TDMA cell
phone. Note that a duplexer is required for analog operation. The transmit-
ter power amplifier is linear, because QPSK requires this. Of course, a linear
amplifier is also satisfactory for FM and FSK. The fact that the channel band-
width and frequencies are the same for analog and digital systems simplifies
the RF design. For instance, only one receiver IF filter is needed.
When we look at PCS systems in the next chapter, we’ll see that the
TDMA system just described is one of three systems used for PCS. This allows
for the possibility of dual-mode, dual-band phones incorporating analog
AMPS, as well as both digital cellular and TDMA PCS modes.
Data Paradoxically, connecting data equipment like modems and fax machines to
Communication digital cell phones can be more complicated than with the ordinary analog
with Digital system. Since voice is encoded using a vocoder, it is not satisfactory simply
Cellular Systems to insert modem tones instead of an analog voice signal. The vocoder is opti-
mized to code voice and will make a mess of any other kind of input. In fact,
even DTMF tones from the phone’s keypad have to bypass the vocoder and
be transmitted digitally in a time slot of their own.
Circuit-switched data communication can be accomplished with the
digital system by inputting the data directly to the voice time slots without
using the vocoder. The data rate is limited to 9600 b/s to allow for additional
error correction and still fit within the 13 kb/s allocated for voice data. Of
course, the system has to be told about this, so that the data can be output
properly at the other end.
FIGURE 10.17 Block diagram of dual-mode cell phone
In practice, since all 800-MHz digital cell phones and cell sites are also
capable of analog operation, the usual way of sending data is to use an
analog channel and a cellular modem as previously described. Similarly,
packet-switched data can be sent via the analog system using CDPD as al-
ready described. Short data messages can be sent using the digital control
channel.
408 ! CHAPTER 10
( Personal communication systems work best when they are portable, con-
nected to the public switched telephone network, and widely available at
reasonable cost.
( Early attempts at personal communication include CB radio, IMTS radio
telephones, and cordless phones.
( Cellular radio systems are much more efficient users of spectrum than
communication systems using a single repeater, but they are also much
more complex and require computing power in the mobile phone.
( AMPS, the North American analog cell phone system, uses FM modula-
tion with 30-kHz channels in the 800-MHz region of the spectrum.
( Cell-splitting, including the use of microcells and picocells, can greatly
increase the capacity of a cellular system.
( Cell sizes have a lower limit due to the number and frequency of handoffs
required for quickly moving vehicles.
( The North American digital cell phone standard builds on AMPS by split-
ting one analog channel into three digital channels using TDMA.
( Both circuit-switched and packet-switched data can be carried on the
cellular radio system using special adapters. Performance is generally not
as good as with the wireline system.
( Equation List
d = 17ht + 17hr (10.1)
Lƒs = 32.44 + 20 log d + 20 log ƒ (10.2)
T = NP (10.3)
( Key Terms
Advanced Mobile Phone Service (AMPS) North American
first-generation cellular radio standard using analog FM
air interface in wireless communication, the radio equipment and the
propagation path
base station controller (BSC) in cellular and PCS systems, the electronics
that control base station transmitters and receivers
blank-and-burst signaling in cellular communication, interrupting the

voice channel to send control information
call blocking failure to connect a telephone call because of lack of
system capacity
cellular digital packet data (CDPD) method of transmitting data on
AMPS cellular telephone voice channels that are temporarily unused
control mobile attenuation code (CMAC) information sent by the base
station in a cellular radio system to set the power level of the mobile
transmitter
digital color code (DCC) signal transmitted by a cell site to identify that
site to the mobile user
dropped call a telephone connection that is unintentionally terminated
while in progress
electronic serial number (ESN) number assigned to a cell phone by the
manufacturer as a security feature. It is transmitted with the phone’s
telephone number to authorize a call.
fast associated control channel (FACCH) in a digital cellular system or
PCS, control information that is transmitted by “stealing” bits that are
normally used for voice information
forward channel communication from a cell site or repeater to the
mobile unit
full-duplex two-way communication in which both terminals can
transmit simultaneously
half-duplex two-way communication in which only one station can
transmit at a time
handoff transfer of a call in progress from one cell site to another
Improved Mobile Telephone Service (IMTS) a mobile telephone service,
now obsolescent, using trunked channels but not cellular in nature
microcell in cellular radio, a small cell designed to cover a high-traffic area
mobile identification number (MIN) number that identifies a mobile
phone in a cellular system; the mobile telephone number
mobile switching center (MSC) switching facility connecting cellular
telephone base stations to each other and to the public switched
telephone network
mobile telephone switching office (MTSO) see mobile switching
center (MSC)
410 ! CHAPTER 10
number assignment module (NAM) in a cellular phone, a memory

location that stores the telephone number(s) to be used on the system
personal communication system (PCS) a cellular telephone system
designed mainly for use with portable (hand-carried) telephones
picocells very small cells in a cellular radio system
radio common carrier (RCC) a company that acts as a carrier of
radiotelephone signals
reverse channel communication channel from mobile station to base
station
roamer a cellular customer using a network other than the subscriber’s
local cellular network
slow associated control channel (SACCH) in a digital cellular system or
PCS, control information that is transmitted along with the voice
station class mark (SCM) code which describes the maximum power
output of a cellular phone
supervisory audio tone (SAT) in the AMPS system, a sine wave above the
voice frequency range, transmitted on the voice channel along with
the voice, used by the base station to detect loss of signal
system identification number (SID) in the AMPS system, a number
transmitted by the base station to identify the system operator
telepoint a very small cell used with some cordless phones to allow their
use in public areas
( Questions
1. What usually limits the distance for CB communication?
2. What usually limits the distance from handset to base for cordless
phones?
3. What advantage does the use of a higher frequency have for hand-held
transceivers?
4. Why do cordless phones and cellular phones need two RF channels for
one conversation?
5. What is a telepoint system?
6. Explain the concept of frequency reuse with reference to the IMTS and
AMPS mobile telephone systems.
7. How do cellular telephone systems connect to the PSTN?
8. Explain why FM is better than AM for cellular radio, in spite of its wider
bandwidth.
9. Why is it more efficient to use a single provider for cellular and PCS
phone service? Why are multiple providers used in spite of this fact?
10. What happens when a mobile phone moves out of a cell? What is the
name for this process?
11. What factors limit the maximum and minimum sizes of cells?
12. What are the functions of the control channels in the AMPS system?
13. What type of modulation and data rate are used on AMPS control
channels?
14. How does the AMPS system attempt to prevent fraudulent billing of
calls to the wrong customer? How effective are these techniques?
15. What are the MIN and ESN for a cell phone and what are the differences
between them?
16. What are the differences between mobile phones, transportable
phones, and portable phones from the point of view of the cellular
system?
17. What are the SID and DCC and what is the difference between them?
18. When, why, and how are control signals sent over the voice channel?
19. How does the AMPS system keep track of which mobiles are turned on
and where they are?
20. List and explain the steps in receiving a call to a mobile.
21. What is SAT and what is its function? Is it analogous to anything in a
conventional local loop?
22. What is blank-and-burst signaling and why is it necessary?
23. What is a cloned phone? How does the AMPS system try to detect these
phones?
24. What is the function of the duplexer in a cell phone?
25. Why is it unnecessary for cell phones and cellular base station transmit-
ters to use linear RF power amplifiers in their transmitters? What is the
advantage of using nonlinear amplifiers?
26. What is the range of transmit power adjustment (in dB) for each class of
cell phone?
27. Under what circumstances would a Class I or Class II phone operate
with higher transmit power than a Class III phone?
28. What is meant by pre-emphasis and how does it improve the signal-
to-noise ratio of an FM cellular radio system?
412 ! CHAPTER 10
29. What advantage does a typical mobile antenna have over a typical por-
table cell phone antenna? Given this advantage, why is the same type of
antenna not used on portable phones?
30. What is the advantage of using two dedicated receive antennas per cell
site or sector over a combined transmit/receive antenna?
31. What is the function of the base station controller?
32. What is meant by the term air interface in a cellular radio system?
33. What is the difference between call blocking and call dropping? Which
is more objectionable?
34. Explain the concept of trunking gain.
35. What are microcells? How do they differ from macrocells and picocells?
36. What limits the size of a microcell?
37. Where are repeaters used in cellular systems?
38. What are the differences between cellular modems and ordinary wire-
line modems?
39. How can a fax machine be made to work with a cell phone?
40. What is the difference between circuit-switched and packet-switched
data communication?
41. How does the CDPD system transmit packet-switched data over the nor-
mally circuit-switched AMPS cell phone system?
42. How does the North American digital cellular system increase system
capacity?
43. What is meant by TDMA and FDMA? Show that the North American
digital cellular system actually uses both TDMA and FDMA.
44. How do privacy and security in the North American digital cell phone
system compare with these aspects of AMPS?
45. Digital wireline telephone systems use 64 kb/s per voice channel, yet
digital cellular systems achieve almost equivalent voice quality using 8
kb/s. Explain briefly how this is possible.
46. State and compare the modulation schemes used by the control chan-
nel in analog AMPS and the digital control channel used in the TDMA
digital cell phone system.
47. TDMA cell phones need a duplexer even though they transmit and re-
ceive in different time slots. Why?
48. What is the problem with connecting a cellular modem to a TDMA
phone?
( Problems
1. Look at Example 10.1.
(a) Calculate the minimum transmitter power that would give line-
of-sight communication at 27 MHz for the maximum distance set
by the radio horizon. Assume free-space propagation.
(b) Repeat the calculation for free-space propagation at 800 MHz.
Account, physically, for the fact that more power is required at the
higher frequency.
2. Repeat the calculation for Problem 1 assuming typical mobile propaga-
tion conditions and a frequency of 800 MHz. What can you conclude
about the feasibility of portable communication without repeaters over
this distance in an urban environment?
3. Now suppose we have a cellular system. The mobile is as described in
Problem 2 but the base station has an antenna with a gain of 6 dBi
located 50 m above average terrain. How much power is needed now? Is
this reasonable?
4. Amateur radio clubs often operate repeaters, usually in frequency bands
near 144 and 430 MHz. They do not have the same power restrictions as
CB radio. Calculate the minimum repeater height and transmitted EIRP
for a repeater that must communicate with mobile stations over a 50 km
radius, maintaining a minimum field strength of 50 mV/m at the
mobile receiver, which has an antenna height of 1.5 m. Assume free-
space propagation conditions.
5. Suppose, that in an attempt to increase range, cordless telephone manu-
facturers double the power output of their transmitters. Assume, that in
the cluttered in-house environment, power density is inversely propor-
tional to the fourth power of distance.
(a) Suppose the usable range, limited by signal strength without inter-
ference, was 50 m before the power increase. What is it after the
increase?
(b) What effect will the power increase have on interference from other
portable telephones?
(i) operating at the old (low) power level?
(ii) operating at the new (high) power level?
(c) What can you conclude from your answer about the effect of in-
creasing power in an environment where transmission range is lim-
ited by
(i) signal-to-noise ratio?
(ii) signal-to-interference ratio?
414 ! CHAPTER 10
6. (a) Calculate the length of a quarter-wave vertical antenna for each of

the following services: CB radio at 27 MHz; cordless phone at 49
MHz; FRS at 460 MHz; cellular radio at 800 MHz; and PCS at
1800 MHz.
(b) What can you conclude about the ease of designing portable trans-
ceivers for each of these services?
7. Suppose we are designing picocells for an indoor cellular system in
an underground mall. If we assume people using the system are walking
at a speed of 5 m/s or less, what is the smallest cell radius we could have
if we wanted to have handoffs occur no more often than once every
45 seconds?
8. Suppose we are communicating on a cell phone using a digital modem
at 9600 b/s. How many bits will be lost during
(a) blank-and-burst signaling?
(b) handoff, if it results in a 0.5 s loss of signal?
9. A cellular system is designed to operate reliably with traffic of 2000 E. If
each customer on average uses the phone for three minutes during the
busiest hour, how many customers can be accommodated, assuming an
even distribution of customers?
10. If the number of channels were doubled in the previous question,
would the maximum number of customers be double, more than dou-
ble, or less than double that found above? Explain.
11. Suppose that a system’s customers use their cell phones for an average
of five minutes each during the busiest hour. The system has 10,000
customers, approximately evenly distributed over 100 square km. Half
of the customers have dual-mode digital phones. The other half have
analog-only phones. We require a maximum of 2% call blocking for
those with digital phones at rush hour.
(a) How many cells are required and with what radius, if we choose an an-
alog system with a 12-cell repeating pattern? Assume hexagonal cells.
(b) Suppose we try to improve service by changing to a digital system,
leaving one analog channel per cell. What happens (qualitatively)
to the blocking rate for customers with each type of phone?
(c) How could we improve the situation in (b), as it applies to analog-
only customers?
12. Suppose two cell phones are in the same room. Draw a diagram showing
the route signals take if the cell phones are:
(a) operating on the same system (A or B)
(b) operating on different systems (one each on A and B)
13. Prepare a chart showing and explaining the steps in placing a call from a
mobile.
14. Prepare a chart showing and explaining the steps in placing a call to a
mobile.
15. (a) Assume that the range of a mobile phone is limited by its transmit-
ter power. If a portable phone (Class III) can communicate over a
distance of 5 km, what would be the maximum range for a perma-
nently installed car phone (Class I) under the same circumstances.
Assume free-space attenuation.
(b) Repeat part (a), but assume that attenuation is proportional to the
cube of distance. This is more typical of a mobile environment.
(c) Examine the assumption made above. What other factors could
limit communication range?
16. (a) By how much would a 2400-Hz baseband signal be boosted by the
transmitter pre-emphasis circuit in a cell phone, compared with a
frequency of 300 Hz?
(b) Why is the boost due to pre-emphasis unlikely to cause over-
modulation of the transmitter?
(c) Compare the gain in the receiver for signals of frequencies 2400 Hz
and 300 Hz after demodulation.
17. Suppose that the same type of vocoder used in full-rate TDMA cellular
radio were to be used for wireline telephony. How many voice channels
could be carried on a T-1 line?
18. A base and mobile are separated by 5 km. What is the propagation time
for a signal traveling between them?
19. Express the minimum power output for a digital Class IV cell phone in
mW. How does it compare with CB radio and cordless phones?
20. (a) Review the discussion of mobile propagation in Chapter 7. Find the
path loss in a typical urban environment in the cell phone fre-
quency range if the base station antenna has a height of 20 m and
the distance is 5 km.
(b) Find the power delivered to the base station receiver under the same
circumstances if the base station antenna has a gain of 6 dBi and the
feedline to the receiver has a loss of 2 dB.
(c) Suppose the phone described in part (a) is used inside a vehicle,
which causes a loss of 20 dB. What would be the effect on the
answers for parts (a) and (b)?
Personal Communication
Systems
11
Objectives
( Compare cellular radio with PCS.
( Describe the operation of each of the personal communication systems
used in North America.
( Compare the North American personal communication systems.
( Perform calculations of spectral efficiency for PCS.
( Calculate open-loop power and processing gain for CDMA PCS.
( Explain and compare the methods in which data can be carried on PCS.
418 ! CHAPTER 11
' 11.1 Introduction

As pointed out in Chapter 10, the term personal communication is somewhat
vague. In general, it refers to direct communication between people rather
than between places. To understand the concept, consider Joe, a typical pres-
ent-day North American. Joe has a home phone and an office phone. The
home phone has an answering machine and the office phone has voice mail,
but there is no connection between them. Joe also has a cell phone in his car.
That system does not have voice mail, though it could if he wanted to pay
extra. If you want to talk to Joe, you really call a place rather than a person.
You decide if he’s more likely to be at home, in the office, or in the car, and
place your call accordingly. If he’s in none of those places, you leave a mes-
sage. Actually, you probably leave two messages, one at home and one at
work.
Now consider Joan, a slightly more modern North American. Joan also
has office and home phones, but her cell phone is a small portable phone
which she usually carries with her. However, she has to pay airtime on the
cell phone (but not on the others) so she only gives out that number to close
friends and important business associates. Those people, trying to prevent
Joan from having to pay airtime, usually try her home and office numbers
first, then the cellular number. Joan is halfway toward a true PCS.
Move forward a little to the future. Ted has only one phone. It’s small
enough for him to slip into his shirt pocket, and he carries it with him all
day. He has only one phone number. If you want to talk to Ted, you call that
number. If he cannot take calls for a while, he turns off the phone and voice
mail takes over. Similarly, if Ted wants to talk to someone, he uses his pocket
phone—no matter where he is. Ted’s phone number is no longer associated
with a place; it is associated with a person.
It is possible to live like Ted today, but for most people it is tricky. How
does the switchboard at Ted’s company find his personal phone? How does
he charge business calls to the company? How does he get a listing in the
phone book? What happens when he is out of range of his service provider?
All these problems are solvable, and there are many people who feel that,
eventually, all phones will be wireless. Your pocket phone will connect to a
wireless PBX at your office when you’re at work, to a telepoint in your house
when you’re at home, to a cellular or PCS network when you’re out, and to a
satellite-based telephone system when you are at a remote location.
We are still some distance from true PCS, but the gap is closing. The sys-
tems described in this chapter are playing a major role in closing that gap.
Perhaps it is appropriate, then, that the name for the larger concept is often
applied to the systems described in this chapter. These so-called personal
communication systems are derived from the cellular concepts introduced
in Chapters 7 and 10, with enhancements to allow the phones to be smaller
PERSONAL COMMUNICATION SYSTEMS ! 419
and lighter, to have improved battery life, and to have extra features not
available in first-generation cellular systems.
There are three competing types of PCS in North America. This contrasts
vividly with Europe, which simply extended its GSM digital cell phone
system to a higher frequency range for PCS. We have already heard of two of
the three North American personal communication systems. One is in fact
the GSM system, which was mentioned, though not discussed, in Chapter
10. The second is IS-136, the North American digital cell phone standard,
which has also been extended for use at PCS frequencies. The third is a
direct-sequence spread-spectrum system developed in the United States by
Qualcomm and known as IS-95, or by its tradename, CDMAone™. In the
next section we look at some features that all these systems have in common
and contrast them with conventional analog AMPS. Following that, we look
at each of the three systems in more detail.
' 11.2 Differences Between Cellular Systems and PCS

Though based on the same cellular idea as the first-generation cell phone
systems described in Chapter 10, PCS have significant differences which
justify the use of a different term. You should realize, however, that many of
the differences are transparent, or at least not immediately obvious, to the
user. The systems described in this chapter are often called second generation
personal communication systems; in other words, the analog cell phone
system is really the first generation of PCS. The third generation, now being
designed, will feature much wider bandwidth for high-speed data communi-
cation and it will be discussed in Chapter 14.
Frequency Range One of the reasons for establishing new PCS was that the cellular frequency
bands were becoming crowded, especially in major metropolitan areas.
There was no room for expansion in the 800-MHz band, so the new service
was established in the 1900-MHz band (1800 MHz in Europe). This has ad-
vantages in terms of portable antenna size. A few years ago, electronics for
this frequency range would have been prohibitively expensive, but advances
in integrated circuit design have reduced the cost penalties.
In North America the broadband PCS band consists of 120 MHz in the
1900-MHz region. The term broadband here is relative. It refers to bandwidth
sufficient for voice communication and distinguishes this service from such
narrowband services as paging, which will be discussed later in this book.
Sometimes the term broadband communication is used to refer to video and
high-speed data; that is not the sense in which it is used here.
420 ! CHAPTER 11
See Table 11.1 for the PCS band plan. Note that there are six frequency
allocations, so up to six licenses can be awarded in any given area. There are
three 30-MHz and three 10-MHz allocations. The reverse channel or uplink
(mobile to base) is 80 MHz above the forward channel or downlink (base
to mobile) frequency. Reverse and forward channel allocations are separated
by a 20-MHz band, from 1910 to 1930 MHz, which is allocated for un-
licensed services like short-range voice communication. In the United States
the frequencies have been assigned by auction; in Canada licenses were
allocated after public hearings. Some PCS carriers are established cellular
providers with 800-MHz licenses; others are new to the field of wireless com-
munication.
TABLE 11.1 Broadband PCS Band Plan
Base Transmit (Forward Mobile Transmit (Reverse

Allocation Channel or Downlink)* Channel or Uplink)*
A 1850–1865 1930–1945
B 1870–1885 1950–1965
C 1895–1910 1975–1990
D 1865–1870 1945–1950
E 1885–1890 1965–1970
F 1890–1895 1970–1975
*Frequencies are in MHz
Smaller Cell Size Cellular telephony was originally conceived as a mobile radio system, with
phones permanently mounted in vehicles. These phones use efficient exter-
nal antennas on the roof of the vehicle and have a maximum ERP of 4 W.
However, in recent years portable cell phones have outsold mobile phones
by a considerable margin. This has implications for the system, as portable
phones have lower power and are often in difficult locations—from a propa-
gation point of view—such as inside vehicles and buildings. For these rea-
sons, portable AMPS cell phones may not work reliably near the edges of the
larger cells. PCS, on the other hand, were designed from the beginning with
handheld phones in mind. At first it was thought that most PCS users would
be on foot, but it is now quite obvious that subscribers expect to use the
phone wherever they are: outdoors, indoors, in an underground shopping
mall, or in their cars.
PCS cells are typically smaller than AMPS cells to accommodate more
traffic and low-power handheld phones. They must hand off calls very
quickly to handle users in moving cars.
All-Digital System Because of technical constraints, all first-generation cellular systems are ana-
log, though some progress has been made in converting them to digital tech-
nology. In fact, some providers have marketed 800-MHz digital cell phones
as “PCS” systems.
Current digital technology is more efficient than analog FM in its use of
bandwidth. It also allows lower power consumption in the portable phone
and more advanced data communication and calling features. Security and
privacy are inherently better with any digital system, since ordinary scan-
ners cannot be used to intercept calls, and digital coding schemes can also
incorporate encryption as required.
There is one major problem with North American digital PCS. Whereas
first-generation cellular systems in North America all use the same analog
technology, there are three incompatible digital systems in North America.
This makes roaming more difficult with PCS than with cellular phones.
Many providers and phone manufacturers have solved this problem by
offering dual-band, dual-mode phones that are capable of both PCS and ana-
log cellular operation. The solution is not ideal, because it results in phones
that are larger and more expensive than they would otherwise have to be.
Extra Features AMPS systems were designed with POTS (plain old telephone service) in mind.
Even features commonly found on wireline phones, such as call display,
present problems in AMPS. Digital systems allow a substantial amount of
data transmission in their control channels, making all sorts of enhance-
ments possible. In addition to obvious features like call display, digital sys-
tems can allow short printed messages, and even e-mail and limited web
browsing are possible without additional modems and computers. The
features available and the way they are implemented vary with the type of
PCS, and we will look further into this later in this chapter.
Coverage At least at present, the coverage for any PCS is much less universal than it is
for AMPS cell phones. This will undoubtedly change in the future, as the sys-
tems acquire more customers and build more infrastructure. In the mean-
time PCS users have to pay more attention to local coverage areas than do
analog cell phone users.
422 ! CHAPTER 11
Rate Structure One of the arguments for PCS is that they should be less expensive than ana-
log cellular radio. The utilization of spectrum space is more efficient, for ex-
ample. In practice, rates tend to be set by a combination of market forces.
The analog systems have a head start in paying for their infrastructure and
have been able to lower prices to match PCS in many cases.
' 11.3 IS-136 (TDMA) PCS

We looked at IS-136, the North American Digital Cellular standard, in Chap-
ter 10. Most people just refer to it as TDMA (time division multiple access)
when they are talking about PCS, though GSM is also a TDMA system. The
most important difference between the 800-MHz and 1900-MHz versions of
TDMA is that there are no analog control channels in the PCS bands. Rather
than go over the ground already covered in Chapter 10, in this chapter we
will take a closer look at the digital control channel and consider how en-
hanced services are provided. Much of this material is very similar for the
GSM system, described next.
TDMA Digital Recall from Chapter 10 that the digital control channel (DCCH) uses two of
Control Channel the six time slots in a TMDA frame (slots 1 and 4, to be precise). Normally
only one DCCH is required per cell or sector. Figure 11.1 shows how the time
slot is divided up for both forward and reverse channels.
FIGURE 11.1
TDMA digital
control channel
Let us look at the forward channel first. The SYNC (synchronizing) bits
have the same function as for the voice channels, allowing the mobile re-
ceiver to lock on the beginning of the transmission. The SCF (shared channel
feedback) bits perform several functions. They provide acknowledgement of
messages from mobiles and inform the mobiles of the status of the reverse
control channel. Just as in analog AMPS, the forward channel is under the
control of the base station, but many mobiles share a single reverse control
channel. By monitoring the status of the reverse channel as reported by the
base in the SCF field, the mobiles can reduce the possibility of collisions.
These still occur occasionally; however, in that case, the message from the
mobile will not be acknowledged by the base, and the mobile will try again
after a random delay time.
The CSFP (coded superframe phase) bits identify the location of this time
slot in a larger frame that extends over 16 TDMA frames or 32 blocks of
control-channel data, representing a time period of 640 ms. Each block is
designated as containing broadcast, paging, messaging, or access response
information. Each of these types of data can be considered a separate logical
channel of data, time-division multiplexed with the other types and with
voice as well, since four of every six timeslots on the RF channel that carries
the control channel are still used for voice. The number of control-channel
blocks assigned to each type of use can be varied within limits. Table 11.2 on
page 424 summarizes the logical channels, and a brief description of each is
given below.
Two superframes comprise a hyperframe. The hyperframe structure al-
lows data to be repeated. This means that a mobile receiver can check the sig-
nal strength on other channels, without missing data. It also provides
redundancy: if the mobile misses some data because of a burst error, it gets a
second chance.
The broadcast channel contains information intended for all mobiles. It
is divided into two components. The fast broadcast channel (F-BCCH) is used
to transmit system parameters to all the mobiles. These include the structure
of the superframe itself, the system identification, and registration and ac-
cess parameters. All of these must be communicated to the mobile before it
can place a call, so all of this information is transmitted at the beginning of
each superframe. The extended broadcast channel (E-BCCH) has less critical in-
formation, such as lists of the channels used in neighboring cells. This infor-
mation can be transmitted over the course of several superframes.
The short message service, paging, and access channel (SPACH) is used for
control messages to individual telephones and for short paging-type mes-
sages to be displayed on the phone. It is not necessary for every phone to
monitor all these messages; the phone is told which block to monitor and
can go into an idle or sleep mode the rest of the time while it waits for a call.
This helps to extend battery life.
424 ! CHAPTER 11
The reverse control channel is quite different from the forward channel.
There is no broadcast information; there is only one logical channel called
the Random Access Channel (RACH). This is used by the mobile to con-
tact the base, for registration, authentication, and call setup. Normally the
mobile will find out from the broadcast channel whether this channel
TABLE 11.2 Logical Channels in the Data Section of the TDMA Digital Control Channel
Time Slots per

Name of Channel Function Superframe
Broadcast Fast Broadcast Urgent information for all mobiles, 3–10

Channel Channel transmitted once per superframe, at
(BCCH) (F-BCCH) beginning of superframe:
( Superframe structure
( System identification
( Access parameters
( Registration parameters
Extended Less urgent information for all mobiles 1–8

Broadcast (transmitted over several superframes):
Channel ( Neighbor lists (control channel frequencies
(E-BCCH) in nearby cells)
( Regulatory configuration (spectrum
allocation)
( Mobile assisted-channel allocation
(frequencies mobiles should monitor)
Reserved As needed by system 0–7
Short Message Short Message ( Short message service Remaining

Service, Paging, Service Channel Slots
( Remote phone programming
and Access (SMSCH)
Channel
(SPACH) Paging Channel Paging (ringing mobile phone)
(PCH)
Access Response Control messages to individual phones

Channel (ARCH)
is clear; if that information is available, it can transmit at random. If there

is no acknowledgement, it probably means a collision has occurred, and the
mobile will try again after a short random delay. As with the reverse voice
channel, time has to be allocated for ramping up the mobile transmitter
power, and guard time is needed to avoid interference between mobiles at
different distances from the base. Ramp time is shown as R and guard time
as G, in Figure 11.1(b).
' 11.4 GSM

GSM is the system used in Europe and most of Asia for both cellular and PCS
bands. It is not found at 800 MHz in North America, but it is used in the
1900-MHz PCS bands. Note that, even though the same modulation scheme
is used, North American PCS phones will not work in Europe because the fre-
quencies allocated to PCS are different—the European bands center around
1800 MHz.
GSM is another TDMA system but the details are different. GSM also has
some unique features that make it arguably more sophisticated and versatile
than IS-136. It is not compatible with existing IS-136 cell site equipment,
but this is not an issue for the new PCS-only providers, as they have no leg-
acy equipment with which to maintain compatibility.
GSM RF Channels GSM channels are 200 kHz wide (compared with 30 kHz for IS-136 TDMA).
and Time Slots The total bit rate for an RF channel is 270.833 kb/s; the modulation is a vari-
ant of FSK called GMSK (Gaussian minimum shift keying) using a frequency de-
viation of 67.708 kHz each way from the carrier frequency. GMSK was
described in Chapter 4.
Voice channels are called traffic channels (TCH) in GSM. One RF channel
is shared by eight voice transmissions using TDMA. In terms of spectral effi-
ciency, GSM works out to 25 kHz per voice channel, compared to about
30 kHz for AMPS and about 10 kHz for TDMA. This is an approximate com-
parison as it ignores differences in control-channel overhead.
As in TDMA, the mobile transmitter operates only during its allotted
time slot (one-eighth of the time, compared with one-third of the time in
TDMA.) Other things being equal, a GSM phone should have longer battery
life than a phone using either AMPS or TDMA.
Figure 11.2 on page 426 shows the structure of an RF channel and its
division into time slots (called bursts in GSM).
Control information in GSM is on two logical channels called the broadcast
channel (BCCH) and the paging channel (PCH). As with TDMA, it is unnecessary
426 ! CHAPTER 11
FIGURE 11.2
GSM RF
channel
to use a whole RF channel for this. Instead, one of the eight time slots on one
RF channel in each cell or sector is designated as a control channel. The broad-
cast information is transmitted first, followed by paging information. See
Figure 11.3 for an illustration.
FIGURE 11.3
GSM control
channel
The BCCH and PCH are forward channels only. The corresponding re-
verse channel is called the random-access channel (RACH) and is used by the
mobiles to communicate with the base. Mobiles transmit on this channel
whenever they have information; if a collision occurs, the mobile waits a
random time and tries again. Transmissions are shorter than the duration of
the slot to prevent interference caused by the propagation delay between
mobile and base. The delay problem is avoided on the traffic channels, be-
cause the base instructs the mobile to advance or retard the timing of its
transmissions to compensate for the changes in propagation delay as it
moves about in the cell.
Just as with TDMA, it is also necessary to send control information on
the traffic channels. This is because the mobile has only one receiver; it can-
not count on receiving the broadcast channel during a call, because both
channels may use the same time slot. Also as with TDMA, there are two con-
trol channels associated with the traffic channel. The Slow Associated Control
Channel (SACCH) uses one of every 26 bursts on the voice channel. It is used
to inform the base of power measurements made by the mobile of signal

strength in adjacent cells. The Fast Associated Control Channel (FACCH)
“steals” bits from the voice signal and is used for urgent messages from the
base, such as handoff instructions.
Voice Each voice transmission is coded at 13 kb/s. A linear predictive coder, which
Transmission models the way sounds are produced in the human throat, mouth, and
tongue, is used. Such coding allows the bit rate to be greatly reduced com-
pared with straightforward PCM. In the future, it is planned to use more ad-
vanced voice coders (vocoders) to allow the bit rate to be reduced to 6.5 kb/s,
doubling the capacity of the GSM system. Note the similarity with full- and
half-rate TDMA, which code voice at 8 and 4 kb/s, respectively. See Chapter 3
for a discussion of vocoders.
The bits from the vocoder are grouped according to their importance,
with the most significant bits getting the most error correction and the least
significant bits getting none. Then the data is spread over several frames by
interleaving it so that the loss of a frame due to noise or interference will
have a less serious effect.
Each voice transmission is allocated one time slot per frame. A frame
lasts 4.615 ms so each time slot is approximately 577 µs in duration. To allow
time for transmitters to turn on and off, the useful portion of the time slot
is 542.8 µs, which allows time for 147 bits. This gives a raw data rate of
31.8 kb/s per voice channel. The timing for mobile transmissions is critical
so that each arrives at the base station in the correct time slot. Since the
propagation time varies with the distance of the mobile from the base,
the mobile has to advance its timing as it gets farther from the base. It does
this by monitoring a timing signal sent from the base on a broadcast chan-
nel. Although the time slots used by a mobile for receiving and transmitting
have the same number, they are actually separated in time by a period equal
to three time slots (uplink lags downlink). This means that the mobile unit,
unlike analog systems, does not have to receive and transmit at the same
time. When neither receiving nor transmitting on the voice channels, the
mobile monitors the broadcast channels of adjacent cell sites and reports
their signal strengths to the network to help it determine when to order a
handoff. See Figure 11.4 for the structure of a voice channel.
FIGURE 11.4
GSM voice
channel
428 ! CHAPTER 11
Frequency When multipath fading is a problem, the GSM system allows for frequency
Hopping in GSM hopping, a type of spread-spectrum communication that was discussed in
Chapter 4. This can often solve the problem, since multipath fading is highly
frequency-dependent. All GSM mobiles are capable of frequency hopping, but
only those cells that are located in areas of severe fading are designated as hop-
ping cells. The system can hop only among the frequencies that are assigned to
the cell, so there will be only a few hopping possibilities (on the order of three
frequencies). Thus GSM is not really a true spread-spectrum system, but rather a
TDM/FDM system with some spread-spectrum capability added on. This fea-
ture is unique to GSM; IS-136 TDMA has nothing like it.
Subscriber The subscriber ID module (SIM) is unique to the GSM system. It is a smart
ID Module card with eight kilobytes of memory that can be plugged into any GSM
phone. SIMs come in two sizes: one is the size of a credit card, the other is
about postage-stamp size. The SIM contains all subscriber information in-
cluding telephone number (called the International Mobile Subscriber
Identification (IMSI) in GSM), a list of networks and countries where the
user is entitled to service, and other user-specified information such as mem-
ories and speed dial numbers. The card allows a subscriber to use any GSM
phone, anywhere. For instance, since the PCS frequencies are different in Eu-
rope and North America, there is no point in a North American traveling in
Europe with a PCS phone. If a traveler takes the SIM, however, it will work
with any phone rented or purchased in Europe, as long as the subscriber has
first contacted his or her North American GSM service provider to arrange
for authorization.
The SIM also offers some protection against fraudulent use. A GSM phone is
useless without a SIM; if the user removes the card when leaving the phone in
a car, for example, the phone cannot be used unless the thief has a valid SIM.
Unfortunately, the cards can be stolen too. The SIM can be set up to require the
user to enter a personal identification number (PIN) whenever the phone is turned
on to provide some security in case the card is lost or stolen.
Once a subscriber has a SIM, buying a new GSM phone is easy. No setup
or programming by the dealer is required. Similarly, a user can have a perma-
nently-installed mobile phone and a portable with the same phone number,
provided that only one is used at a time. However, the TDMA system also
makes purchasing a new phone fairly easy; it allows a phone to be activated
and programmed over the air, using the control channel.
GSM Privacy The GSM SIM just discussed is only a part of the effort that has gone into
and Security securing this system. Both the data used in authorizing calls, such as the
subscriber’s identifying numbers, and the digitized voice signal itself, are
usually encrypted. (It is possible to weaken or turn off the voice encryption,
if a government requires it.) The security in GSM is better than in IS-136 and
much better than in analog AMPS.
' 11.5 IS-95 CDMA PCS

This United States-designed system has an air interface that is radically dif-
ferent from either of the others, though its control and messaging structure
is quite similar to GSM. CDMA is used to a limited extent on the 800-MHz
band, but is much more common in the 1900-MHz PCS band. It uses code-di-
vision multiple access by means of direct-sequence spread-spectrum modu-
lation. See Chapter 4 for an introduction to spread-spectrum radio and
code-division multiple access (CDMA).
CDMA One CDMA RF channel has a bandwidth of 1.25 MHz, using a single carrier
Frequency Use modulated by a 1.2288 Mb/s bitstream using QPSK. CDMA allows the use of
all frequencies in all cells (not one-seventh or one-twelfth of the frequencies
in each cell, as required by other systems). This gives a considerable increase
in system capacity. Because of the spread-spectrum system, co-channel in-
terference simply increases the background noise level, and a considerable
amount of such interference can be tolerated. As with the other personal
communication systems, base and mobile stations transmit on separate
channels separated by 80 MHz.
Frequency diversity is inherent in any spread-spectrum system. This is
especially beneficial in a mobile environment subject to multipath propaga-
tion. The GSM system discussed earlier can use a limited amount of fre-
quency diversity by hopping among several (typically three) discrete
channels. The CDMA system, on the other hand, uses the full 1.25-MHz
bandwidth for all voice channels on a given RF channel. If a small portion of
this spectrum suffers a deep fade due to reflections, the only effect will be a
slight increase in the error rate, which should be compensated for by the er-
ror correction built into the coding of the voice and control signals.
Space diversity is also built into a spread-spectrum system. Other cellular
systems and PCS typically employ two receiving antennas per cell or sector
at the base station to provide some space diversity, but they use only one an-
tenna at the mobile location. Multiple receiving antennas are also used with
CDMA; but since all frequencies are used in all cells, it is possible to receive
the mobile at two or more base stations. Similarly, a mobile can receive sig-
nals from more than one base station. Each can make a decision about the
strongest signal and can, in fact, combine signals to obtain an even stronger
430 ! CHAPTER 11
one. Since there is no need for the mobile to change frequency on handoff,
the CDMA system can use a soft handoff, in which a mobile communicates
with two or more cells at the same time, rather than having to switch
abruptly from one to another. This gives the ultimate in space diversity, with
receiving antennas up to several kilometers apart. See Figure 11.5 for a com-
parison of soft and hard handoffs.
FIGURE 11.5 Hard and soft handoffs
CDMA Channels Each RF channel at a base station supports up to 64 orthogonal CDMA chan-
nels, using direct-sequence spread-spectrum, as follows:
( 1 pilot channel, which carries the phase reference for the other
channels
( 1 sync channel, which carries accurate timing information (synchro-
nized to the GPS satellite system) that allows mobiles to decode the
other channels
( 7 paging channels, equivalent to the control and paging channels in

TDMA and GSM
( 55 traffic channels
CDMA thus uses a bandwidth of 1.25 MHz for 55 voice channels, which
works out to about 22.7 kHz per channel. This is similar to GSM and, at first
glance, not as efficient as TDMA. However, the fact that all channels can be
used in all sectors of all cells makes CDMA more efficient in terms of spec-
trum than any of the other systems. Since CDMA degrades gracefully with
increasing traffic, it is difficult to arrive at a definite maximum for its capac-
ity. Proponents of CDMA claim spectrum efficiencies ten to twenty times as
great as for GSM; those using other systems dispute this and put the gain
nearer two. Once there is a large body of data from all the PCS schemes, it
will be easier to get at the truth.
Along with the other personal communication systems discussed in
this chapter, the CDMA system also uses FDMA. Each PCS carrier has a spec-
trum allotment of either 5 MHz or 15 MHz in each direction (refer back to
Table 11.1), so a cell site can have more than one RF channel.
Forward Channel The forward and reverse channels are quite different in the CDMA system.
Let us look at the forward channel first. We already know that sync, paging,
and speech channels are combined on the same physical RF channel using
CDMA. We learned in Chapter 4 that the direct-sequence form of CDMA is
created by combining each of the baseband signals to be multiplexed with a
pseudo-random noise (PN) sequence at a much higher data rate. Each of the
signals to be multiplexed should use a different PN sequence. In fact, it can
be shown that if the various sequences are mathematically orthogonal, the
individual baseband signals can, at least in theory, be recovered exactly
without any mutual interference. The math involved in proving this is be-
yond the scope of this text, but we should note that the number of possible
orthogonal sequences is limited and depends on the length of the sequence.
If the PN sequences are not orthogonal, CDMA is still possible, but there will
be some mutual interference between the signals. The effect of this will be
an increased noise level for all signals; eventually, as the number of non-
orthogonal signals increases, the signal-to-noise ratio becomes too low and
the bit-error rate too high for proper operation of the system. However, at no
time do we hear audible crosstalk, as we do with two analog signals on the
same frequency.
From the foregoing it would seem that using orthogonal PN sequences
for CDMA is highly desirable, and this is what is done at the base station.
A class of PN sequence called a Walsh code is used. The base station uses
432 ! CHAPTER 11
64 orthogonal Walsh codes; each repeats after 64 bits. This allows for 64 in-
dependent logical channels per RF channel, as mentioned earlier. Walsh
code 0 is used for the pilot channel to keep mobile receivers phase-aligned
with the base station. This is a requirement for coherent demodulation,
which is the only way to avoid interference among channels using the same
carrier frequency.
In addition to the Walsh codes, two other codes are in use at a CDMA
base station: a short code for synchronizing, and a long code, which is used
for encryption of both voice and control-system data and is not used for
spreading.
Figure 11.6 shows how the spreading works for one voice signal. The
vocoder produces a voice signal with a maximum bit rate of 9.6 kb/s.
Error-correction bits are added, and the samples are interleaved over time,
just as they are for TDMA and GSM. This process increases the bit rate for one
voice channel to 19.2 kb/s. Next the signal is exclusive-or’d with the long
code. This code repeats only after 242 − 1 bits and is used for encryption, not
spreading. The signal remains at 19.2 kb/s after this process.
FIGURE 11.6 CDMA forward voice channel
Spreading occurs when the 19.2 kb/s baseband data stream is multi-
plied by one of the 64 Walsh codes. Each of the Walsh codes has a bit rate of
1.2288 Mb/s. The multiplication works as follows: when the data bit is zero,
the Walsh code bits are transmitted unchanged; when the data bit is one, all
Walsh code bits are inverted. The output bit stream is at 1.2288 Mb/s, which
is 64 times as great a data rate as for the baseband signal at 19.2 kb/s. There-
fore, the transmitted signal bandwidth is 64 times as great as it would be for
the original signal, assuming the same modulation scheme for each.
Recall from Chapter 4 that the processing gain can be found as follows:
BRF
Gp = (11.1)
BBB
where
G p = processing gain
B RF = RF (transmitted) bandwidth
B BB = baseband (before spreading) bandwidth
Here,
BRF
Gp =
BBB
1.2288 × 106
=
19.2 × 103
= 64
In decibels, this is
G p (dB) = 10 log 64
= 18.06 dB
If we consider that the error-correction codes are a form of spreading as well,

since they increase the data rate, the total spreading becomes
BRF
Gp =
BBB
1.2288 × 106
=
9.6 × 103
= 128
= 21.1 dB
A signal-to-noise ratio of about 7 dB is required at the receiver output for

a reasonable bit-error rate. This means that the signal-to-noise ratio in the RF
channel can be about −14 dB for satisfactory operation; that is, the signal
power can be 14 dB less than the noise power. This takes a little getting used
to, but is typical of spread-spectrum systems.
The 64 orthogonal channels are transmitted on one RF carrier by sum-
ming them, as in Figure 11.7, and using QPSK to modulate them on a single
carrier.
434 ! CHAPTER 11
FIGURE 11.7
Multiplexing of
CDMA channels
Reverse Channel The mobile units cannot use truly orthogonal channels because they lack a
phase-coherent pilot channel. Each mobile would need its own pilot chan-
nel, which would use too much bandwidth. Therefore, they use a more ro-
bust error-control system. It outputs data at three times the input data rate.
Follow Figure 11.8 to see what happens to the signal.
FIGURE 11.8 CDMA reverse voice channel

ƒ 2 = 3 × 9.6 kb/s
= 28.8 kb/s
The 28.8 kb/s signal is combined with one of the 64 Walsh codes and a
long code to reach the full data rate of 1.2288 Mb/s. However, the purpose of
each of these codes is different on the reverse channel. Here the long code is
used to distinguish one mobile from another, as each uses a unique (though
not necessarily orthogonal) long code. The Walsh codes are used to help the
base station decode the message in the presence of interference. Each block
of six information bits (64 different possible combinations) is associated
with one of the 64 Walsh codes, and that code, rather than the actual data
bits, is transmitted. Since each Walsh code is 64 bits long, this in itself does
some spreading of the signal: the bit rate is increased by a factor of 64/6. The
Walsh code mapping thus increases the data rate as follows:
64
ƒ 3 = 28.8 kb/s ×
6
= 307.2 kb/s
The long code is now multiplied with the data stream to produce a re-
verse channel bit rate of 1.2288 Mb/s, the same as for the forward channel.
Each mobile transmits at the same rate to produce the spread-spectrum sig-
nal received at the base.
The modulation scheme is also slightly different on the reverse and for-
ward channels. Both use a form of quadrature phase-shift keying (QPSK).
The base station uses conventional QPSK. With this system the transmitter
power has to go through zero during certain transitions. See Figure 11.9(a)
on page 436.
The mobiles delay the quadrature signal by one-half a bit period to pro-
duce offset QPSK, which has the advantage that the transmitter power never
goes through zero, though the amplitude does change somewhat. Linear am-
plifiers are still required in the mobile transmitter, but the linearity require-
ments are not as strict for offset QPSK as they are for conventional QPSK. See
Figure 11.9(b).
Offset QPSK would have no advantage for the base station because a sin-
gle transmitter is used for all the multiplexed signals. The summing of a large
number of signals would result in a signal that still went through the zero-
amplitude point at the origin.
Voice Coding CDMA uses a variable rate vocoder. Four different bit rates are possible:
9600, 4800, 2400 and 1200 b/s. The full rate of 9600 b/s is used when the user
is talking. During pauses, the bit rate is reduced to 1200 b/s. The other two
rates are also in the specifications but are seldom used.
436 ! CHAPTER 11
FIGURE 11.9
Standard and offset
QPSK
For many years it has been realized that each user typically talks less
than fifty percent of the time during a conversation. Theoretically, the band-
width allocated to that customer can be reassigned during the pauses while
the other person is talking. However, until CDMA PCS came along there
were at least two problems with this. The first was that it does not sound nat-
ural to have the voice channel go dead when someone stops talking. It
sounds as if the phone has been disconnected. The reason is that there is
always background noise, even in a quiet room. The CDMA system transmits
this noise, but codes it at a lower rate (1200 b/s) because it is not important
that it be rendered accurately.
The other problem, with either FDMA or TDMA, was finding a use for the
vacated channel or time slot. The slot is usually available for only a few sec-
onds or less, and the amount of time is not known in advance. In CDMA the
reduced amount of information to be sent can be translated directly into re-

duced interference to other transmissions on the same frequency, which au-
tomatically increases the capacity of the system. The way in which this is
done is different for the forward and reverse channels.
On the forward channel, data bits are repeated when the coder is run-
ning at less than the maximum rate of 9.6 kb/s. For instance, if the coder
operates at 1.2 kb/s, as it does during pauses in speech, each block of data is
transmitted eight times. Because the error rate at the receiver depends on
the energy per received data bit, the power in the transmit channel can be
reduced under these circumstances.
The mobile transmitter handles this situation differently. Rather than
reduce power, it simply transmits only one-eighth of the time, reducing
interference and increasing battery life.
Mobile Power Controlling the power of the mobile stations is even more important with
Control CDMA than with other schemes. The power received at the base station from
all mobiles must be equal, within 1 dB, for the system to work properly. The
power level is first set approximately by the mobile, and then tightly con-
trolled by the base. When first turned on, the mobile measures the received
power from the base, assumes that the losses on the forward and reverse
channels are equal, and sets the transmitter power accordingly. This is called
open-loop power setting. The mobile usually works with the equation:
P T = −76 dB − P R (11.2)
where
P T = transmitted power in dBm
P R = received power in dBm
EXAMPLE 11.1 Y
A CDMA mobile measures the signal strength from the base as −100 dBm.
What should the mobile transmitter power be set to as a first approxi-
mation?
SOLUTION
From Equation (11.2):
PT = −76 dB − P R
= −76 dB − (−100 dBm)
= 24 dBm
≈ 250 mW
X
438 ! CHAPTER 11
The mobile begins by transmitting at the power determined by Equa-

tion (11.2) and increases power if it does not receive acknowledgement from
the base. This could happen if a substantial amount of the received power at
the mobile is actually from adjacent cells. We should also remember, that
just as for the other systems, the forward and reverse channels are at differ-
ent frequencies, so the amount of fading may be different.
Once a call is established, the open-loop power setting is adjusted in 1 dB
increments every 1.25 ms by commands from the base station, to keep the
received power from all mobiles at the same level. This closed-loop power
control is required; for CDMA to work properly, all the received signals must
have equal power. Otherwise the system suffers from the near/far effect, in
which the weaker signals are lost in the noise created by the stronger ones.
Careful power control has the added benefit of reducing battery drain in the
portable unit, as the transmitted power is always the minimum required for
proper operation of the system.
Rake Receivers and One of the advantages of the CDMA system is that multipath interference
Soft Handoffs can be reduced by combining direct and reflected signals in the receiver. The
receivers used are called rake receivers; the reason can be seen in the dia-
gram in Figure 11.10, which somewhat resembles a rake with several teeth
for the reception of signals having different amounts of delay.
FIGURE 11.10 Rake Receiver

The mobile unit can combine three RF signals, delaying two of them to
match the third. One of these signals can be assumed to be the base station
in the current cell. The other two may be reflections or neighboring base
stations. The base-station receiver can combine four signals: the direct sig-
nal from the mobile and three reflections.
In addition, two base stations may receive a signal from the same mo-
bile. The base stations each send their signals to the MSC, which uses the
higher-quality signal. Decisions about quality are made on a frame-by-frame
basis every 20 ms. It is possible to have two base stations communicating
with the same mobile indefinitely in what is referred to as a soft handoff.
This avoids the dropping of calls that sometimes occurs when a handoff is
unsuccessful in other systems, perhaps because there are no available chan-
nels in the new cell. The disadvantage is a considerably increased load on the
base stations and the switching network.
CDMA Security CDMA offers excellent security. A casual listener with a scanner will hear
only noise on a CDMA channel. In order to decode a call it is not only neces-
sary to have a spread-spectrum receiver, but also to have the correct
despreading code. Since this so-called “long code” is 242 − 1 bits long before
it repeats and is newly generated for each call, the chances of eavesdropping
are small. Identification is done using private-key encryption, as for GSM.
' 11.6 Comparison of Modulation Schemes

All of the North American PCS have advantages. TDMA is compatible with
much existing North American cell-site equipment. GSM has a long history
and a large installed base, which tends to lead to lower prices. It also has
more advanced features than IS-136. CDMA is the most sophisticated tech-
nically, offers the best security, and makes the best use of system bandwidth,
at least in theory. All three are in wide use in the United States and Canada.
Table 11.3 on page 440 compares the three systems under several headings.
Compatibility From Table 11.3 there appears to be an obvious compatibility problem in

Issues: Multi- PCS. The three systems have only their frequency range in common; none of
Mode Phones the systems is compatible with either of the others. Consequently, roam-
ing in the PCS band is possible only among providers that use the same sys-
tem. At this writing, one or more of these systems is available in most
populous areas, but not all areas have all three. Eventually the problem may
disappear, as PCS coverage becomes ubiquitous with all three systems, but in
the meantime, there is substantially less roaming capability with PCS than
440 ! CHAPTER 11
TABLE 11.3 Comparison of North American PCS
Property IS-136 TDMA GSM IS-95 CDMAone
RF Channel Width 30 kHz 200 kHz 1.25 MHz
Voice/Data Channels 3 8 64, including 1 sync

per RF Channel and 7 control channels
Multiplexing Type TDMA TDMA plus limited Direct-sequence

frequency-hopping spread-spectrum
Voice Coding Rate 8 kb/s full rate 13 kb/s full rate Variable
4 kb/s half rate 6.5 kb/s half rate 9600 b/s max.
1200 b/s min.
Bit Rate for RF Channel 48.6 kb/s 270.833 kb/s 1.2288 Mb/s
Control Channel 2 time slots of 6 in 1 time slot of 8 in 7 of 64 orthogonal

one 30-kHz RF channel one 200-kHz RF codes in one 1.25-MHz
channel RF channel
with 800-MHz analog AMPS, which still has by far the widest distribution of
any system in North America.
There is an obvious, though rather unwieldy, solution to the compatibil-
ity problem. This is to manufacture dual-band, dual-mode phones, which
work with analog, 800-MHz AMPS as well as with one of the 1900-MHz per-
sonal communication systems. Dual-mode phones are currently available
for all three of the PCS. Those PCS providers who do not also have an
800-MHz license often form alliances with a cellular provider to allow seam-
less roaming with only one monthly bill. Figure 11.11 shows examples of
dual-mode phones.
FIGURE 11.11
Dual-mode
phones
(Courtesy of
Nokia, Inc.)
' 11.7 Data Communication with PCS

When we studied the TDMA cellular system, we observed that data commu-
nication can actually be more complex with a digital than with an analog
system. This is because vocoders will not work properly with modems, so
that the classic technique of connecting a modem to an analog voice chan-
nel does not work. We saw that at 800 MHz it is common for a digital phone
to revert to analog mode for circuit-switched data communication and to
use the CDPD system for packet-switched data.
The above techniques are still possible with a dual-mode PCS phone, but
each of the three personal communication systems has developed its own
techniques for data communication. This can be expected to become more
important as a new generation of “smart phones” incorporating larger dis-
plays, (some even including web browsers) and portable computers incorpo-
rating RF communication modules are introduced.
At present the most popular use for PCS data seems to be short paging-
type messages, followed by electronic mail. Worldwide web access is gaining
in importance, but is currently limited by slow connection speeds and the
limited graphics capability of PCS phone displays. Let us see how data trans-
mission is handled with each of the three PCS.
TDMA Data The TDMA PCS standard allows for short messages and packet-switched data
Communication to be sent on the digital control channels (DCCH) or the digital traffic chan-
nels (DTC). Circuit-switched data is possible on the digital traffic channels.
The digital control and traffic channels support two main types of
packet-switched data communication. A format called cellular messaging
teleservice (CMT) is employed for a short messaging service (SMS). This
allows for brief paging-type messages and short e-mail messages (up to
239 characters), which can be read on the phone’s display and entered using
the keypad. For longer messages and extended services like web brows-
ing, the Generic UDP Transport Service (GUTS) protocol is used. The acronym-
within-an-acronym UDP stands for User Datagram Protocol.
Both of these services require extra equipment in the PCS network to
translate between wireline protocols and those used with the radio link.
With GUTS, the user connects to a network server that relays messages to
and from the internet. The CMT system also requires the servers in the PCS
network to assemble messages and interconnect with other services such as
the user’s e-mail service. See Figure 11.12 on page 442 for an illustration of
packet-switched PCS data.
Circuit-switched data communication is accomplished on the digital
traffic channels. The vocoder is bypassed and data is coded and sent directly
442 ! CHAPTER 11
FIGURE 11.12
PCS packet-
switched data
over a traffic channel at up to 9.6 kb/s. It is also possible to combine two or

three traffic channels to send data at higher rates. This is done by using four
or six of the time slots on a single RF channel, rather than the usual two.
Circuit-switched data is more expensive for the network to carry, as
it uses at least one traffic channel full-time. On the other hand, it allows
the user more freedom as to the type of data and the network called. For in-
stance, a user could dial directly into a company mainframe computer. All
that the PCS network has to provide is an interface to convert the protocols
used on the air interface to an ordinary wireline modem standard for com-
munication with the PSTN. At the mobile phone, a serial-port interface is
typically provided for plugging in a computer. There is no modem in the
phone, but it is set up to appear like a standard wireline modem to the com-
puter. Figure 11.13 shows how circuit-switched data works.
GSM Data The types of data communication possible with GSM are similar to those
Communication used with TDMA. Short messages are available (up to 160 characters) using
either the control or traffic channels, depending on whether the phone is in
use for a voice call at the time. Circuit-switched data (including fax) can be
accommodated at up to 9600 b/s using a traffic channel, just as for TDMA. A
device especially designed to take advantage of GSM data communication,
the Nokia 9000il Communicator, is shown in Figure 11.14.
FIGURE 11.13
PCS circuit-
switched data
FIGURE 11.14
Nokia communicator
(Courtesy of Nokia, Inc.)
CDMA Data There are some differences between CDMA and the other two systems
Communication in terms of data communication. Like the others, CDMA offers short mes-
sages via control channels. Its circuit-switched data capability using a single
traffic channel is much greater, though, at 14.4 kb/s.
444 ! CHAPTER 11
Wireless Web Any of the PCS schemes just described can be used to access content on the
Browsing World Wide Web. There are three major problems with all of them however:
the data rate is low, even in comparison with ordinary telephone modems; the
on-board computing power is low compared with a personal computer; and
the handheld devices have very small, low-resolution displays. Many of these
displays are not suitable for graphics. A typical web page would take a long time
to load and when loaded would be almost, if not completely, unusable.
Third-generation wireless systems, which are described in Chapter 14,
will help to solve the first problem, and perhaps make a start on the second.
The third is more intractable: large displays and pocket-sized devices are
simply not compatible. Therefore, even with third-generation systems,
there will be a need for a means to display web pages on the small screens of
PCS devices.
Until recently there have been many proprietary standards for display-
ing web content on wireless devices. Each worked only with a small number
of specially created sites. Many of the major wireless manufacturers, includ-
ing Ericsson, Nokia, and Motorola, have now combined to create a set of de
facto standards for creating this content, known as the Wireless Application
Protocol (WAP). The idea is to include a small program called a microbrowser
in the wireless device, with most of the required computing done on net-
work servers. These servers have access to specially modified pages on web
sites and can also attempt to translate conventional sites so that they can be
used by wireless devices. The pages have minimal graphics and condensed
text so that they can be used with portable devices.
WAP is compatible with all of the current (second generation) systems
and will be compatible with all third-generation systems as well. As more
sites begin to provide pages compatible with WAP, the web should become
quite accessible to portable wireless devices.
' 11.8 Testing Cellular Systems and PCS

We saw in Chapter 7 that all calculations of signal strength in a mobile envi-
ronment are necessarily approximate, as there are too many variables for
even a computer analysis to be accurate. Nonetheless, these predictions are
generally accurate enough to locate cell sites and repeaters. Once these have
been built, it is necessary to go into the field with a receiver to verify that the
signal strength is satisfactory. The transmitter used for preliminary tests is
often a portable model that puts out a carrier only. This simplifies the mea-
surements, particularly when the system will eventually be CDMA with all
channels active in all cells. Figure 11.15 shows a typical portable transmitter
and receiver suitable for such testing.
(a) Transmitter (b) Receiver
FIGURE 11.15 Test transmitter and receiver

(Courtesy of Berkeley Varitronics Systems)
In addition to the RF equipment, it is desirable to have the test vehicle

equipped with a global positioning system (GPS) receiver and a computer. It
is then possible to plot signal strengths on a map and then to draw signal
contour lines. In this way, any problem areas can be identified.
Once the system is in use, signal measurements can still be taken, but the
process is more complex. This is especially true for CDMA, because all cells
transmit on all frequencies with low power spread over a wide bandwidth. A
conventional receiver would not be able to distinguish signals from noise, or
one cell from another. A specialized receiver capable of extracting the pilot
channel from a CDMA transmission is required. Each cell transmits a pilot
with a slightly different delay added to the short code, and thus it can be
identified if an accurate timing reference is available. Again, GPS is used by
the CDMA system for this function. Figure 11.16 shows a specialized instru-
ment designed for this purpose.
Testing of individual phones presents a problem, especially since it
must be done quickly and economically; otherwise it will be less costly to
446 ! CHAPTER 11
FIGURE 11.16
CDMA test receiver
(Courtesy of Berkeley
Varitronics Systems)
replace the phone than to repair it. Specialized and quite costly equipment is
required. Figure 11.17 shows a unit that can be used to test all the functions
of a PCS phone.
FIGURE 11.17
Mobile phone tester
(Courtesy of Agilent
Technologies, Inc.)
( The eventual goal of personal communication systems is to allow each in-
dividual to have one personal phone and phone number which will take
the place of home, office, car, and portable phones.
( Current PCS resemble cellular radio systems except that they operate at a
higher frequency and are completely digital.
( The PCS frequency range is divided in such a way that there can be up to
six service providers in a given region.
( There are three incompatible PCS in current North American use: IS-136
TDMA, GSM1900, and IS-95 CDMAone. TDMA is also widely used at
800 MHz, and there is some use of CDMA in the 800-MHz band.
( All three PCS are digital. IS-136 and GSM1900 use TDMA, but with differ-
ent modulation schemes and protocols. CDMAone uses a direct-sequence
spread-spectrum system.
( Dual-mode, dual-band phones allow roaming on the analog cellular
system when the correct type of PCS service is unavailable.
( In addition to voice, all systems have provisions for short written mes-
sages and also for circuit-switched data on their traffic channels.
( Equation List
BRF
Gp = (11.1)
BBB
PT = −76 dB − PR (11.2)
( Key Terms
downlink signal path from a base station or a satellite to a mobile station
or a ground station
interleaving changing the time order of digital information before
transmission to reduce the effect of burst errors in the channel
International Mobile Subscriber Identification (IMSI) in the GSM system,
a telephone number that is unique to a given user, worldwide
near/far effect in a spread-spectrum system, the tendency for stronger
signals to interfere with the reception of weaker signals
rake receiver a radio receiver that is capable of combining several
received signals with different time delays into one composite signal
448 ! CHAPTER 11
short messaging service (SMS) transmission of brief text messages, such

as pages or e-mail, by cellular radio or PCS
smart card a card with an embedded integrated circuit, that can be used
for functions such as storing subscriber information for a PCS system
soft handoff in a cellular or PCS system, connecting a mobile to two or
more base stations simultaneously
subscriber ID module (SIM) in the GSM system, a “smart card”
containing all user information, which is inserted into the phone
before use
uplink transmission channel to a satellite or base station
Walsh code class of orthogonal spreading codes used in CDMA
communication
( Questions
1. The systems described in this chapter are sometimes called second-
generation PCS. What was the first generation? What is expected from
the third generation?
2. What frequency range is used for North American PCS? How were fre-
quencies assigned to service providers?
3. How is interference between base and mobile transmissions prevented
in PCS?
4. Give two other terms for the RF channel from base to mobile.
5. Give two other terms for the RF channel from mobile to base.
6. How is it possible for a digital system to use less bandwidth than a tradi-
tional analog FM system?
7. What other advantages do digital systems have over analog?
8. Do current PCS implementations have any disadvantages compared
with analog 800-MHz cellular radio?
9. How does IS-136 PCS TDMA differ from the TDMA used at 800 MHz?
10. What are the meaning and function of the SCF bits in the TDMA digital
control channel?
11. What happens if two mobiles transmit simultaneously on the reverse
control channel in the TDMA system?
12. What is meant by a logical channel, and how is the control channel di-
vided into logical channels in the TDMA system?
13. What is the meaning of a superframe in TDMA? What information must

be transmitted completely in one superframe?
14. What is a hyperframe in TDMA, and what is its function?
15. What is sleep mode for a cell phone, and what is the advantage of being
able to use a sleep mode?
16. Why won’t a North American TDMA phone work in Europe?
17. Compare the number of TDMA channels on one RF channel for the
TDMA and GSM systems.
18. Compare the RF channel bandwidths for the TDMA and GSM systems.
Which uses bandwidth more efficiently?
19. Why do GSM and TDMA sometimes have to use traffic channels for con-
trol information?
20. How does GSM handle handoff requests from base to mobile?
21. What does a GSM mobile do, during a call, in those time slots when it is
neither transmitting or receiving voice information?
22. How does frequency hopping improve the reliability of a communica-
tion system in the presence of multipath interference?
23. Which of the three North American PCS can use frequency hopping to
improve its performance?
24. Describe the GSM SIM and explain how it makes traveling to other
countries easier for GSM users than for those with other types of PCS
phones.
25. How does GSM prevent people from using the system without authori-
zation?
26. Why won’t a North American GSM phone work in Europe?
27. How would a North American GSM user obtain service on a trip to
Europe?
28. Explain the difference between frequency-hopping and direct-sequence
spread-spectrum systems. Which of these is used with the CDMAone
personal communication system?
29. How does CDMA improve the frequency reuse situation compared with
TDMA and GSM?
30. Explain what is meant by a soft handoff. What advantages does it have?
31. Why are soft handoffs not possible with TDMA or GSM?
32. What is the use of the Walsh codes for each of the forward and reverse
links of the CDMA system?
450 ! CHAPTER 11
33. What is the use of the long code in each of the forward and reverse links
in the CDMA system?
34. How does the CDMA system take advantage of pauses in speech to
increase system capacity? Explain how it is done in both forward and
reverse channels.
35. Why does QPSK require linear amplifiers while GMSK does not? Which
PCS use each of these modulation schemes?
36. What advantage does offset QPSK have over normal QPSK? Where is
offset QPSK used in PCS?
37. Why is control of the mobile transmitter power more important in
CDMA than in the other PCS?
38. Explain the difference between open-loop and closed-loop power
control.
39. What is a rake receiver, and what benefits does it provide?
40. Explain the difference in the way PCS handle short messages and longer
data transmissions.
41. What is the difference between circuit-switched and packet-switched
data transmissions?
42. Which of the personal communication systems has the highest speed
for circuit-switched data?
43. Why is it difficult to measure signal strength in a CDMA PCS?
( Problems
1. (a) Calculate the length of a quarter-wave monopole antenna for the
center of the cellular radio band (860 MHz) and the center of the
PCS band (1910 MHz).
(b) Assuming equal transmitter power applied to both of the above anten-
nas, would there be any difference in the power density generated at a
receiving antenna? Assume free-space propagation. Explain.
(c) If the above antennas are used as receiving antennas, which of the
two would produce a stronger signal at the receiver input, and by
how many decibels? Assume equal power density at the antenna.
Explain.
2. Compare the spectral efficiency of the TDMA and GSM PCS and analog
AMPS in terms of kHz used per voice channel. Proceed as follows:
(a) Take a simplistic approach to get a rough estimate. Find the width of
each RF channel and the number of voice channels per RF channel.
Ignore control channels. Calculate spectrum use in kHz per chan-

nel.
(b) Refine your estimate by considering the proportion of the overall
bandwidth used by control channels in each system and therefore
unavailable for voice. Assume a 12-cell repeating pattern with one
control channel for each cell in TDMA and GSM. Also assume the
use of one of the larger PCS frequency blocks. Check Chapter 10 to
refresh your memory about AMPS control channels.
(c) Compare CDMA with the others. First find the bandwidth per voice
channel for one cell, then take into account the gain in capacity
that results from the ability to use all channels in all cells to arrive at
an equivalent bandwidth per voice channel.
3. (a) Suppose that the SMSCH in a TDMA system occupies five slots
in each superframe. What is the total bit rate available for short
messages?
(b) Suppose that five users in the cell or sector have messages addressed
to them at the same time. What is the bit rate available for each
user?
4. As we saw earlier in this book, ordinary digital wireline telephony uses a
bit rate of 64 kb/s for each voice channel. Find the ratio by which the
data has been compressed in each of the digital systems discussed.
Assume full-rate vocoders for TDMA and GSM.
5. Compare TDMA, GSM, and CDMA by finding for each (take note that
there are two different modulation schemes in use):
(a) the RF channel rate in bits per second per hertz of bandwidth
(b) the RF channel rate in symbols per second per hertz of bandwidth
6. (a) What is the difference in propagation time for signals from two
mobiles, one 150 m from the base, and the other 2 km away?
(b) How many bit periods in the RF channel correspond to this differ-
ence, for each of the three systems discussed in this chapter?
7. (a) Calculate the time for one complete repetition of the Walsh code on
the forward channel.
(b) Repeat (a) for the long code. (Now do you see why it’s called the
long code?)
8. (a) A CDMA mobile measures the signal strength from the base as −85
dBm. What should the mobile transmitter power be set to as a first
approximation?
(b) Once the connection is made with the above power level, the base
station needs the mobile to change its power to +5 dBm. How long
will it take to make this change?
452 ! CHAPTER 11
9. A rake receiver in a mobile receives a direct signal from a base 1 km away

and a reflected signal from a building 0.5 km behind the mobile. It also
receives a signal from another base station 3 km away. See Figure 11.18
for the situation. Calculate the amount of time delay each “finger” of
the receiver needs to apply.
FIGURE 11.18
Satellite-Based
Wireless Systems
12
Objectives
( Calculate velocity and period for artificial satellites in circular orbits.
( Distinguish between low-, medium-, and geostationary earth orbits, and
explain the advantages and disadvantages of each for communication.
( Explain how elliptical orbits can be used in communication satellites.
( Explain the need for satellite tracking, and describe how it is done.
( Describe the types of satellite transponders used for wireless
communication.
( Perform signal, noise, and signal-to-noise ratio calculations with satellite
links.
( Describe several current and projected projects in the field of wireless
communication by satellite, and discuss the merits of each.
454 ! CHAPTER 12
' 12.1 Introduction

Cellular and personal communication systems work very well in populous
areas, but they do require extensive infrastructure. Such systems are imprac-
tical for use in remote areas or at sea. Until quite recently the only means of
portable communication in such areas was high-frequency (HF) radio.
Over-the-horizon propagation is possible at frequencies in the range from
about 3–30 MHz, due to reflection from the ionosphere; but HF propagation
is unreliable, noisy, and suitable only for low-bandwidth applications. The
antenna sizes required for wavelengths in the 10 to 100 meter range are
awkward for mobile and all but impossible for handheld use.
A better alternative is to use satellites to relay signals from mobile or por-
table users to the public switched telephone network. Problems of cost, time
delay, and equipment size exist, however, and will be addressed as we pro-
ceed. It seems unlikely that satellite wireless communication systems will
ever be as popular as terrestrial systems, but they will fill an important role
where terrestrial systems are not practical.
' 12.2 Satellite Orbits

We looked very briefly at satellite orbits in Chapter 7. In theory, a satellite
can orbit at any altitude, but air resistance makes satellites impractical below
about 300 km. A satellite can have any elliptical orbit but most (not all) of
those used for communication have orbits that are at least approximately
circular.
Satellite orbits can circle the equator (equatorial orbit); they can pass
over both poles in a polar orbit, which has a 90° angle with respect to
an equatorial orbit; or they can have any angle between these. Figure 12.1
shows some examples of possible circular orbits.
Orbital Satellites are held in orbit by their momentum. Gravity continually bends a
Calculations satellite’s path toward the earth, but the satellite’s momentum is sufficient
to prevent its trajectory from reaching the earth. This phenomenon is com-
monly called centrifugal force, though technically there is no such thing.
Any satellite orbiting the earth must satisfy this equation:
4 × 1011
v = (12.1)
(d + 6400)
SATELLITE-BASED WIRELESS SYSTEMS ! 455
FIGURE 12.1
Circular orbits
where
v = velocity in meters per second
d = distance above earth’s surface in km
Several important things can be seen from Equation (12.1). First, the far-
ther a satellite is from the surface of the earth, the slower it travels. Second,
since a satellite that is farther from the earth obviously has farther to go to
complete an orbit, and since it also travels more slowly than one closer
to earth, then the orbital period of a distant satellite must be longer than the
period of one closer to the earth. An example will illustrate this.
EXAMPLE 12.1 Y
Find the velocity and the orbital period of a satellite in a circular orbit
(a) 500 km above the earth’s surface
(b) 36,000 km above the earth’s surface
SOLUTION
(a) From Equation (12.1):
4 × 1011
v =
(d + 6400)
4 × 1011
=
(500 + 6400)
= 7.6 km/s
456 ! CHAPTER 12
The circumference of the orbit can be found from its radius, which is
that of the earth (6400 km) plus the distance of the satellite from the earth.
In this case the total distance is
r = 6400 km + 500 km = 6900 km
and the circumference of the orbit is
C = 2πr
= 2π × 6900 km
= 43.4 Mm
The period of the orbit can be found by dividing the circumference by

the orbital velocity,
C
T =
v
43.4 × 106 m
=
7.6 × 103 m/s
= 5.71 × 103 s
= 1.6 hours
(b) Again, start with Equation (12.1):
4 × 1011
v =
(d + 6400)
4 × 1011
=
(36,000 + 6400)
= 3.07 km/s
Note that the speed is less than before. The new radius is
r = 6400 km + 36,000 km
= 42.4 Mm
The circumference of the orbit is, as before,
C = 2πr
= 2π × 42.4 Mm
= 266.4 Mm
Now we can find the period of the orbit.

C
T =
v
266.4 × 106 m
=
3.07 × 103 m/s
= 86.8 × 103 s
≅ 24 hours
X
Geostationary The satellite in Example 12.1(b) has a particularly interesting orbit. It is, at
Orbit least approximately, geosynchronous; that is, the satellite orbits the earth
in the same amount of time it takes the earth to rotate once on its axis. If the
orbit is circular and above the equator and the satellite travels in the same di-
rection as the earth’s rotation, it will also be geostationary; that is, it will
appear stationary from the ground, because it rotates at the same rate and
in the same direction as the earth. Though it is theoretically possible for a
satellite to be geosynchronous without being geostationary by orbiting
the earth in the direction opposite to the earth’s rotation, in practice this
is never done, and the two terms are used interchangeably. The orbit for a
geostationary satellite has a radius more than five times as large as that of
the earth. See Figure 12.2 for an illustration of the geostationary orbit.
FIGURE 12.2
Geostationary
orbit
(not to scale)
458 ! CHAPTER 12
Elliptical Orbits Technically all satellite orbits are elliptical. A circle is a special case of ellipse
where the maximum distance from the earth (apogee) is equal to the mini-
mum distance (perigee).
When a satellite has an elliptical orbit in which the earth is at one focus,
it spends more time in that part of the orbit that takes it farthest from earth.
This can be seen from Kepler’s second law, which states that such a satellite
sweeps out equal areas in space in equal times. See Figure 12.3 for the idea.
FIGURE 12.3
Elliptical orbit
Such orbits can be used to advantage. The Russian Molniya series of com-
munication satellites, for instance, uses highly elliptical polar orbits ar-
ranged so that the satellites spend about 11 hours of a 12-hour orbit over the
northern hemisphere. This facilitates communication in regions near
the North Pole. Geostationary satellites, which must be in a circular orbit
above the equator, are near the horizon in these regions. The proposed
Ellipso satellite system, described later, will also use elliptical orbits.
' 12.3 Use of Satellites for Communication

The traditional way to communicate with a satellite in non-geostationary
orbit is to use a movable, directional antenna and point it at the satellite.
The approximate azimuth (angle in the horizontal plane) and elevation (verti-
cal angle measured from the horizon) can be calculated from orbital data,
and fine position adjustments can be made for the strongest signal. The
antenna must then follow, or track, the position of the satellite as it moves.
At some point, of course, the satellite goes below the horizon and communi-
cation is lost.
There are two obvious problems with this for mobile and portable com-
munication. The first is that rapidly changing the orientation of a highly
directional antenna on a moving vehicle is difficult. This can be done on
ships where changes in direction are relatively slow and there is room for a
dish antenna, but it is very difficult on cars and just about impossible for
a portable, hand-carried unit. Moving vehicles require antennas that either
have a broad beam or are steered electronically using phased elements.
The disappearance of satellites over the horizon is another serious prob-
lem. If real-time voice communication is to be possible, at least one satellite
must be visible at all times. Two solutions to the problem are explored below.
Using Satellites There is an orbit in which a satellite appears to be stationary above a particu-
in Geostationary lar spot at the equator. This is very useful, because such a satellite never goes
Orbit below the horizon and the antenna position with respect to the earth, once
found, never needs to be changed. In fact this orbit is so convenient that
until recently most communication satellites used it. It is still the method of
choice for point-to-point service and broadcasting.
The geostationary orbit does leave something to be desired for portable
and mobile use, however. Being able to point the antenna always in the same
direction is not as useful on a car or with a handheld device as it is for a fixed
installation. The orientation of the antenna with respect to the earth is sta-
ble, but the orientation of the antenna with respect to the vehicle must vary
rapidly as the vehicle changes direction.
Another problem is the extreme path loss at the distance needed for a
geostationary satellite. We have seen that if we require a geostationary or-
bit the altitude of 36,000 km, though greater than we would like, is set by the
laws of physics. The large path loss usually means that a high-gain
directional antenna is needed, and this is problematic for portable/mobile
operation.
One of the main reasons for using satellites for wireless communication
is to provide coverage in remote areas, which are out of reach of terrestrial
cellular systems. Many of these areas are in high latitudes (northern Canada
and Alaska, for example). Unfortunately, as the earth station moves toward
the poles, the geostationary satellites are closer to the horizon. This makes
it harder to achieve a direct line of sight to the satellite, which is free from
obstructions such as buildings or hills. Thus the geostationary satellites,
which must be above the equator, become less useful in extreme northern or
southern latitudes. Another glance at Figure 12.2 should make this clear.
Yet another problem, though less important than those already men-
tioned, is propagation time. We saw in Chapter 7 that the propagation delay
is about 0.25 s for a round trip to a geostationary satellite. This delay, while
not fatal, is annoying in real-time conversations. It also causes delays in data
460 ! CHAPTER 12
transmission whenever a protocol requires acknowledgment from the

receiving station before the transmission can continue.
Geostationary A geostationary satellite can be “seen” from almost half the earth’s surface.
Satellite Beams Therefore, three such satellites should be sufficient to cover the entire earth,
and Footprints except for the polar regions, with some overlap. A satellite designed for such
wide coverage has an antenna with a relatively large beamwidth called a
hemispheric beam. Because of its wide beam, such an antenna necessarily
has relatively low gain.
Many geostationary satellites are not intended to cover an entire hemi-
sphere. They use much more directional antennas, producing spot beams to
cover populated regions. Such antennas have higher gain and can deliver a
stronger signal on earth for the same transmitter power. Similarly, when
used for receiving, such antennas can achieve a better signal-to-noise ratio at
the satellite for a given transmitter power at the ground. This is very impor-
tant in portable and mobile communication where ground-station transmit-
ter power and antenna gain are strictly limited.
It is certainly possible for a satellite to have a combination of hemi-
spheric and spot beams in order to provide basic service over a wide area and
the ability to use smaller earth stations in particular areas.
When designing receiving installations on the ground, it is the EIRP of
a satellite and not simply its transmitter power that is important. EIRP
depends on the gain of the transmitting antenna in a particular direction, as
well as the transmitter power, so it is different at different points on earth.
Satellite operators publish maps showing the footprint of each geostation-
ary satellite on the earth. These show the effective EIRP (in dBW) for the sat-
ellite at each point on the earth where reception is possible. See Figure 12.4
for an example of a satellite footprint.
Use of Low- and Geostationary satellites are usable for wireless communication, but as we
Medium-Earth have seen, there are reasons to prefer satellites with lower orbits. The two
Orbits main problems with such satellites are (1) their position in space is not fixed
with respect to a ground station and (2) the annoying tendency of such satel-
lites to disappear below the horizon. (Another smaller problem is the Dopp-
ler effect, which causes frequencies to change. Transmitted frequencies are
shifted higher as the satellite approaches a point on the ground and lower as
the satellite recedes.)
The first problem is less important than it might seem. Shorter range re-
sults in much less propagation loss and removes the requirement for highly
directional antennas. This makes antenna tracking less critical. In any case,
if the antenna is mounted on a moving vehicle or person, the direction to
the satellite is constantly changing, even with geostationary satellites.
461
FIGURE 12.4 Footprint for Anik E-1 satellite (Courtesy of Telesat Canada)
462 ! CHAPTER 12
When real-time communication is required, the only way to address the

second problem is to use a constellation containing more than one satellite.
The closer the satellites are to earth, the more of them are needed in order
to have at least one satellite visible from a given point at all times. This
can make the system quite complex and expensive, even allowing for the
fact that it is less expensive to put satellites in lower orbits than in geosta-
tionary orbit.
The Doppler effect requires careful receiver design for both satellite and
ground stations, so that the receiver can lock onto an incoming signal and
track its frequency changes.
Satellite orbits are usually divided into three ranges. Low-earth-
orbit (LEO) satellites range from about 300 to 1500 km above the earth.
Medium-earth-orbit (MEO) satellites are about 8,000 to 20,000 km in alti-
tude. The gap between LEO and MEO orbits is there to avoid the lower of the
two Van Allen radiation belts that surround the earth; this radiation can
damage satellites. These radiation belts extend from 1,500 to 5,000, and
from 13,000 to 20,000 km above the earth’s surface. MEO satellites are typi-
cally near the lower end of the MEO range to avoid the upper Van Allen belt.
Finally, of course, there is the geostationary earth orbit (GEO) already men-
tioned. It would, of course, be possible to orbit satellites at still greater dis-
tances from the earth, but since the GEO is already farther from Earth than
we would like, there is no point in using more distant satellites for commu-
nicating between points on Earth. Figure 12.5 shows the LEO, MEO, and
GEO ranges, along with the Van Allen belts.
FIGURE 12.5
LEO, MEO and
GEO orbits
(not to scale)
' 12.4 Satellites and Transponders

The satellite as a spacecraft, with its attendant guidance systems and posi-
tioning jets, is outside the scope of this book. We are concerned, however,
with the satellite as a radio repeater. We need to know what the satellite
looks like from the ground.
The traditional way to build a communication satellite is to design it as a
frequency-shifting repeater or collection of repeaters (for some reason, a re-
peater on a satellite is called a transponder.) One satellite may have many
transponders, each of which has a block diagram that looks like Figure 12.6.
A range of frequencies is received from the ground via the uplink, amplified
and shifted in frequency, and retransmitted on the downlink. No signal pro-
cessing other than amplification and frequency shifting is done on the satel-
lite. For obvious reasons, this type of transponder has what is known as the
bent-pipe configuration.
FIGURE 12.6
Bent-pipe
transponder
This transponder design is extremely versatile. A bent-pipe transponder

can be used for anything from broadcast television using analog wideband
FM to digital telephony using either time-division or frequency-division
multiplexing or both. When used for one wideband FM signal, the satellite
power amplifier can be operated in a saturated mode, much like a Class C
amplifier, for greatest efficiency. If required to handle amplitude-varying
signals, the amplifier can be “backed off” to a lower-power linear mode by re-
mote control from the ground. Such a transponder is ready for just about
any signal, or combination of signals, that will fit into its bandwidth.
It is also possible to design satellite transponders for specific applications.
Some are designed to store digital information and retransmit it at a later time.
Using this store-and-forward technique, data can be communicated using a
low-earth orbit satellite that may not be visible to the transmitting and receiv-
ing stations at the same time. Also, satellites can be designed to communicate
with each other. Such crosslinks can improve the efficiency of communication
between earth stations, as indicated in Figure 12.7.
464 ! CHAPTER 12
FIGURE 12.7
Crosslinks
It is possible to turn a satellite transponder into a switching center so

that a signal from a transmitter on the ground can be relayed to one of a vari-
ety of ground- or satellite-based receivers as required. However, the gain in
efficiency is counterbalanced by an increase in satellite cost and complexity.
It should be remembered that it is much easier to build and maintain com-
plex equipment on the ground than in space. When we look at practical ex-
amples of wireless satellite communication later in this chapter, we’ll see
examples of both philosophies. Figure 12.8 shows a satellite with beam
switching.
FIGURE 12.8
Beam switching
' 12.5 Signal and Noise Calculations

Because of the very weak received signals, satellite systems require low-noise
receivers and relatively high-gain antennas for both the satellite and the
earth station, especially when geostationary satellites are used. This is a dif-
ferent situation from the terrestrial systems we studied earlier, where inter-
ference or distance to the horizon was more likely than thermal noise to
limit the range. This is, therefore, a good place to consider in more detail the
factors that determine signal-to-noise ratio.
The signal strength is determined by the factors noted in Chapter 7. The

important parameters are the EIRP of the transmitting installation, the dis-
tance to the receiver, and the gain of the receiving antenna, less any losses
between antenna and receiver. Free-space propagation can usually be
assumed in satellite communication, though for frequencies much above
10 GHz, losses in the atmosphere should be taken into account.
Noise As discussed in Chapter 1, the noise level is determined by the bandwidth of

Temperature the system and its equivalent noise temperature. To review,
P N = kTB (12.2)
where
PN = noise power in watts
T = system noise temperature in kelvins (K)
The system noise temperature is not necessarily the actual temperature
at which it operates as measured with a thermometer. It depends on the con-
tribution of the various electronic components, as well as any noise picked
up by the antenna or contributed by the feedline from antenna to receiver.
The total noise temperature of a system can be found by adding together
the noise temperatures of its various components, provided all are refer-
enced to the same point. Usually the receiver input is used as a reference
point. Receiver noise temperatures are specified at their input. Sometimes,
noise figure, rather than noise temperature is given. The noise figure is a mea-
sure of how much an electronic system degrades the signal-to-noise ratio of a
signal at its input; that is,
( S /N )i
NF = (12.3)
( S /N )o
where
(S/N) i = signal-to-noise ratio at the input
(S/N) o = signal-to-noise ratio at the output
The noise figure, expressed this way, is a dimensionless ratio. In practice,
it is always specified in dB, where
NF dB = 10 log NF (12.4)
It is simple to convert noise figure to equivalent noise temperature using
the equation:
T eq = 290(NF − 1) (12.5)
466 ! CHAPTER 12
where
T eq = equivalent noise temperature in kelvins

NF = noise figure as a ratio (not in dB)
EXAMPLE 12.2 Y
A receiver has an equivalent noise figure of 2 dB. Calculate its equivalent
noise temperature.
SOLUTION
First convert the noise figure to a ratio.
 NFdB 
NF = antilog  
 10 
 2
= antilog  
 10 
= 1.58
Now use Equation (12.5) to find the equivalent noise temperature:
Teq = 290(NF − 1)
= 290(1.58 − 1)
= 170 K
X
To find the equivalent noise temperature of the receiving installation,
we need to add the antenna noise temperature, as modified by the feedline,
to the receiver equivalent noise temperature. That is,
T = T eq + T a (12.6)
where
T = system noise temperature in K
T eq = receiver equivalent noise temperature in K
T a = antenna noise temperature in K
The antenna noise temperature results from thermal noise picked up
from objects in the beam of the antenna. This depends on the angle of eleva-
tion of the antenna: when the antenna beam includes the ground, the noise
level increases because of radiation from the ground itself. Luckily, this is
seldom the case with ground stations used with geostationary satellites,
except in very high latitudes where the satellite is just above the horizon.
This means that noise entering the antenna originates mainly from extrater-
restrial sources (stars, for instance) and from the atmosphere. Occasionally
the sun passes through the main lobe of the antenna pattern; the sun is a
very powerful noise source and makes communication impossible for the
few minutes it takes to pass through the antenna beam. Otherwise the sky
noise temperature for an earth-station receiving antenna is quite low, typi-
cally 20 K or less. The situation is very different for less directional antennas
such as those used with portable phones in LEO systems. The noise tempera-
ture of these antennas may be about the same as the ambient temperature
because of noise picked up from the surroundings; signals are much stronger
at these antennas as well.
Losses in the antenna system contribute to its noise temperature. The
noise temperature of an antenna system, at the far end of a feedline, is
given by
( L − 1)290 + Tsky
Ta = (12.7)
L
where
T a = effective noise temperature of antenna and feedline,
referenced to receiver antenna input, in kelvins
L = loss in feedline and antenna as a ratio of input to output
power (not in decibels)
T sky = effective sky temperature, in kelvins
EXAMPLE 12.3 Y
An earth station for use with a geostationary satellite has a dish antenna
which sees a sky temperature of 25 K. It is connected to the receiver with a
feedline having 1 dB loss. The receiver equivalent noise temperature is 15 K.
Calculate the noise temperature for the system.
SOLUTION
First convert the feedline loss to a power ratio:
L 
L = antilog  dB 
 10 
 1
= antilog  
 10 
= 1.26
468 ! CHAPTER 12
Now use Equation (12.7) to find the antenna noise temperature:
( L − 1)290 + Tsky
Ta =
L
0.26 × 290 + 25
=
1.26
= 80 K
Now add this to the receiver equivalent noise temperature to find the system
temperature, as in Equation (12.6).
T = Teq + Ta
= 15 + 80
= 95 K
X
Signal-to-Noise Once the system temperature is known, it is easy to calculate the noise
Ratio power in any given bandwidth.
EXAMPLE 12.4 Y
Calculate the noise power for the system in the previous example, if the
bandwidth is 2 MHz.
SOLUTION
P N = kTB
= 1.38 × 10 −23 × 95 × 2 × 10 6
= 2.62 fW
= −116 dBm
X
Now, if we know the signal strength, we can easily find the signal-
to-noise ratio. Usually the satellite power is specified as an EIRP in dBW, so
all we need is the path loss and antenna gain, less any feedline losses, to find
the received signal strength. We’ll review free-space path loss after we look at
another way to specify the noise performance of a receiving installation.
G/T A figure of merit called G/T has evolved to measure the combination of an-
tenna gain and equivalent noise temperature for a receiving installation.
G/T is defined as:
G/T(dB) = G R (dBi) − 10 log (T a + T eq ) (12.8)

where
G/T(dB) = figure of merit for the receiving system
G R (dBi) = receiving antenna gain in dBi
Ta = the noise temperature of the antenna
T eq = the equivalent noise temperature of the receiver
The gain and noise temperatures should all be taken at the same refer-
ence point. The gain required is the antenna gain less any losses up to the
reference point. As before, the reference point is usually the receiver input.
See Figure 12.9. In a satellite receiver, the first stage is often located sepa-
rately from the rest of the receiver and very close to the antenna. This stage is
called the low-noise amplifier (LNA).
FIGURE 12.9
Reference plane for
G/T calculations
The antenna noise temperature is found as before.
EXAMPLE 12.5 Y
A receiving antenna has a gain of 40 dBi and looks at a sky with a noise tem-
perature of 15 K. The loss between the antenna and the LNA input, due
to the feed horn, is 0.4 dB, and the LNA has a noise temperature of 40 K. Cal-
culate G/T.
SOLUTION
First, we find G in dB. This is simply the antenna gain less any losses up to
the reference point.
G = 40 dBi − 0.4 dB
= 39.6 dBi
470 ! CHAPTER 12
The calculation of T is a little more difficult. We have to convert the

feedhorn loss into a ratio, as follows:
 0.4 
L = antilog   = 1096
.
 10 
Substituting into Equation (12.7), we get
( L − 1)290 + Tsky
Ta =
L
(1096
. − 1)290 + 15
=
1096
.
= 39 K
The receiver noise temperature is given with respect to the chosen reference
point, so it can be used directly. Therefore,
G/T(dB) = GR(dBi) − 10 log (Ta + Teq)

= 39.6 − 10 log (39 + 40)
= 20.6 dB
X
Carrier-to-Noise Once G/T has been found, it can be used to help calculate the carrier-to-noise
Ratio Using G/T ratio for a system. We can use the following equation:
C
(dB ) = EIRP(dBW ) − Lƒ s (dB ) + G/T − k(dBW ) (12.9)
N
where
C/N(dB) = carrier-to-noise ratio in decibels
EIRP(dBW) = effective isotropic radiated power in dBW
L ƒs (dB) = free space loss in decibels
G/T = figure of merit as given in Equation (12.8)
k(dBW) = Boltzmann’s constant expressed in dBW
(−228.6 dBW)
The free-space loss for this application is as found in Chapter 7, that is
L ƒs = 32.44 + 20 log d + 20 log ƒ (12.10)

where
L ƒs = free space loss in decibels

d = path length in km
EXAMPLE 12.6 Y
The receiving installation whose G/T was found in Example 12.2 is used
as a ground terminal to receive a signal from a satellite at a distance of
38,000 km. The satellite has a transmitter power of 50 watts and an antenna
gain of 30 dBi. Assume losses between the satellite transmitter and its an-
tenna are negligible. The frequency is 12 GHz. Calculate the carrier-to-noise
ratio at the receiver.
SOLUTION
The earth station was found to have G/T = 20.6 dB. The satellite transmitter
power, in dBW, is
PT(dBW) = 10 log 50 = 17 dBW
The EIRP in dBW is just the transmitter power in dBW plus the antenna gain
in dBi, less any feedline losses, which are negligible here. Here, we have
EIRP(dBW) = 17 dBW + 30 dBi = 47 dBW
Next we find Lƒs(dB). Using Equation (12.10),
Lƒs = 32.44 + 20 log d + 20 log ƒ

= 32.44 + 20 log 38,000 + 20 log 12,000
= 205.6 dB
Now we can find C/N. From Equation (12.9),
C/N(dB) = EIRP(dBW) − L ƒs (dB) + G/T − k(dBW)

= 47 dBW − 205.6 dB + 20.6 dB + 228.6 dBW
= 90.6 dB
X
472 ! CHAPTER 12
' 12.6 Systems Using Geostationary Satellites

In spite of their disadvantages, the relative simplicity of geostationary sys-
tems has made them attractive for the first generation of mobile systems.
Global coverage can be achieved with only three GEO satellites, and all of
North and South America can be covered with one. Let us look first, then, at
two of the pioneers in satellite wireless communication. Both of them are
still very much alive, and both use geostationary satellites.
Inmarsat Inmarsat (International Maritime Satellite Organization) was established in

1979 as an intergovernmental treaty organization (it has since been privat-
ized) and began service in 1982. It is now into its third generation of satel-
lites, known as Inmarsat-3. Originally Inmarsat’s mandate was to provide
voice and data services to ships at sea, supplementing and eventually partly
supplanting high-frequency radio. Since then its services have expanded to
include land and aeronautical mobile communication.
Inmarsat uses a total of nine GEO satellites (four are in service, the rest
are spares or leased out), covering the whole world except for the polar
regions. Each satellite has a hemispherical beam and five spot beams with a
maximum EIRP of 48 dBW in the spot beams. Power and bandwidth can be
dynamically allocated among the beams. The satellites operate in the L-band
(1.5/1.6 GHz).
Inmarsat offers several different types of telephone services. Inmarsat-A
is the original analog telephone service. Inmarsat-B is a more modern digi-
tal service with similar capabilities. Both of these are intended mainly
for communication with ships and need dish antennas with diameters of
about 80 cm.
The Inmarsat mini-M service is more closely related to the theme of this
book. Using the satellite spot beams, it is designed mainly for operation
on land and in coastal waters. A typical portable transceiver, including
antenna, is about the size of a laptop computer. It cannot be carried about
when in use, but must be set up where its antenna can be aimed at one of
Inmarsat’s geostationary satellites. See Figure 12.10.
Handheld Inmarsat terminals are available, but only for text messaging
services. Inmarsat is a major partner in ICO, a proposed MEO system for use
with handheld phones.
MSAT (Mobile This joint Canadian and United States project uses one GEO satellite to pro-
Satellite) vide coverage for North and Central America, the Caribbean and Hawaii (via
a spot beam), and the surrounding coastal waters. See Figure 12.11 for the
satellite footprint.
FIGURE 12.10
Inmarsat portable
terminal
(Courtesy of Inmarsat)
FIGURE 12.11
MSAT footprint
MSAT’s satellite is about ten times more powerful than those used
by Inmarsat and has an EIRP of at least 57.3 dBW in its coverage area so an-
tennas can be smaller. Mobile terminals use a reasonably compact roof-
mounted antenna, and portable terminals are about the size of a notebook
computer and have a lid-mounted antenna. See Figure 12.12(a) for a typical
mobile installation. The geostationary system does not allow for portable
telephones that can be carried when in use.
The use of a single geostationary satellite allows MSAT to use a relatively
simple network. Only one ground station with an 11 m dish antenna is
needed. All calls are relayed through the satellite to the single ground station
and from there to the PSTN. See Figure 12.12(b).
474 ! CHAPTER 12
(a) Mobile antenna
(b) Portable terminal
FIGURE 12.12 MSAT terminals
Comparison of Table 12.1 shows a comparison of some of the important features of

Geostationary Inmarsat-3 and MSAT. Both use bent-pipe transponders connecting mobile
Systems users to a gateway that is in turn connected to the PSTN. Figure 12.13 shows
the relatively simple network that results.
TABLE 12.1 Geostationary Satellite Systems
System Inmarsat-3 MSAT
Date of operation 1996–1998 1996
Major uses and coverage Voice, data, especially for ships Voice, data, mainly for land mobile
Worldwide except polar regions Western hemisphere
Number of satellites in
4 1
active service
Uplink frequency (mobile

1.6 1.6315–1.6005
to satellite), GHz
Downlink frequency
1.5 1.53–1.559
(satellite to mobile), GHz
Uplink frequency (gateway

6.4 10.75–10.95
to satellite), GHz
Downlink frequency
3.6 13–13.15, 13.2–13.25
(satellite to gateway), GHz
Satellite EIRP, dBW 48 57.3
FIGURE 12.13 Mobile telephone network using geostationary satellites

476 ! CHAPTER 12
' 12.7 Systems Using Low-Earth-Orbit Satellites

LEO satellite systems are very attractive, especially for use with handheld
portable phones. The short distance to the satellite allows transmitter power
and antenna gain requirements to be relaxed. This permits the use of porta-
ble phones that are only somewhat larger than a conventional cell phone.
On the other hand, such a system requires many satellites (on the order of
40 to 70) and a complex network.
LEO systems are the most complex and expensive wireless communica-
tion systems yet devised. Several proposed systems have been cancelled
before becoming operational due to a shortage of funds. As of April 2000,
one LEO wireless telephony system (Globalstar) was operating, and another
(Iridium) had been forced to discontinue service after only a year in opera-
tion, due to bankruptcy. If their technical advantages can be coupled with
the necessary financial resources, LEO systems look like the ultimate in satel-
lite telephony.
Iridium Iridium began service in November 1998—and applied for bankruptcy pro-
tection in August 1999. As of April 2000, many attempts to restructure and
continue to operate the network had failed, all commercial operation had
ceased, and it seemed very likely that the satellites would be destroyed by de-
liberately taking them out of orbit. Nonetheless, a brief look at the Iridium
system will give useful insights into the possibilities of LEO satellite tele-
phony. Technically, the system was a success, and if the economic obstacles
could be overcome, a similar system could be built in the future.
The Iridium system comprises 66 LEO satellites in a complex constella-
tion, such that at least one satellite is visible from any location on earth at all
times. The satellites are crosslinked; telephone calls can traverse the network
from one satellite to another before being relayed to a ground station. This
means that it is not necessary for every satellite to be in range of a ground
station at all times and reduces the number of ground stations required. See
Figure 12.14 for an illustration of this idea.
Iridium uses digital modulation, with a combination of FDMA and
TDMA to assign channels.
Because Iridium’s satellites are powerful and close to the ground, porta-
ble phones are usable with this system. The phone weighs about half a kilo-
gram and has a fairly awkward helical antenna which must be positioned
vertically and with a clear view of the sky for reliable operation. For opera-
tion where terrestrial cellular radio is available, a cellular cassette with its own
antenna can be installed in the back of the phone. Although the Iridium
satellite system was available nearly everywhere (in a few countries local
FIGURE 12.14 Mobile telephone network using LEO satellite and crosslinks
authorities forbade it), users generally preferred to use their local cellular
system where possible to save on airtime costs. Airtime for Iridium ran
several dollars per minute, compared to a few cents per minute for most ter-
restrial cellular systems. The high cost of airtime, and of the phones them-
selves, may have been contributing factors to Iridiums’s financial problems.
Globalstar The Globalstar system began commercial operation in 1999, and by April
2000 service was available in more than 100 countries.
This system is slightly less ambitious than Iridium. It uses a constellation
of 48 LEO satellites (plus four spares) at an altitude of 1414 km. The satellites
use simple “bent pipe” transponders but have high power (about 1 kW per
satellite). The satellites are in eight orbital planes of six satellites each, in-
clined at 52 degrees with respect to the equator. This allows the system to
provide service from 70 degrees North latitude to 70 degrees South, which
includes most of the Earth except for the polar regions.
CDMA is used, allowing a ground user to access two or more satellites si-
multaneously, provided they are in view, and utilizes the soft handoff tech-
niques introduced for CDMA PCS services.
Because there is no switching on the satellites, communication is possi-
ble only when at least one satellite is visible simultaneously from the mobile
phone and a ground station. This will require at least 38 ground stations,
called gateways, for worldwide coverage. By April 2000 there were 18 gate-
ways in operation.
478 ! CHAPTER 12
Like Iridium, the Globalstar system is usable with handheld phones that
resemble cell phones, but are larger and heavier. Some dual-mode phones are
already available, so that Globalstar users can access lower-cost terrestrial
cellular radio where it is available.
Teledesic The Teledesic system, still under development, is the most ambitious of the
proposed LEO systems. When operational it is expected to use 288 satellites
plus spares, orbiting at an altitude of 1375 km. It is intended to be a
high-speed data service, designed more for fixed terminals in homes and
businesses than for mobile use. The frequency band it uses is much higher
than the other LEO services: 28.6–29.1 GHz for the uplink and 18.8–19.3
GHz for the downlink. Atmospheric losses are high at these frequencies. It is
necessary for earth stations to use satellites at a high angle, because signals
pass through less of the atmosphere en route to and from these satellites
than when satellites at lower elevation angles are used.
The main application for Teledesic is expected to be high-speed commer-
cial data, in competition with fiber optics, and for individual use,
high-speed internet access. Here Teledesic will have competition from cable
modems, telephone lines using ADSL, and existing satellite access using geo-
stationary satellites. Standard terminals are expected to support data rates
from 16 kb/s (for a single voice channel) to 2.048 Mb/s (for high-speed data).
Table 12.2 provides a comparison of the three “Big LEO” systems just
described.
“Little LEOs” In addition to the huge projects sometimes referred to as “big LEOs,” there
are a number of more modest schemes, both existing and proposed, that ex-
ist only to provide low-data-rate digital services such as paging, short
messaging, and vehicle tracking for trucking companies. These operations
are called “little LEOs” because the systems are simpler and smaller. It is not
necessary that there be a satellite in view of all stations at all times for these
services because messages can be stored briefly and forwarded when a satel-
lite becomes available (typically within a few minutes). Here are some exam-
ples of “little LEO” systems.
ORBCOMM The ORBCOMM system went into operation in 1998. It uses 28

LEO satellites and plans to increase this to 48. The system is used for short
messaging, e-mail, and vehicle tracking. Unlike the big LEOs, little LEOs typ-
ically use frequencies in the VHF range to communicate with customer earth
stations. ORBCOMM’s uplink from mobile to satellite is at 148–150.05 MHz,
with downlink at 137–138 MHz.
TABLE 12.2 “Big LEO” Systems
System Iridium Globalstar Teledesic
Date of operation 1998 1999 2000
Major uses Voice, paging, low-speed Voice, paging, low-speed High-speed data,
data data voice
Number of satellites 66 48 288
Satellite altitude (km) 780 1414 1375
Uplink frequency (mobile

1.616–1.6265 2.4835–2.500 28.6–29.1
to satellite), GHz
Downlink frequency 1.616–1.6265 1.610–1.6265 18.8–19.3

(satellite to mobile), GHz (same range as
uplink, TDM used)
Uplink frequency (gateway

29.1–29.3 5.025–5.225 27.6–28.4
to satellite), GHz
Downlink frequency
19.4–19.6 6.875–7.075 17.8–18.6
(satellite to gateway), GHz
Intersatellite crosslink
23.18–23.38 (No crosslinks) 65–71
frequency, GHz
LEO One LEO one is a proposed “little LEO” system with a similar structure
to that of ORBCOMM. It will be designed to use 48 satellites at an altitude
of 950 km for paging and short messaging. It is expected to be operational
in 2002.
E-Sat E-Sat, expected to begin this year (2000), is an interesting “spe-

cial case” LEO system. Using only six satellites, orbiting at an altitude of
1260 km, and using CDMA, it is intended to concentrate on the niche mar-
ket of remote meter reading, especially in rural areas. Since meter readings
are not urgent, store-and-forward technology is very appropriate for this
type of system.
Table 12.3 summarizes the specifications of the little LEO satellite sys-
tems discussed in this chapter.
480 ! CHAPTER 12
TABLE 12.3 “Little LEO” Systems
System ORBCOMM LEO One E-Sat
Date of operation 1998 2002 2000
Major uses Paging, short messaging/ Paging, short messaging/ Remote meter
e-mail, vehicle location e-mail, vehicle location reading
Number of satellites 28 48 6
Satellite altitude (km) 785 950 1260
Uplink frequency
(mobile to satellite), 148–150.05 148–150.05 148–148.905
MHz
Downlink frequency
(satellite to mobile), 137–138, 400 137–138 137.0725–137.9725
MHz
Uplink frequency
(gateway to satellite), 149.61 148–150.05 148–148.905
MHz
Downlink frequency
(satellite to gateway), 137–138 400.15–401 137.0725–137.9725
MHz
' 12.8 Systems Using Medium-Earth-Orbit Satellites

Satellites in medium earth orbit are a compromise between the LEO and GEO
systems. More satellites are needed than for GEO (on the order of 6 to 20 for
real-time communication), but fewer than for LEO. Delay and propagation
loss are much less than for GEO, but greater than LEO. Portable phones are
possible with MEO systems, but they must be heavier and bulkier than for
LEO systems.
The main advantage of using MEO rather than LEO satellites is financial.
These systems promise rates for airtime that are at least on the same order of
those for terrestrial cellular systems, unlike LEO systems.
At this printing no MEO systems are up and running, but some proposed
systems appear likely to become operational in the near future.
Ellipso Ellipso uses an interesting combination of elliptical and circular orbits. Its
constellation is based on the fact that there is far more land mass and far
greater population in high northern latitudes than in similar southern lati-
tudes. A glance at any globe will confirm that most of the world’s land mass
is north of 40° south latitude.
The Ellipso system is designed to take advantage of this asymmetry. It
will initially include six satellites, later increasing to ten, in a circular orbit
about 8000 km above the equator for worldwide coverage at low latitudes.
These are to be complemented by eight active satellites (plus two spares) in
inclined elliptical orbits, designed so that they spend most of their time
above the northern hemisphere. The elliptical orbits have a maximum
height of approximately 7800 km and a minimum height of approximately
520 km. The elliptical-orbit satellites will have an orbital period of about
three hours. See Figure 12.15.
FIGURE 12.15
Ellipso system
orbits
Ellipso’s coverage is scheduled to be phased in gradually, beginning with

the deployment of the circular-orbit satellites in 2002. These satellites can
provide coverage to about 80% of the world’s population.
The main focus of the Ellipso system is expected to be voice communica-
tion using portable and mobile terminals that are projected to be roughly
the size of conventional cell phones but are likely to be considerably larger.
CDMA is used, with the uplinks of all satellites receiving on the same fre-
quency bands. This allows use of the soft-handoff feature of CDMA and
avoids any necessity for portable phones to switch frequencies. The satellites
will relay signals directly to ground-station gateways, with no on-satellite
processing or intersatellite links.
ICO ICO stands for Intermediate Circular Orbit. The plan, initiated by Inmarsat but
since spun off and privatized, is to launch ten operational satellites and two
spares in two orthogonal planes at an altitude of 10,355 km, each at a 45°
482 ! CHAPTER 12
angle to the equator, thus providing global coverage. In fact, most of the
time a given location will be in view of two satellites to provide diversity in
case the view of one is blocked (by a tall building, for instance). Each satellite
would be able to support 4,500 telephone calls using TDMA technology. The
satellites are to use a simple bent-pipe configuration. The frequency ranges
are projected to be similar to those for Ellipso, in the area of 1.6 GHz for
satellite-to-mobile links.
( Satellites are especially useful for telecommunication in remote areas

where terrestrial cellular systems are prohibitively expensive or impossi-
ble to build.
( The orbital period of a satellite in a circular orbit depends on its distance
from the earth, with satellites farther from the earth having a longer period.
( Satellite orbits are classified according to their distance from earth, as low
earth orbit (LEO), medium earth orbit (MEO), and geostationary earth
orbit (GEO). Geostationary satellites appear stationary at a point above
the equator.
( Systems using satellites in lower orbits have lower path loss and shorter
propagation times, but require more satellites for real-time coverage.
( Satellites can act as simple repeaters or can contain elaborate switching
systems to route calls. They can also store data for later forwarding.
( Current satellite systems for mobile communication use either GEO or
LEO satellites. Only the latter systems allow handheld transceivers. There
are MEO systems in the planning and construction stages.
( Equation List
4 × 1011
v = (12.1)
(d + 6400)
PN = kTB (12.2)
( S /N )i
NF = (12.3)
( S /N )o
NFdB = 10 log NF (12.4)

Teq = 290(NF − 1) (12.5)
T = Teq + Ta (12.6)
( L − 1)290 + Tsky
Ta = (12.7)
L
G/T(dB) = GR(dBi) − 10 log (Ta + Teq) (12.8)
C
(dB ) = EIRP(dBW ) − Lƒ s (dB ) + G/T − k(dBW ) (12.9)
N
Lƒs = 32.44 + 20 log d + 20 log ƒ (12.10)
( Key Terms
apogee point in a satellite orbit that is farthest from the earth
bent-pipe configuration a design of satellite transponder in which
signals are amplified and shifted in frequency but do not undergo any
other processing
constellation in satellite telephony, a group of satellites coordinated in
such a way as to provide continuous communication
crosslink a radio or optical connection directly between satellites,
without going through an earth station
equatorial orbit a satellite orbit that is entirely above the equator
footprint depiction of the signal strength contours from a satellite
transmitter on the earth
geostationary orbit satellite orbit in which the satellite appears to
remain stationary at a point above the equator
geosynchronous orbit satellite orbit in which the satellite’s period of
revolution is equal to the period of rotation of the earth
hemispheric beam antenna beam on a geostationary satellite that is
adjusted to cover the whole earth
low-earth-orbit (LEO) satellite an artificial satellite orbiting the earth at
an altitude less than about 1500 kilometers
medium-earth-orbit (MEO) satellite a satellite in orbit at a distance
above the earth’s surface of approximately 8,000 to 20,000 km
perigee the point in a satellite orbit that is closest to the earth
polar orbit a satellite orbit passing over the north and south poles
484 ! CHAPTER 12
spot beam in a satellite system, a focused beam of energy that covers a

relatively small area on the earth, produced by a high-gain antenna
on the satellite
store-and-forward technique in digital communication, the use of a
device (repeater) to receive one or more data packets, store them, and
retransmit them at a later time
tracking in satellite communication, continuously adjusting the position
of a directional antenna on the ground, so that it always points at the
satellite
transponder in satellite communication, a repeater located on a satellite
( Questions
1. Compare the following in terms of cost and practical communication
distance:
(a) repeaters on towers
(b) satellites in low earth orbit
(c) geostationary satellites
2. How does the orbital period of a satellite change as it moves farther
from the earth?
3. Sketch the earth and the orbit of a geostationary satellite, approxi-
mately to scale.
4. Why do all geostationary satellites orbit the earth at the same distance
and above the equator?
5. Why are geostationary satellites unusable from earth stations in the
polar regions?
6. What are the Van Allen belts and what effect do they have on the place-
ment of satellites?
7. Explain the advantages of the use of elliptical orbits for satellite com-
munication.
8. Why are spot beams used with geostationary satellites?
9. How many geostationary satellites are necessary for a system to have
worldwide coverage (except in the polar regions)?
10. What is Doppler shift? Why does the effect increase as the height of a
satellite’s orbit decreases?
11. Why is it necessary for mobile systems to use antenna tracking, even
with geostationary satellites?
12. Why is it necessary to use multiple satellites for real-time coverage with
LEO and MEO systems?
13. How can the use of multiple satellites be avoided for data communica-
tion with MEO and LEO satellites?
14. What advantages and disadvantages do satellites in low earth orbit
have compared with geostationary satellites for mobile communication
systems?
15. What is meant by a “bent-pipe” satellite transponder?
16. What is meant by the term backoff with respect to a transponder power
amplifier, and when is it necessary to use it?
17. What are crosslinks and how are they useful in a satellite communica-
tion system?
18. What satellite communication system is especially designed for use by
ships at sea?
19. Compare the Inmarsat mini-M and MSAT services in terms of coverage
and convenience.
20. Compare Iridium with MSAT in terms of coverage and convenience.
21. Compare Iridium with Globalstar in terms of the way the networks are
organized.
22. How does the Teledesic system differ from the other LEO systems de-
scribed in this chapter?
23. What are the differences between “big LEO” and “little LEO” systems?
24. Why are store-and-forward techniques unsuitable for voice communi-
cation?
25. Explain how the Ellipso system can achieve worldwide coverage with
fewer satellites than the other systems discussed in this chapter.
26. Compare the LEO and MEO concepts for voice communication. What
advantages does each have?
( Problems
1. Find the orbital velocity and period for a satellite that is 1000 km above
the earth’s surface.
2. The moon orbits the earth with a period of approximately 28 days. How
far away is it?
3. What velocity would a satellite need to have to orbit just above the sur-
face of the earth? Why would such an orbit be impossible in practice?
486 ! CHAPTER 12
4. A receiving installation has a G/T of 30 dB. If its antenna gain is 50 dBi

and the antenna noise temperature is 25 K, what is the equivalent noise
temperature of the receiver?
5. How could the G/T of the system described in the preceding problem be
improved to 35 dB? Give two ways, and perform calculations for each.
6. Find the noise temperature of a receiver with a noise figure of 1 dB.
7. Find the noise temperature of an antenna on a satellite if it looks at the
earth (giving it a “sky” noise temperature of 290 K) and is coupled to the
reference plane by a waveguide with a loss of 0.3 dB.
8. Compare the time delay for a signal transmitted by geostationary satel-
lite to that for a signal transmitted via terrestrial microwave radio link
over a distance of 2000 km. Assume the distance from ground to satel-
lite is 37,000 km at both ends of the path. Ignore any time delay in the
electronics for both systems.
9. Compare the signal strengths from two satellites as received on the
ground. One is geostationary, with a path length of 40,000 km; the other is
in low earth orbit, with a path length of 500 km. Assume all other factors
are equal. Which signal is stronger, and by how many decibels?
10. It is possible to communicate between points on earth by using the
moon as a passive reflector. In fact, it is done routinely by radio ama-
teurs. Calculate the time delay and round-trip path loss at 1 GHz using
the moon as a reflector. (The distance to the moon is calculated in prob-
lem 2.) Note that your path loss calculation ignores losses in the reflec-
tion at the moon’s surface. What can you conclude about the likely
commercial possibilities of this idea?
11. (a) Calculate the path loss for a satellite in the Iridium system on the
link from satellite to mobile, assuming that the satellite is directly
overhead.
(b) Calculate the round-trip time delay for the Iridium system under
the same conditions as in part (a).
12. Repeat the previous problem, but use the ICO MEO system. Compare
the results and draw some conclusions about the differences between
LEO and MEO systems.
13. Calculate the velocity and orbital period for each of the following types
of satellite:
(a) Iridium
(b) Globalstar
(c) Teledesic
(d) ORBCOMM
Paging and Wireless
Data Networking
13
Objectives
( Describe and explain the operation of several systems used for one- and
two-way paging.
( Compare paging systems with respect to capabilities and complexity.
( Describe the operation of voice paging systems.
( Describe the operation of wired Ethernet LANs.
( Explain the need for wireless LAN equipment.
( Discuss the IEEE 802.11 and Bluetooth standards and suggest which would
be preferred for given applications.
( Explain the need for and the operation of wireless Ethernet bridges and
modems.
( Describe the operation of infrared LANs and compare them to wired LANs
and wireless LANs using radio.
( Describe and compare public packet-data networks and compare them with
other kinds of wireless data communication.
488 ! CHAPTER 13
' 13.1 Introduction

In this chapter we look at some forms of wireless communication that are
perhaps less glamorous than PCS and satellite telephony but are nonetheless
important. Pagers have been with us for many years, and many people pre-
dicted that cellular and PCS phones would make them obsolete. That hasn’t
happened; in fact, paging features have been built into personal communi-
cation systems. Meanwhile, paging systems have become more capable.
Wireless modems and LANs have also been around for a good many
years but have not been very popular outside niche markets until recently.
The current vogue for notebook and subnotebook computers and for con-
nected organizers like the Palm Pilot™ has spurred interest in these technol-
ogies as well. In this chapter we will look at some of the emerging standards
in this field.
' 13.2 Paging and Messaging Systems

Paging systems have undergone changes not unlike those that have occurred
with wireless voice communication. The traditional paging system uses
widely spaced transmitters, each covering a considerable geographic area. In
addition, all transmitters in a given system operate on the same frequency,
either in the VHF range at about 152 or 158 MHz or in the UHF range from
454.025 to 454.650 MHz, and all pages are sent by all the transmitters in the
system. The system, wasteful of spectrum though it is, works because short
paging messages require little time or bandwidth.
The traditional pager is simply a fixed-tuned receiver that uses a trans-
mitted code to identify messages meant for it. The earliest pagers simply
beeped when they received a message. That signaled the pager owner to call
a central office to find out the details of the page. Most modern systems
transmit a short numeric or alphanumeric message to the pager. The nu-
meric message is usually a phone number for the owner to call, but some in-
ventive pager owners, particularly teenagers, have invented numerical codes
for standard messages.
Simple paging systems like these remain very popular in spite of the in-
creased popularity of cellular and PCS telephones. The advantages of pagers
are their small size, low cost, and long battery life (typically several weeks,
compared to hours for cell phones and days for PCS phones). The extended
battery life of one-way pagers results partly from the lack of any need for a
transmitter or for power-consuming vocoders and audio amplifiers in the re-
ceiver, and partly from the ability of pagers to go into a “sleep” mode, emerg-
ing at intervals to check for messages.
PAGING AND WIRELESS DATA NETWORKING ! 489
The ultimate in paging is for the user to be able to send a reply. Such
systems exist and we’ll look at them too. Some systems even allow a voice
message to be received and answered.
One-Way Paging The traditional way to handle paging is to have a network of relatively
Systems powerful transmitters (on the order of hundreds of watts), all of which trans-
mit all pages on the same frequency. Figure 13.1 shows how this works.
Superficially the system may resemble a cellular network, but in reality the
technology is much simpler. Frequencies are reused by using the same fre-
quencies for every transmitter. With a voice system this would give a very
low capacity, since one call would tie up a pair of frequencies over the whole
network. Paging messages, however, typically require only a few seconds to
transmit and are relatively infrequent, so many pagers can share a frequency
using TDMA. Each pager has a unique address called a capcode, which is
sent every time it is paged. Transmissions addressed to other pagers are
simply ignored.
The fact that all pages are sent from all transmitters means that there is
no need for the system to know the location of any pager. In fact, since the
pagers are receive-only devices with no transmission capability, there is no
way for the system to tell where an individual pager is or even whether it is
turned on. The use of multiple transmitters has the added advantages of
reducing the number of dead spots for the signal and completely avoiding
any need for handoffs.
FIGURE 13.1 Basic paging network

490 ! CHAPTER 13
When a paging system covers a very wide area, such as the United States
or Europe, roaming systems may be used in which the pager owner informs
the system when he or she moves away from the local network. This removes
the need to send all pages over a very large area. Satellites are often used to
transmit pages great distances to local systems, which relay them via terres-
trial transmitters. In addition, as we discussed in the previous chapter, there
are LEO satellite systems that send paging messages directly to individual
pagers. These systems are more expensive but are useful in remote areas
without terrestrial paging systems. See Figure 13.2 for an illustration of the
difference between these two types of satellite-based paging.
FIGURE 13.2
Satellite-based
paging systems
The most common protocol for one-way paging systems is called

POCSAG, which is an acronym for Post Office Code Standardization Advisory
Group; it is so named because it was originated by the British Post Office.
It can be transmitted at three different data rates: 512, 1200, and 2400 b/s.
Currently the 1200 b/s variation is the most common; the 512 b/s system
is gradually being phased out, and the 2400 b/s system is used mainly for
nationwide systems. The modulation scheme is FSK with ±4-kHz deviation,
in an RF channel that is 25 kHz wide. The protocol can handle about two
million pager addresses. This seems like a lot but remember, even in a
nationwide system, each pager must have a unique capcode.
POCSAG messages are sent in batches. Several batches may be sent to-
gether in one transmission. See Figure 13.3 for an illustration of a POCSAG
signal. Each transmission begins with a preamble in the form of 576 bits of
the sequence 1010101010. Each batch begins with a 32-bit synchronizing
pattern followed by eight frames with two codewords per frame. Each code-
word can be defined as an address or a message codeword. These are easily
distinguished: the address codeword begins with a 0 and the message code-
word begins with a 1. The address codeword has an 18-bit address field. Fol-
lowing are two bits that allow for four different paging sources, and ten
error-correction bits.
The number of possible addresses that can be transmitted in one frame is
218 = 262,144 addresses. This is not enough for a large system. To increase the
number of possible addresses, each pager is programmed with a 3-bit code
that tells it in which of the eight frames in the batch it should look for its ad-
dress. This gives a total possible number of addresses equal to 221 = 2,097,152
addresses.
FIGURE 13.3
POCSAG paging
protocol
492 ! CHAPTER 13
Codewords in a batch that are not used for addresses are used for mes-
sages. (If there are not enough messages to fill a batch, then an idle message
is sent.) There are 20 message bits per codeword, and as many codewords
may be used as necessary to transmit the message. Simple tone pagers do not
need any message frames; an address is sufficient. Alphanumeric pagers
require more message frames than do the simpler numeric pagers.
EXAMPLE 13.1 Y
Suppose the POCSAG system is used with simple tone pagers, which require
only an address field. If all the frames are used for addresses, how many
pages could be transmitted by this system in one minute if it operates at the
slowest POCSAG rate of 512 b/s? Assume that only one preamble is needed.
SOLUTION
The preamble uses 576 bits. To find how many bits remain in one minute we
can find the total number of bits and subtract 576. Each frame can contain
two address codewords. Let us assume that things are balanced and all
codewords can be used. That is, one-eighth of the pagers being paged have
addresses in each of the eight possible codewords. This is obviously a
best-case assumption.
Total number of usable bits/minute = 512 × 60 − 576 = 30,144 bits
Each batch has one 32-bit synchronizing codeword and sixteen 32-bit
address codewords for a total of 17 × 32 = 544 bits. Therefore,
bits/min
batches/min =
bits/batch
30144
,
=
544
= 55.412 batches/min
Each batch has 16 addresses so the number of pages that can be transmitted
in one minute is
addresses/min = batches/min × addresses/batch

= 55.412 × 16
= 886
X
A newer family of paging protocols, introduced by Motorola, is the

FLEX™ group, consisting of FLEX for one-way paging, ReFLEX™ for two-way
alphanumeric paging, and InFLEXion™ for voice paging.
The FLEX system uses four-level FSK to allow operation at 1600, 3200,
and 6400 b/s. There are four possible transmit frequencies rather than two
as for conventional FSK, which is used by POCSAG. The possible frequency
deviations from the carrier frequency are ±800 Hz and ±2400 Hz. This
provides four possibilities so that two information bits can be transmitted
per symbol, allowing a data rate of 6400 b/s with a symbol rate of 3200 baud
in the same 25-kHz channel as is used for a POCSAG system.
EXAMPLE 13.2 Y
Calculate the efficiency, in terms of bits per second per hertz of RF band-
width, for the FLEX system at its maximum data rate.
SOLUTION
The channel width is 25 kHz and the maximum bit rate is 6400 b/s. This
represents
6400 b/s
= 0.256 b/s/Hz
25 kHz
X
This is a rather low figure, indicating a robust system. In fact, this system
can operate in much less bandwidth. The two-way ReFLEX system, described
next, can operate with three paging channels in one 50-kHz RF channel.
Two-Way Paging The Motorola ReFLEX™ system is the de facto standard for two-way alpha-
Systems numeric pagers. Two-way paging is much more complex and expensive than
the more common one-way operation. Every pager needs a transmitter, with
a power output on the order of one watt; this adds greatly to the cost of these
devices.
On the other hand, there are advantages in addition to the obvious
one of being able to reply to pages without having to find a phone. The pres-
ence of a transmitter allows pagers to acknowledge the receipt of messages,
making for more reliable service. It is also possible for a two-way paging sys-
tem to employ frequency reuse, much like a cellular or PCS telephone
system. Since the pagers have transmitters, they can respond to queries from
the system and can be located within a particular cell.
494 ! CHAPTER 13
ReFLEX paging systems operate in the frequency ranges of 928–932 and

940–941 MHz for the outbound channel (base to pager) and 896–902 MHz
for the inbound channel (pager to base). One outbound carrier can be car-
ried in a 25-kHz channel, or three carriers in a 50-kHz channel. The available
outbound data rates are 1600, 3200, and 6400 b/s. Inbound communication
uses a 12.5-kHz channel and can support up to a 9600 b/s data rate.
Voice Paging Voice paging has been available for many years but was not widely used until
recently. It allows a pager to function in a manner similar to a telephone
answering machine.
The Motorola InFLEXion™ system is the most popular voice paging pro-
tocol. It uses analog compression and SSB AM to transmit voice messages
from the base station to the pagers. Both upper and lower sidebands are
used, but each sideband constitutes a separate voice channel. This system
allows two voice messages to be transmitted simultaneously on a channel
6.25 kHz wide. A pilot carrier is also transmitted for ease in demodulation.
See Figure 13.4. Time-division multiplexing is also used so that many pagers
can share a single voice channel.
FIGURE 13.4
InFLEXion voice
paging signal
InFLEXion voice pagers normally allow a text (not voice) reply in a man-
ner similar to two-way alphanumeric pagers. In that respect they are less
flexible than cellular or PCS telephones. Since voice paging is fairly expen-
sive and the price of cellular and PCS phones is dropping, it remains to be
seen how popular voice paging will become.
' 13.3 Wireless Local-Area Networks

Many offices and even some homes use local-area networks to connect mi-
crocomputers. Traditionally, these networks have used coaxial cable,
twisted-pair wire, or in some cases fiber optics, to connect the computers to
each other, usually through a central hub. This is quite suitable for installa-
tions where the computers are fixed in place and where wiring can be run
without too much difficulty.
Some offices with staff who bring notebook computers to meetings or
who move about regularly find wireless network connections useful. There is
also an emerging market in home networking. The problem in the home is
not so much mobility as the inability or reluctance of homeowners and
tenants to run wiring through or along the walls. Wireless networks seem to
be much easier to install—and in some cases that may even be true.
LAN Topologies There are three basic ways to organize a local-area network. As shown in Fig-
ure 13.5, they are the star, the ring, and the bus. The star is the traditional
way to connect a central mainframe to dumb terminals. The ring topology is
used with some LANs, notably those using the IBM token-ring system. By far
FIGURE 13.5 LAN topologies

496 ! CHAPTER 13
the most common logical topology for LANs, however, is the bus. The major-
ity of local-area networks use the Ethernet standard, a logical bus topology
usually running at either 10 or 100 Mb/s (now up to 1 Gb/s). Ethernet can
use either coaxial cable or twisted-pair wiring, but the latter is cheaper and
more commonly seen nowadays. In this chapter we shall focus on networks
using Ethernet, because practically all wireless networking equipment is
designed to connect to Ethernet systems.
With a basic Ethernet system, all stations are connected to the bus at all
times, and any station can transmit data packets at any time. Each packet has
an address and so does each station; the station to which a packet is addressed
will accept it and read the data, while all other stations simply ignore it.
Any bus network has the possibility of collisions; an occurrence where
two stations transmit at the same time, thereby rendering the data from
both unusable. Ethernet systems try to minimize collisions with a set of
simple rules. Each station “listens” before transmitting to make sure the
bus is not being used. If it is, the station waits until all is quiet. In the event
that two stations do transmit at the same time, each will detect this,
stop transmitting immediately, and wait a random time before beginning
the transmission again. This is a type of protocol known as carrier-sense
multiple-access with collision detection (CSMA/CD).
The physical layout of an Ethernet LAN using twisted-pair wiring usually
looks more like a star than a bus. All the computers are connected to a cen-
tral hub, as shown in Figure 13.6, using two pairs of wire in a single cable.
Appearances are deceiving, however. Logically the network is still a bus,
and the central hub simply makes wiring easier and less expensive. Hubs can
usually do some network management, such as disconnecting a defective
station. It is also possible to use a switch instead of a hub at the central loca-
tion, which can make the network more efficient by avoiding collisions. The
address of each packet is examined as it arrives at the switch, and the packet
FIGURE 13.6
Ethernet network
using twisted-pair
wiring
is sent only to the node to which it is addressed, eliminating the possibility

of a collision.
The alternative, for 10 Mb/s networks only, is to use a true bus using 50 Ω
RG-58/U coaxial cable. This is sometimes done, but it has disadvantages.
Coaxial cable is more expensive than twisted-pair wiring, and connect-
ing a station to the network requires breaking the cable and inserting a tee
connector. See Figure 13.7.
FIGURE 13.7
Ethernet using
coaxial cable
There is a third type of Ethernet, which uses thicker RG-8/U coaxial

cable and transceivers, but this is seen so seldom that it will not be covered
here.
Ethernet systems using twisted-pair wiring are called 10BaseT (for 10 Mb/s,
baseband, twisted pair) or 100BaseT for the 100 Mb/s version. Practically all
wireless LAN equipment is designed to connect to this type of network.
Ethernet networks have restrictions on the distances between terminals.
The reason for this limitation can be found in the CSMA/CD protocol. Before
transmitting, a node “listens” for a period of 9.6 microseconds. If another
station is transmitting at the beginning of this period, it will be detected by
the end of the listening period, provided that its distance from the receiving
station is less than the distance a signal can travel during the listening pe-
riod. Recall that the speed of transmission along a line can be found by mul-
tiplying a number called the velocity factor, which is a characteristic of the
cable, by the speed of light.
The velocity factor for typical coaxial cable is about 0.66, so the distance
traveled in 9.6 µs is given by:
d = vt
= v ƒ ct
= 0.66 × 3 × 10 8 m/s × 9.6 µs
= 1900 m
498 ! CHAPTER 13
where
d = distance traveled
v = velocity of propagation along the cable
t = propagation time
vƒ = velocity factor
c = speed of light
It would not be wise to have an Ethernet segment this long, however. If a
station begins transmitting during the listening period for another station,
it might not be detected. The Ethernet specifications call for maximum cable
runs of 500 m on thick coax, 200 m on thin coax, and 100 m on twisted-pair.
Of course, it is still possible that another station will begin transmitting
while the first station is listening. If the station is too far away, the transmis-
sion will not be detected and a collision may result. In that case, the collision
should be detected relatively early in the transmission, so that the process
can be stopped and restarted with as little wasted time as possible. The sys-
tem is most efficient when the packets of data are relatively long.
To determine how long it can take to detect a collision, consider the
scenario shown in Figure 13.8 on page 499. Nodes A and B are at opposite
ends of the bus. Suppose node A transmits. The signal from node A moves
down the line toward node B, but it takes some time to get there. Meanwhile,
node B listens and, hearing nothing, begins to transmit just as the signal
from A reaches it. Node B stops transmitting, but both transmissions now
have errors and have to be discarded. Meanwhile, node A knows nothing
about the collision until the signal from node B reaches it. Then it stops
transmitting and the network is clear for one or the other to try again.
EXAMPLE 13.3 Y
Two nodes on an Ethernet LAN using coaxial cable are 200 m apart. How
long will it take the stations to detect a collision?
SOLUTION
The total time used by this process is the time it takes for a signal to travel
down the line and back. That is, it is two times the propagation delay along
the line. Recall that for any line,
v = vƒc
The time for a signal to travel a given distance along the cable is given by
d
t =
v
FIGURE 13.8
Ethernet distance
restrictions
For an Ethernet system with a length of 200 m on coaxial cable with a

velocity factor of 0.66, the propagation delay is:
d
t =
v
200 m
=
0.66 × 300 × 106 m/s
= 1 µs
500 ! CHAPTER 13
The maximum time it takes to detect a collision is twice this, or 2 µs. At

10 Mb/s, this corresponds to the time taken by 2 µs × 10 Mb/s = 20 bits.
X
Ethernet packets are generally much longer than this. Long packets can ex-
tend the distances possible on the net, but they reduce the network efficiency
when there are collisions, since the whole packet must be retransmitted.
Wireless connections to LANs have timing constraints similar to those
seen with wired LANs. Of course, the velocity factor is equal to one for a
wireless connection. On the other hand, since wireless LAN transceivers can-
not receive while they are transmitting, a collision cannot be detected until
the end of the packet. This makes shorter packets more desirable with wire-
less systems.
Ethernet Bridges When a network becomes too large to satisfy the distance constraints just de-
scribed, it can be broken into two sections connected by a bridge, as shown
in Figure 13.9. A bridge connects to both network segments. It looks at each
packet of data on each segment of the network, and from its address, deter-
mines whether or not its destination is on the same segment as its source. If
the source and destination are on the same segment, it does nothing. If they
are on different segments, it places the packet on the segment holding the
destination. This is called selective forwarding.
FIGURE 13.9
Ethernet bridge
Wireless LANs In general, wireless local-area networks are slower and more expensive than
wired LANs. Therefore, the use of wireless LANs is indicated only where
there are compelling advantages to being free of cabling. Offices where peo-
ple frequently change work stations, work locations like warehouses or hos-
pitals where people have a need to communicate while moving about, and
places where it is inconvenient or expensive to run wires, are examples. Or-
dinary wired Ethernet networks commonly run at either 10 or 100 Mb/s, and
gigabit Ethernet is now available. By contrast, wireless LAN speeds tend to
be in the 1- to 2-Mb/s range. This is fast enough for many applications, how-
ever, particularly if the application program runs locally and the network is
used only to transfer data. For instance, a spreadsheet program running on a
notebook computer can usually share files with others on the network with-
out requiring a great deal of bandwidth.
Radio LANs There has been a proliferation of proprietary radio standards for LANs. Most
have used unlicensed frequency bands in the 900-MHz range, and more re-
cently the 2.4-GHz range. However, there was no standardization until very
recently. This meant that network components from different manufactur-
ers would not work together. Wired Ethernet, by contrast, has quite rigid
standards: any network adapter designed for a 10BaseT system will work
with any other adapter made for the same system.
Recently, however, there has been some improvement in this chaotic
situation. Two relatively new standards are particularly interesting. The
first that we will consider, called IEEE 802.11, is a general-purpose standard
for LANs that cover a typical office environment. The second, with the inter-
esting name of Bluetooth, is intended for much shorter range, up to 10 m,
and is intended to provide a low-cost, built-in capability for devices such
as cellular phones and personal digital assistants (PDAs). Let us look at each
in turn.
IEEE 802.11 The Institute of Electrical and Electronic Engineers worked for many years to
bring some order to the chaotic wireless LAN environment, finally approv-
ing the 802.11 standard in 1998. It envisions spread-spectrum operation on
the unlicensed ISM (industrial, scientific, and medical) frequency band from
2.4 to 2.484 GHz. This is the same part of the spectrum that is used for micro-
wave ovens; the use of spread-spectrum transmission allows these networks
to operate in the presence of interference, both from microwave ovens
and from other nearby LANs. The 2.4-GHz band was chosen for the 802.11
standard primarily because, unlike other unlicensed frequency ranges such
502 ! CHAPTER 13
as 902–928 MHz, it is available in most of the world and until recently has
been largely unused because of the interference problem from microwave
ovens. The standard also allows for infrared operation, but this provision is
seldom used.
Both the direct-sequence and frequency-hopping variations of spread-
spectrum operation can be used. The following statements establish a few
basics on which the standard is built:
( A set of wireless nodes is called a Basic Service Set (BSS), and a network
consisting of only a BSS with no access points is clled an ad-hoc net-
work.
( A network can consist only of wireless nodes communicating with
each other—an Independent Basic Service Set (IBSS)—but more usually a
number of wireless nodes in a BSS communicate with an access point
that is also part of a wired Ethernet network.
( There can be multiple access points for extended coverage.
( A network with multiple access points is called an Extended Service Set
(ESS).
( Wireless units can roam within the ESS. See Figure 13.10 on page 503.
The 802.11 system specifies a CSMA/CA (Carrier-Sense Multiple-
Access with Collision Avoidance) protocol. The main difference between
this and the CSMA/CD system used with wired Ethernet is that the radios are
only half-duplex, so they cannot detect collisions once they have begun to
transmit a packet of data. This can reduce the network capacity by increasing
the time taken to detect collisions.
The choice between frequency-hopping and direct-sequence spread-
spectrum versions of the specification is not clear-cut, and both types of
products are available. In current practical implementations, the direct-
sequence version offers greater range, but the frequency-hopping system
often seems to perform better in cluttered offices with many reflections and
sources of interference.
The frequency-hopping version of 802.11 uses 79 separate channels,
1 MHz apart and visited at a rate of approximately ten channels per second.
The maximum transmitter power allowed is 1 W, and devices have to be able
to reduce this to not more than 100 mW when the full power level is not
needed. When systems are interference-limited, as is often the case in LANs,
there is no benefit in increased power: it only shortens battery life in porta-
ble equipment.
The basic frequency-hopping standard uses frequency-shift keying with
two transmitted frequencies. The bit rate is 1 Mb/s. There is an optional stan-
dard, using four-level frequency-shift keying (two different shifts, much like
the FLEX paging system that we looked at earlier in this chapter). With the
same symbol rate of 1 Mbaud, this system achieves a transmitted bit rate of
2 Mb/s. The idea behind the two speed standards is to allow an intermix of
lower-cost lower-speed and higher-cost higher-speed devices.
The standard does not envisage any central control over hopping se-
quences to prevent two transmitters from transmitting at once. The protocol
will prevent this from happening at the start of a transmission, but it can
happen later. There are 78 predefined hopping patterns organized into three
FIGURE 13.10 IEEE 802.11 networks

504 ! CHAPTER 13
sets of 26 patterns each. In spite of collisions, the data throughout the whole
system keeps increasing until there are about 15 networks operating in the
same area.
The direct-sequence version of the specification is quite different, as you
would expect. The bit rates allowed are the same: 1 and 2 Mb/s. The way this
is achieved is different, however. Either binary (for 1 Mb/s) or quadrature
delta phase-shift keying (DBPSK or DQPSK) can be used. In either case, the
symbol rate is 1 Mbaud. See Figure 13.11.
FIGURE 13.11
DBPSK and DQPSK Phase change from
Bit previous symbol
0 0°
1 180°
Phase change from

Bits previous symbol
00 0°
01 90°
10 −90°
11 180°
The symbol rate is increased to 11 Mbaud by using ten extra bits, called
chips, for each data bit. The result is an 11:1 spreading gain. That is, the
bandwidth is increased by a factor of 11. The strength of an interfering sig-
nal will also be reduced by a factor of 11 (10.4 dB) during the de-spreading
process.
Even with the spreading due to the chipping bits, the signal from a sin-
gle direct-sequence wireless node occupies only about 5 MHz. Therefore,
several frequency channels can be assigned; this system combines FDMA
with CDMA. The exact number of channels varies with the country, since
the width of the band allocated varies somewhat. In North America there are
11 available channels.
Many companies have introduced wireless LAN equipment using the

802.11 standard. Most of the wireless nodes consist of network cards that
are intended to be installed in laptop and notebook computers. Access
points are more likely to be stand-alone boxes. In both cases there is a small
antenna that should be oriented for best results. Examples are shown in Fig-
ure 13.12.
FIGURE 13.12
Wireless Ethernet
devices
(Courtesy of Lucent
Technologies)
The distance obtainable with a wireless Ethernet connection varies

greatly with the environment. As we saw when we discussed propagation,
the square-law rule for attenuation does not apply once we leave a free-space
environment. The typical office is anything but free space, being full of ob-
jects of varying conductivity as well as quite a few sources of electrical noise.
Usable ranges are therefore very difficult to predict. Typical systems work
over a distance of about 100 to 200 m in an office environment.
IEEE 802.11—a and b Versions

( The IEEE 802.11b standard was introduced in September 1999. It al-
lows for a maximum bit rate of 11 Mb/s using direct-sequence
spread-spectrum operation in the 2.4-GHz band. Equipment designed
to this standard can interoperate with the1-Mb/s and 2-Mb/s equip-
ment previously described.
( The IEEE 802.11a project is designed as a future standard for data rates
of up to 54 Mb/s in the 5-GHz band. It is still being developed.
506 ! CHAPTER 13
Bluetooth The Bluetooth specification is a joint venture involving several companies:

Ericsson, IBM, Intel, Nokia, and Toshiba. It is designed to be an open stan-
dard for short-range systems; the usable operating range is specified as 10 cm
to 10 m and can be extended to 100 m using RF amplifiers for the transmit-
ters. The intention is to make Bluetooth devices small enough and inexpen-
sive enough that they can be built into many types of equipment: cellular
and PCS phones; notebook computers and personal digital assistants (PDAs);
and computer peripherals such as printers, modems, and loudspeakers. Any
device using the Bluetooth standard should be able to communicate with
any other such device. For instance, at present it is necessary to connect a
notebook computer to a PCS phone for wireless data by using a proprietary
cable. If both were equipped with Bluetooth transceivers it would only
be necessary to have them reasonably close together to make a truly wire-
less connection. The portable phone could even be left, powered on, in a
briefcase.
Technically, the Bluetooth standard resembles the frequency hopping
version of the IEEE 802.11 wireless Ethernet standard. Like the LAN system,
Bluetooth operates in the 2.4-GHz ISM band. The reason is the same as cited
by the IEEE: this band is available worldwide, though the exact frequency
limits vary somewhat from place to place.
Bluetooth radios employ a rather simple spread-spectrum technique us-
ing frequency hopping. Channels are 1 MHz apart, giving room for 79 chan-
nels in North America and most of Europe. The transmitter uses two-level
FSK, with a frequency deviation (measured each way from the carrier fre-
quency) of between 140 kHz and 175 kHz. The channel bit rate is 1 Mb/s, and
since this is a two-level system, the bit rate and baud rate are equal.
Bluetooth systems divide data into packets and send one packet per hop.
The length of the packet (and the corresponding hop) can be from one to
five slots; the length of a slot is 625 µs.
EXAMPLE 13.4 Y
Calculate the maximum and minimum hopping rate for the Bluetooth
system.
SOLUTION
The maximum hopping rate is
1
ƒ h (max) =
625 µs
= 1600 Hz
and the minimum is
1
ƒ h (min) =
5 × 625 µs
= 320 Hz
X
Bluetooth differs from the wireless Ethernet systems in that it envisions

both audio and data transmission. Generally the audio is synchronous and
the data asynchronous. Audio is coded at 64 kb/s per stream, using either
linear CVSD (continuously variable slope delta modulation) or companded
PCM. The companding can follow either the North American µ-law or the
European A-law standard, in order to be compatible with wireline tele-
phony. The Bluetooth system can carry up to three of these 64 kb/s streams
in each direction.
When used for data, the system can support 432 kb/s in each direction
simultaneously, or alternatively, 721 kb/s in one direction and 57.6 kb/s in
the other. The latter alternative is useful for asymmetrical applications like
file downloads, where most of the data flows one way with the other direc-
tion used only for acknowledgements.
Full-duplex operation is accomplished by switching from transmit to re-
ceive between slots. This is called TDD (time-division duplex) and is simi-
lar to the procedure used in the TDMA and GSM cellular systems and PCS.
The simplest Bluetooth networks, called piconets, can have from two
to eight nodes. All Bluetooth nodes have the same capabilities, but which-
ever node initiates the communication acts as the master for timing and
frequency-hopping information, with the others acting as slaves.
One Bluetooth device can operate simultaneously on two piconets, act-
ing as a bridge between the two. A conglomeration of two or more piconets is
called a scatternet. See Figure 13.13 on page 508 for a depiction of Bluetooth
piconets and scatternets.
When a Bluetooth device is in standby mode and not connected to a net-
work, it listens every 1.28 seconds on a set of 32 hop frequencies, going
through each in turn. When a station wishes to connect, it sends a Page
message (if it already knows the network address of the intended recipient
from previous communication) or an Inquiry message (if it does not know
what other devices are on the network). Page and Inquiry messages are first
sent to 16 channels, then to the other 16 wake-up channels if there is no re-
sponse. Thus, it can take up to 2 × 1.28 = 2.56 seconds for a connection to
be set up.
Bluetooth can be used to set up a wireless LAN or to connect to a wired
LAN, just as with the IEEE 802.11 standard. The main differences are that the
508 ! CHAPTER 13
FIGURE 13.13
Bluetooth
networks
data rate and range are less with Bluetooth. On the other hand, once compa-
nies tool up for production, the cost of adding Bluetooth capability to a
piece of equipment is expected to be only a few dollars. This compares with
typical wireless Ethernet costs of several hundred dollars per node.
Wireless Bridges Wireless bridges connect LAN segments. Often these are in different build-
ings, so it is necessary to use a system with more range than those described
above. Some bridges use high-speed dedicated microwave links. These can
be fast but they are expensive and require licensing. Lower-cost bridges us-
ing the 902-MHz or 2.4-GHz unlicensed bands are available from several
vendors. A typical example is shown in Figure 13.14. It can operate over a
distance of about 10 km, under good line-of-sight conditions, at a data rate
FIGURE 13.14
Wireless bridge
(Courtesy of Wi-LAN Inc.)
of about 2 Mb/s. Line-of-sight propagation conditions are more likely

between buildings than within them. The built-in antenna shown can be
replaced with a directional rooftop antenna for greater range.
Connections Though the great majority of the wireless systems described in this book use
Using Infrared radio communication, we should note that light waves can also be em-
ployed. You are probably familiar with fiber optics; it is not really a wireless
technology because the fiber cable, though it may not contain copper, cer-
tainly has the effect of restricting mobility. Obviously it is possible to trans-
mit light directly through the air, however, without benefit of fiber.
Generally invisible infrared light is used. The common television remote
control is an example of a simple infrared communication system.
Anyone who has used an infrared remote control is familiar with some of
the limitations of optical systems. The range tends to be short, the signal will
not pass through walls or other opaque objects, and usually the transmitter
must be carefully pointed in the direction of the receiver. All of these issues
tend to limit the usefulness of infrared communication systems. We should
note in passing, though, that the inability of the light signal to pass through
walls can be an advantage in some cases, as it reduces the danger of deliber-
ate eavesdropping and of accidental interference between systems.
510 ! CHAPTER 13
A short-range infrared system called Infrared Data Association (IRDA) has

been used for some time to allow two devices to communicate with each
other. A typical example would be the synchronization of data between a
desktop and a notebook computer. The system is deliberately restricted to
a range of one meter to allow several independent infrared links to operate
simultaneously in the same room. LEDs operating at a wavelength of about
860 nm are used. Several data rates are specified, up to a maximum of 4 Mb/s,
but most systems operate at a significantly lower data rate, since it is com-
mon to connect an infrared transceiver to the serial port on computers with-
out built-in infrared ports.
At least one commercial system does take advantage of the 4-Mb/s data
rate possible with built-in infrared ports. Clarinet Systems makes the
EthIRLAN™, which uses a wired hub with infrared transceivers connected to
it by wires. The hub is connected to a portable computer by placing one of its
transceivers near the infrared port on the computer. What results is similar
to a wired LAN except that portable computers can be added to and removed
from the LAN at will without disturbing any electrical connections. See
Figure 13.15.
FIGURE 13.15
Infrared connections to LAN
(Courtesy of Clarinet Systems
“EthIRLAN” System, © 1999 Clarinet
Systems. All rights reserved.)
It is also possible to build a complete wireless LAN with relatively widely

separated computers. Each computer needs an infrared transceiver, which
typically connects to a standard Ethernet card. A central transceiver con-
nects to the network. In this case it is better to use diffused infrared light
than a narrow beam. Diffused infrared does not need a direct line of sight:
reflections from walls and ceiling can be used. This allows the computers to
be moved around, but they must, of course, remain in the same room.
Figure 13.16 shows one of these systems. It has a range of about 10 m.
Several computers with infrared transceivers can form a simple peer-to-peer
network. If one transceiver is connected to an existing Ethernet LAN, all of
the computers with transceivers become part of that network.
FIGURE 13.16
Infrared LAN
(Courtesy of iRLan Ltd.)
Systems like this, while they work, have failed to become very popular.
Perhaps this is because they are rather cumbersome for use with portable de-
vices, and they are much more expensive than wired systems for desktop
computers. If a network is confined to one room, in most cases it is not diffi-
cult to install the wire for a conventional Ethernet system.
Wireless Modems In a sense all wireless data communication devices are modems, since the
data must be modulated onto a carrier for transmission. This section deals
with point-to-point, rather than networked, data transfer at distances of
up to a few kilometers: larger distances than are found in a typical LAN,
but within line of sight. For distances greater than line of sight, a wireless
modem can be used as a repeater, or the user can resort to commercial
packet-radio, paging, cellular, or PCS networks.
A typical wireless modem is shown in Figure 13.17. It can work with one
other modem, or with several by polling each in turn. The data rate is fairly
low, with a maximum of 19.2 kb/s, but this is sufficient for many applica-
tions. Such devices are used for traffic control, process monitoring (at oil
wells, for instance), and similar applications.
512 ! CHAPTER 13
FIGURE 13.17
Wireless modem
(Courtesy of Wi-LAN Inc.)
Many wireless modems use the 2.4-GHz band; most of the others use an-
other ISM unlicensed band from 902 to 928 MHz. Most use spread-spectrum
operation to reduce the possibility of interference. Since the frequency
ranges where they operate are unlicensed, there is no coordination of fre-
quencies.
Most wireless modems operate at transmitter power levels similar to the
wireless Ethernet systems previously described, but they use much lower
speeds. These systems can therefore work over greater distances, up to sev-
eral kilometers in open terrain.
Wireless modems are similar in operation to conventional telephone
modems and quite different from wireless LAN systems. Typically they
attach to a computer’s serial port and operate with any ordinary modem
software package.
' 13.4 Wireless Packet-Data Services

So far we have looked at a number of wireless data communication schemes,
each with its own niche. Wireless LANs give high-speed short-distance con-
nectivity; wireless bridges and modems can extend proprietary systems for a
few kilometers. Paging systems give local, national, and even international
(using satellite systems) access to short text messages.
Note that there is some overlap. PCS have short-messaging features that
are very similar to those offered by paging systems, for instance. Personal
communication systems are also introducing packet-radio services, as we
noted earlier. These duplicate the functions of CDPD (which we studied ear-
lier) and also of dedicated wireless packet-data networks (which we shall
look at now). The competition between services is vigorous, and it is impos-
sible to tell at this writing which services will prosper, which will find niche
markets, and which will disappear.
Public packet-data networks predate PCS and even cellular telephony.
They are intended strictly for data transfer, which can include such jobs as
vehicle tracking, remote order entry, and even internet access.
There are two major public wireless packet-data networks in North
America, Mobitex™ and ARDIS™. There are also a number of private net-
works. Keep in mind that both these services compete with CDPD, analog
cell phones using modems, and PCS messaging.
Mobitex The Mobitex standard for wireless packet-switched data was created by the
Swedish telephone company for use by its field personnel. It was introduced
as a commercial system in 1986 and is now developed by Ericsson, a Swedish
manufacturer of wireless equipment, which also manufacturers most of the
equipment for this system. The specifications are public, though, so there
are other suppliers as well. In the United States, RAM Mobile Data, a subsid-
iary of BellSouth, is the main user of Mobitex; in Canada, Cantel AT&T oper-
ates a Mobitex network. Between them these systems cover most North
American cities.
Mobitex networks use a cellular structure in the 900-MHz band (896–902
MHz and 935–941 MHz) in North America (frequencies differ in other coun-
tries), but they are separate from the AMPS cellular system—except that the
two systems sometimes use the same towers. Narrow channels only 12.5 kHz
wide are used with GMSK (Gaussian minimum shift keying) modulation to
support a data rate of 8 kb/s. Since this is a packet-switched system, there
is no need for connections or multiplexing. Data is sent by landline to a
Mobitex transmitter, which transmits packets as received. Each mobile ter-
minal receives all packets, but ignores all except those addressed to it. Simi-
larly, the mobile transmits packets to the base station when it is available
using a contention protocol similar to that used with Ethernet.
Being a packet-switched system, Mobitex can at best be called “near
real-time.” This is not good enough for audio or video, but is quite satisfac-
tory for e-mail, database access, dispatching, vehicle tracking, web surfing,
and many other activities.
514 ! CHAPTER 13
ARDIS The Advanced Radio Data Information Services (ARDIS) system was created by
IBM as a joint venture with Motorola for its own use in working with its out-
side sales and service personnel. It is now owned by the American Mobile
Satellite Corporation. It is available in major metropolitan areas in the
United States, and also in Canada, where the Bell Mobility data network uses
the ARDIS system.
ARDIS differs from Mobitex in that all its cells use the same frequencies,
much like a paging system. This is wasteful of bandwidth but it does allow
for diversity in that a mobile unit can receive signals from, and transmit to,
more than one base station.
The normal data rate with ARDIS is 4.8 kb/s in a 25-kHz channel; there is
a high-speed version with a data rate of 19.2 kb/s but having limited avail-
ability. ARDIS systems compete directly with Mobitex for the portable fax,
e-mail, and tracking market.
( There are many options for wireless data communication, including pag-
ing systems, cellular and PCS radio, wireless LANs and modems, and
packet-data networks.
( There are many different types of paging systems, including simple
one-way beepers, one-way numeric, one-way alphanumeric, two-way
alphanumeric, and voice pagers.
( One-way paging systems transmit all pages throughout the whole system
to avoid having to locate the recipient.
( Two-way paging systems can locate pagers in much the same way as a
cellular radio system.
( The most common type of voice pager, the Motorola InFLEXion system,
uses analog compressed voice in an SSB AM system.
( Most wireless LAN systems work as extensions to wired Ethernet systems.
( Most current wireless LAN equipment follows the IEEE 802.11 standard,
and uses either frequency-hopping or direct-sequence spread-spectrum
radio in the 2.4-GHz unlicensed frequency band.
( The Bluetooth standard resembles IEEE 802.11, but is modified for
low-cost short-range operation. It uses frequency hopping at 2.4 GHz.
( Many wireless LAN bridges and wireless modems use proprietary systems
at 902 MHz or 2.4 GHz in order to achieve ranges of several kilometers.
Some long-range bridges use microwave links that require licenses.
( Infrared light is common for very short-range connections and is occa-

sionally used for wireless LANs.
( Public packet-data networks are used for e-mail, keeping contact with em-
ployees in the field, limited worldwide web browsing, and similar
low-data-rate applications.
( Key Terms
bridge a device to connect two segments of a network
capcode unique address for a pager
carrier-sense multiple-access with collision avoidance (CSMA/CA) method
of reducing contention in a network, involving each station checking
for interference before transmitting
carrier-sense multiple-access with collision detection (CSMA/CD) method
of reducing contention in a network, involving each station checking
continuously for interference before and during transmissions
collision attempt by two transmitters to use the same channel
simultaneously
Ethernet form of local-area network using CSMA/CD and a logical bus
structure
inbound channel communication channel from mobile to base station
outbound channel a radio channel used for communication from a base
station to mobile stations
time-division duplexing (TDD) transmission of data in two directions on
a channel by using different time slots for each direction
( Questions
1. Why is it necessary for one-way paging systems to transmit all messages
from all transmitters?
2. What factors account for the long battery life of pagers compared with
cell phones?
3. What modulation schemes and data rates are used with the POCSAG
protocol?
4. What limits the number of pagers in a POCSAG system, and approxi-
mately what is the maximum number of pagers that can be used in such
a system?
516 ! CHAPTER 13
5. How does the POCSAG system accommodate the differing mes-

sage-length requirements of tone-only, numeric, and alphanumeric
pagers?
6. Compare POCSAG with FLEX in terms of data rate and modulation
scheme.
7. How does the ReFLEX system differ from FLEX?
8. What advantages does a two-way paging system have over a one-way
system?
9. How does the InFLEXion system transmit voice?
10. What are the three topologies that are used with local-area networks?
11. What logical topology is used with Ethernet?
12. What physical topology is used in a typical Ethernet network? Is there
any other possible physical topology?
13. What transmission speeds are commonly used with wired Ethernet?
14. Does a practical wired Ethernet network always attain the data rates
specified in your answer to the previous question? Explain.
15. What limits the distance between nodes in a wired Ethernet?
16. Describe the function of an Ethernet bridge.
17. How can using a switch instead of a hub improve the performance of an
Ethernet LAN?
18. What frequency range is used for the IEEE 802.11 and Bluetooth wire-
less network standards, and why?
19. Why is it necessary to use spread-spectrum technology in the frequency
range used by IEEE 802.11 and Bluetooth?
20. What data rates are possible with IEEE 802.11(b) wireless LANs?
21. Explain the difference between a Basic Service Set, an Independent Ser-
vice Set, and an Extended Service Set in the IEEE 802.11 specification.
22. What two types of modulation are used for the frequency-hopping ver-
sion of 802.11? Explain how they differ.
23. Define the two terms CSMA/CD and CSMA/CA, describe how they dif-
fer, and state which is used in wireless LANs, and why.
24. What modulation schemes are used for the direct-sequence version
of 802.11?
25. How is the spreading of the spectrum accomplished in each of the two
RF variants of 802.11?
26. What is the spreading gain in the direct-sequence variant of 802.11?
27. What is the anticipated range for devices using the Bluetooth system?
28. What modulation scheme and spread-spectrum technique are used for
Bluetooth?
29. What is the maximum bit rate for the Bluetooth system?
30. How is audio transmitted using Bluetooth?
31. What is a piconet? How does it differ from a scatternet?
32. What are the advantages and disadvantages of infrared wireless LAN
systems, compared with radio-frequency systems?
33. What is the maximum range for the IRDA system? Why is it designed to
have such a short range?
34. Why is diffused infrared preferred for an optical LAN in a situation
where computers may be moved?
35. How does a wireless bridge differ from a wireless Ethernet card for a
computer?
36. How does a wireless modem differ from a wireless bridge?
37. Compare the Mobitex and ARDIS systems in terms of data rate.
( Problems
1. Calculate the efficiency of the POCSAG paging system in terms of
bits/second per hertz of RF bandwidth at each of its three specified data
rates.
2. Suppose a POCSAG system is set up so that each frame contains one
address codeword and one message codeword. A certain type of page
requires 500 message bits. Allowing for all the other types of bits in-
volved, how many frames and how much time will it take to transmit
this message at 1200 b/s? Assume that all the required frames can be
transmitted in one transmission with one preamble.
3. Suppose a POCSAG system at 1200 b/s is replaced with a FLEX system at
6400 b/s. Give an estimate of the increase in capacity. (The answer will
not be precise because the two protocols differ in the amount of redun-
dancy they include.)
4. Suppose that instead of four-level FSK the FLEX system used eight-level
FSK. What would be the maximum bit rate, assuming the baud rate re-
mained unchanged? Would there be any disadvantage to doing this?
5. Calculate the efficiency in bits per second per hertz of bandwidth of the
ReFLEX system operating at its maximum rate with three paging chan-
nels in one 50-kHz RF channel.
518 ! CHAPTER 13
6. Suppose two stations on a wireless LAN are 150 m apart. How long does
it take one station to detect that another is transmitting?
7. What is the spreading gain for a wireless LAN system using the fre-
quency-hopping version of the 802.11 system?
8. Suppose the Bluetooth system is used to connect stereo headphones to a
computer. What data rate could be supported for each of the two chan-
nels?
9. Prepare a table comparing the data rates delivered by the Mobitex
and ARDIS systems described in this chapter with the CDPD, cellular
modem, and PCS data transmission systems previously discussed.
10. Which of the systems described in the previous problem uses frequency
spectrum most efficiently?
The Future of Wireless 14
Objectives
( Describe and compare wireless alternatives to the copper local loop.
( Explain the process by which standards for the third generation of wireless
communication are being developed.
( Compare third-generation wireless proposals.
( Explain the third-generation requirements and compare them with systems
currently available.
( Explain the difference between MMDS and LMDS and discuss their
contribution to the total wireless picture.
( Explain what is meant by convergence and discuss the place of wireless
communication in the future.
( Discuss safety and esthetic factors that could affect the progress of wireless
communication.
( Make some predictions of your own about the future of wireless
communication.
520 ! CHAPTER 14
' 14.1 Introduction

There is no doubt that we are still near the beginning of the wireless revolu-
tion. The cellular concept started the revolution by allowing frequency reuse
at relatively short distances. As higher frequencies become economical for
everyday use, there is almost no limit to the amount of data that can be sent
by wireless means.
At the same time, the more traditional wired media have not remained
static. The development of high-speed fiber-optic networks, as well as new
technologies that increase the capacity of existing twisted-pair and coaxial
cables, have implications for wireless systems as well. In this chapter we look
at some emerging wireless technologies, then consider the interactions
among wireless and wired technologies. Finally, the author will go out on a
very thin limb and make some rash predictions about the future. New devel-
opments in the wireless field are announced almost daily, so this chapter
cannot hope to be exhaustive. Rather, we will try to find some pattern to cur-
rent and possible future developments, so that the reader can make sense of
a very confusing area of technology.
' 14.2 Wireless Local Loop

The “last mile” of any telephone system is the most costly. Running
twisted-pair wire to every house is expensive even in urban settings, but in a
remote area the costs and difficulties are much greater.
Wireless local loops using radio transmission in the VHF and UHF bands
have been employed in isolated areas for some time, because they can trans-
mit signals for several kilometers without the need for stringing cable. Until
recently, however, this was definitely a niche market. These systems use the
traditional method of building one tall tower and using it to get as much
range as possible without frequency reuse. One or sometimes two subscrib-
ers use a dedicated RF channel. The amount of spectrum allocated to this
type of service is not sufficient to allow wireless local-loop service in any but
the most sparsely populated areas.
Figure 14.1 shows a traditional wireless local-loop setup. Modulation is
analog FM and transmitter power is 1 to 10 W, depending on the distance to
be covered.
Recent developments in wireless technology, especially the cellular con-
cept and spread-spectrum communication, have caused the idea of wireless lo-
cal loops to be revisited. Several companies offer fixed wireless systems based
on this modern technology. See Figure 14.2(a) for a block diagram of this type
THE FUTURE OF WIRELESS ! 521
FIGURE 14.1
FM wireless
local loop
of system, with examples. Figure 14.2(b) shows a commercial system using

CDMA in a dedicated frequency band. The subscriber terminal provides the
proper signals, including analog voice, dial tones, and ringing voltage, to work
with standard phones and fax machines. The system uses QPSK on dedicated
RF channels in the range from 1.35 to 3.6 GHz. Each RF channel can carry 30
voice channels at the full landline-standard 64 kb/s. This standard has the ad-
vantage of allowing the lines to work with 56-kb/s modems, but it is also possi-
ble to operate the system at a reduced bit rate of 32 or 16 kb/s per voice channel
to achieve greater capacity.
FIGURE 14.2(a) Digital wireless local loop

522 ! CHAPTER 14
FIGURE 14.2(b) Commercial system using CDMA

There has also been considerable interest in using established PCS net-
works to provide fixed service. The most obvious way to do this is simply to
use a PCS phone at home, instead of a wired phone, or perhaps instead of
getting an additional phone line. However, several companies provide inter-
face units that will allow ordinary telephones to connect through existing
house wiring to a PCS transceiver.
Wireless local loops are still not very popular in areas which already have
telephone wiring in place, but they are attractive for new installations (par-
ticularly in developing countries that lack extensive copper infrastructure).
One of the advantages of a wireless system in this kind of environment is
that it is possible to begin slowly, with a small number of base stations, and
increase the number of cells as the number of customers grows. A wired sys-
tem, on the other hand, must be fully installed before any revenue comes in.
This can cause severe cash-flow problems in a relatively undeveloped area
where only a few people are likely to become customers at first.
In developed countries, far-sighted thinkers expect that eventually the
wired local loop will be replaced for telephony, not by a fixed wireless sys-
tem, but by a variation of PCS that is flexible enough, and cheap enough, to
be used everywhere for both voice and data. However, such a system might
well involve tiny cells (picocells) in users’ homes. These would have to be
connected to the PCS network somehow, so the copper local loop might well
survive in a new guise—as part of a PCS network rather than part of the wired
PSTN. The next section deals with developments in PCS that could lead to
this situation.
' 14.3 Third Generation PCS

Cellular radio and PCS have been very successful, but they are not the ulti-
mate in wireless personal communication. Manufacturers, service providers
and standards bodies were already working on improvements when the first
systems went into operation. At this time, third-generation (3G) wireless
systems are about to be deployed.
Requirements There are several areas in which improvements to the present systems are
for the Third needed. The following paragraphs provide a brief description of some of
Generation them.
Improved Data Communication All the digital personal communication

systems we have looked at so far were designed mainly for voice communica-
tion. The emphasis was on keeping bandwidth requirements low, along with
524 ! CHAPTER 14
minimizing power requirements for the mobile unit. We have seen that
these systems also work well for e-mail and other short messages, but the
available bit rate for data communication is quite low, even compared with
ordinary telephone modems. The reasons for this are easily understood, es-
pecially once we realize that these systems transmit digitized voice at much
lower bit rates than does the PSTN. The recent development of PCS phones
with internet browsing capabilities, and of notebook computers and per-
sonal digital assistants that can connect to a portable phone or can have a
cellular or PCS radio built in for web browsing, only underscores the need for
a higher data rate for serious data communication. A great deal of work has
gone into specialized browsers and web sites, all with the aim of minimizing
the amount of data that must be transmitted over the air interface. Never-
theless, users who are used to connecting to the internet over office LANs or
from home using cable modems or ADSL telephone-line connections still
find wireless access painfully slow. Other high-speed data applications, such
as streaming multimedia or video telephony, are simply impossible with cur-
rent systems.
Greater Capacity For some time it has seemed obvious that the next gener-
ation of wireless communication must incorporate the possibility of us-
ing much higher data rates when needed. On the other hand, there is no
need to use higher rates for ordinary telephone-quality voice; indeed, using
higher rates than necessary merely wastes spectrum. It also wastes battery
power in the mobile, since higher data rates need greater bandwidth and
therefore greater transmitter power for an equivalent signal-to-noise ratio at
the receiver. In fact, the trend in vocoder design is toward lower bit rates for
telephone-quality voice. Therefore, the third generation should have a vari-
able data rate, using whatever bit rate is necessary for the application.
One of the reasons for the creation of second-generation (PCS) wireless
systems was that the first-generation cell phone systems were becoming
overloaded in major metropolitan areas. At current growth rates, it will not
be long before the same thing happens to the second-generation networks
(it is already happening in Japan). Therefore, a third-generation wireless net-
work should have greater capacity for ordinary voice calls, as well as allow-
ing higher data rates for digital communication.
Adaptability to Mobile, Pedestrian, and Fixed Operation For some time

there has been considerable interest in making it possible to use the same
wireless phone in a variety of situations. We have already noted, that even
though the AMPS cellular system was really designed with mobile use (from
vehicles) in mind, at the present time there are more portable (hand-carried)
than mobile AMPS phones in use. The second-generation systems were
designed for portable phones and incorporate more extensive use of micro-
cells and picocells (very small cells in office buildings, shopping malls, and
so on.) For the third generation it is expected that in addition to vehicle
and pedestrian use some people will use their wireless phones from fixed
locations, such as a home or office. The requirements and the difficulties
in meeting these requirements differ in each situation. There is probably
less need for high-speed data in a mobile than in a fixed environment, for
instance. This is actually rather fortunate, since the frequent handoffs
required when a vehicle moves at high speed make high-speed data connec-
tions difficult.
The idea is to have a Universal Personal Telecommunications (UPT) ter-
minal that would work anywhere, using a variety of RF standards for fixed,
mobile and perhaps even satellite operation. The whole system would be
transparent to the user, who would not need to know how the connection
was made any more than a present-day wireline telephone user needs to
know the details of call switching and routing.
Greater Standardization As we noted earlier, there is a lack of standardiza-

tion in the second generation of wireless systems, especially in North Amer-
ica, where three incompatible systems compete and where backward
compatibility is still required with analog AMPS if a phone is to be used out-
side cities. In addition to the various terrestrial systems, each existing and
proposed satellite system uses its own protocols and modulation scheme. It
would certainly be useful to have one worldwide standard for the third gen-
eration. Failing that, a small group of standards that allowed manufacturers
of portable terminals to produce equipment that would work anywhere in
the world would be an improvement on the current situation.
Any of the second-generation systems could be extended by increasing
its data rate but leaving the basic system unchanged. Obviously, adopting a
single third-generation specification using any of the existing standards
would give an advantage to those manufacturers and service providers who
already have experience with that system. Consequently, there has been a
good deal of argument about the merits of various proposals for a 3G system.
Not all of the arguments have had a strictly technical basis.
The IMT-2000 By autumn of 1999, general agreement was reached on the outline of a
Specification third-generation system, designated by the International Telecommunica-
tions Union (ITU) as IMT-2000 (IMT stands for International Mobile Telecom-
munications, and 2000 refers both to the approximate implementation date
and to the fact that the proposed system will operate at about 2000 MHz),
which meets most of the above requirements. Maximum data rates will be
526 ! CHAPTER 14
144 kb/s for mobile users in high-speed vehicles, about 384 kb/s for pedestri-
ans and perhaps for slow-moving vehicles, and up to 2 Mb/s for stationary
users. There are two major reasons for the differences in data rates. The first
is that pedestrians, because of their lower speed, encounter handoffs less fre-
quently than users in moving vehicles. Stationary users, of course, are not
subject to handoffs at all. Secondly, stationary users are much less affected
by multipath fading.
While no one standard for the air interface has been approved, the num-
ber of different standards to be included has, after much negotiation, been
reduced from about fifteen to five. This is a decided improvement, though
still far from ideal. The specification includes one TDMA standard, one
FDMA standard, and one CDMA standard with three variations. There will
probably be multimode phones that can cope with all these standards. Fig-
ure 14.3 illustrates how the standards come together.
The need for a capacity increase necessitates a greater spectrum alloca-
tion for the new system. The World Administrative Radio Conference in
1992 assigned the frequency bands from 1885–2025 MHz and 2110–2200
MHz for this service. In North America, as we have seen, much of this band
is already in use for PCS; nonetheless, there will be some increase in the spec-
trum allocation. Part of this range, from 1980–2010 and 2170–2200 is
also used for satellite telephony, but this sharing of bands is deliberate as sat-
ellites are incorporated in the IMT-2000 standard.
FIGURE 14.3 Evolution of third-generation PCS standards

It is questionable whether the 1992 spectrum allocation will be suffi-

cient. If there is a great deal of demand for high-speed data services, it will
probably be necessary to revisit the question of spectrum allocation for
third-generation PCS.
Wideband CDMA The IMT-2000 system will incorporate three variations of CDMA. See
Systems Table14.1. The modes differ in how duplexing is accomplished and how
many carriers are used. All variations operate in a 5-MHz channel, as com-
pared to 1.25 MHz for CDMAOne (also called IS-95), the current CDMA PCS.
The new CDMA standard is an amalgam of several proposals that eventu-
ally were combined into three. One variant, called cdma2000 and proposed
by users of the North American IS-95 CDMA system, can use three separate
carriers within a single 5-MHz channel, each modulated with a chipping rate
of 1.2288 Mb/s. This is the same rate as used with IS-95, so this variation
essentially uses three IS-95 signals in one larger channel, combining the data
from the three to give some combination of greater numbers of voice calls
and higher-speed data transmission. In fact, the three carriers do not have to
be carried in contiguous spectrum; three 1.25-MHz channels at different
points in the spectrum can be used. This is to allow for backward compatibil-
ity with the existing IS-95 system. The cdma2000 system can also use a sin-
gle carrier operating at a higher chipping rate for new systems where
backward compatibility is not required.
The other variation, proposed by a consortium of European and Japa-
nese interests and called W-CDMA (for Wideband CDMA), uses a chipping
rate of 4.096 Mb/s with only one carrier per 5-MHz channel and is therefore
TABLE 14.1 CDMA Comparison
IS-95 cdma2000 W-CDMA

Specification (Current system) (U.S.A.) (Japan/Europe)
Channel 1.25 MHz 5 MHz 5 MHz

bandwidth
Number of 1 3 1
carriers
Chip rate 1.2288 Mb/s 1.2288 Mb/s per 4.096 Mb/s

carrier
Data rate 9.6 kb/s Variable to 2 Mb/s Variable to 2 Mb/s

528 ! CHAPTER 14
not backward compatible with IS-95. It also does not rely on the synchroni-
zation of base stations by means of the American GPS satellites. This pro-
vides advantages in the case of indoor base stations where the reception of
satellite signals is difficult.
See Table 14.1 for a comparison of the two major CDMA systems.
The harmonized standard will allow two CDMA modulation schemes: a
multi-carrier system resembling cdma2000, and a single-carrier scheme as in
W-CDMA.
In addition to two ways of dividing the channel, there are two ways of
providing full-duplex operation. You will recall that the IS-95 system, like all
first- and second-generation wireless systems, provides full-duplex opera-
tion by using separate carrier frequencies for the forward and reverse chan-
nel. With a digital system there is another possibility called time-division
duplexing (TDD), in which different time slots are used for each direction,
but the same RF channel is used for both. Digital PCS already use different
time slots for the two directions, but they also use different RF channels. The
new third-generation specification will allow for either TDD or the more
conventional frequency-division duplexing. The IMT-TC specification
allows for either spread-spectrum or TDMA operation in each direction.
Wideband TDMA The other variation that seems likely to be incorporated in the third-
generation PCS standard is a wideband version of the North American
TDMA system, with some elements incorporated from the European GSM
system. This proposal is known as UWC-136 and was proposed by the Univer-
sal Wireless Communications (UWC) consortium, representing operators of
and equipment providers for North American TDMA cellular systems and
PCS using the IS-136 standard.
The new TDMA standard envisions a wide variety of channel widths,
data rates, and modulation schemes, so as to allow a gradual migration from
current TDMA and GSM systems to the third-generation standard. North
American TDMA operators can begin by increasing the data rate in their
30-kHz channels, while GSM systems can do the same with their 200-kHz
channels. The plan is eventually to move to channels 1.6 MHz wide. The
wider channel will support a channel bit rate of up to 5.2 Mb/s using 8-DPSK,
that is, phase-shift keying using eight different phase angles, which allows
three bits to be transmitted per symbol. See Figure 14.4. Once the final stan-
dard is in place, all the IMT-2000 requirements should be supported. As men-
tioned above, TDMA systems can use either frequency- or time-division
duplexing.
In general, it appears that the likely result of all the complicated negotia-
tions is that the third generation of wireless communication standards will
be only slightly less chaotic than the second. A fourth generation, with data
FIGURE 14.4
8-DPSK
rates of 150 Mb/s and more, is already being considered. Perhaps by then
it will be possible to arrive at one worldwide standard. Don’t count on it
though.
' 14.4 Residential Microwave Communication Systems

Local multipoint distribution systems (LMDS) use terrestrial microwave
transmission to provide a number of services to homes and businesses. These
services can include broadcast television, high-speed internet access, and te-
lephony. In Canada the same technology is called LMCS (local multipoint
communication system). Most systems use frequencies in the 28-GHz range.
There is also a variant known as MMDS (multichannel multipoint distri-
bution system) which operates in the 2-GHz range. We’ll look at both, be-
ginning with MMDS because it is simpler and was first historically.
MMDS Multichannel multipoint distribution systems have been operating for sev-
eral years in the 2-GHz range. The first systems were analog but recent ver-
sions are digital. MMDS began as an alternative to cable television using
coaxial cable and is sometimes called by the rather contradictory name wire-
less cable for that reason.
Figure 14.5 shows the layout of a typical MMDS. Local television stations
are received off the air at the headend, which also picks up cable-network
stations by satellite. The signals are rebroadcast in a digital format called
MPEG (for Motion Picture Experts Group) by microwave from one or more tall
towers to small wall-mounted antennas on houses and apartment buildings.
530 ! CHAPTER 14
FIGURE 14.5 MMDS system
The system works over line-of-sight distances of up to about 50 km, depend-

ing on the antenna heights. Receiving antennas must have a direct line of
sight to the transmitter, so the system is uneconomical in hilly areas or
dense metropolitan areas with many tall buildings that may block the sig-
nal. There is sufficient spectrum for about 100 channels.
MMDSs are also used for internet access. Typically, all data requests from
all subscribers are transmitted on one RF channel, and each home terminal
accepts only those packets that are addressed to it. In this way the system re-
sembles a LAN.
Multichannel multipoint distribution systems are one-way: a bidirectional
system would require dividing the coverage area into much smaller cells, which
would be expensive. With an existing MMDS, any return link that may be
required for uploading data to the internet or for ordering pay-per-view tele-
vision programs must use some other technology. Usually upstream communi-
cation is accomplished using the PSTN, but subscribers find this inconvenient,
especially for internet access, as it ties up a telephone line. Recently some
MMDS operators have begun to investigate the possibility of using a low-speed
wireless link, such as PCS, for upstream communication. The only problem
with this is that it adds considerably to the cost of the system, which must
remain competitive with cable television in order to survive.
Most observers expect that MMDS services will eventually be replaced
by LMDS, which we look at next. Another possibility, since many MMDS
services are owned by telephone companies, is that these services will be

replaced by video services delivered by telephone line, when these services
become available.
LMDS The concept behind an LMDS is similar to that behind the earlier MMDS, but
with some major changes. First, moving the frequency to 28 GHz allows
much freer use of spectrum; about 1.3 GHz is allocated to this service in the
United States, and 1 GHz in Canada. Second, the high frequency causes se-
vere attenuation problems in the presence of rain, and even foliage from
trees can block the signal. This, coupled with the difficulty of generating
substantial amounts of power at 28 GHz, forces the system to use much
smaller cells with a maximum diameter on the order of 5 km. Finally, the
LMDS is designed to be bidirectional. This allows it to be used for internet ac-
cess without using the PSTN and, in fact, allows the LMDS to be used for tele-
phone communication instead of the PSTN. See Figure 14.6.
However, early versions of LMDS, such as SPEEDUS™ in New York City,
still rely on a wireline modem and a PSTN connection for upstream traffic.
Because of the very large amount of spectrum available, LMDS can offer
very high data rates. SPEEDUS, for example, advertises rates up to 48 Mb/s,
which is much greater than its two nearest competitors (cable modems using
FIGURE 14.6 LMDS system

532 ! CHAPTER 14
cable-television coaxial cable and ADSL using twisted-pair telephone wir-

ing). Both of these offer data rates on the order of 1 Mb/s.
As with MMDS, LMDS is essentially a line-of-sight technology. However,
some research shows, that due to the high frequency and short wavelength,
LMDS may be able to take advantage of reflections from buildings to achieve
coverage in some areas that are not in the direct line of sight from a transmit-
ter. See Figure 14.7.
FIGURE 14.7 Multipath propagation and LMDS systems
' 14.5 Convergence

There has been a good deal of talk, most of it rather vague, about the conver-
gence of various forms of communication. To some extent this is indeed
happening, but it seems from what we have seen so far that an equally good
case could be made for divergence.
Insofar as the idea of convergence has a technical, rather than merely a
marketing, component, it could loosely be summed up in the statement: as
all forms of communication become digital they begin to resemble each
other more than previously. A few years ago such systems as television
broadcasting, radio broadcasting, voice communication, and data transfer
were considered separately. Each had its own technology, modulation
schemes, technical and corporate structures, history, and indeed culture. All
these and more can now be viewed as data streams, differing from one an-
other in bit rate, need for error correction, need for real-time delivery and so
on, but definitely all aspects of the same technology. Therefore, any medium
capable of carrying a sufficient bit rate could be used for any of the above ser-
vices, as well as others yet to be developed.
This kind of thinking leads companies that are already involved in one
or more of the above areas to speculate about moving into others. Thus we
have internet access and, shortly, telephony on cable-television systems; in-
ternet access and, shortly, television on telephone wire (in addition to the
basic telephone service, which can be used simultaneously with the new
high-speed services); and, more germane to the focus of this book, internet
access and, eventually, video on portable wireless communication devices.
The dream of many people involved in the industry is of one medium
bringing all communication services to the consumer. Cable television peo-
ple think of this medium as the coaxial cable they already have installed to
most places. Telephone people are stuck with lower-bandwidth twisted-pair
wiring, but they point out that their switched network gives them a big ad-
vantage. Both telephone and cable-television people dream of running fiber
optics to every residence in order to provide virtually unlimited bandwidth,
but that is so expensive it is unlikely to happen for some time. Look for fiber
to the home in new housing developments, though.
From our point of view as students of wireless technology, convergence
means that we can take another look at wireless as a means of carrying ser-
vices and traffic that were formally the domain of wired systems or broadcast
radio and television. LMDS proponents claim that their system can carry
video, audio, internet access, and telephony, and do it more economically
than any of the other systems. Even third-generation PCS, with a maximum
data rate of 2 Mb/s, can be used for reasonably high-speed data and can make
a reasonable attempt at video, as well as providing telephony services reli-
able enough to make a wired telephone unnecessary for most people. These
systems have the additional advantage of portability.
' 14.6 Divergence

In the previous section we looked at convergence. Now let us look again, this
time from the point of view that the multiplicity of media offering very simi-
lar services can really be seen more as divergent than otherwise.
A few years ago the communication situation was much simpler than
now, with fewer choices. For telephone, there was the Bell System with
twisted-pair wire. For television there were three choices: over-the-air, cable
TV, and satellite systems. None of these were interactive. When internet ac-
cess came along, it was provided via telephone lines. Later, download was
available by satellite at quite a large cost premium, but uploads still needed a
telephone line.
534 ! CHAPTER 14
Portable telephony within the house used a cordless phone. Outside,

AMPS cell phones were the rule for those who could afford them. Airtime
was too expensive for anyone to use a cell phone when at home, instead of a
wired or cordless phone. Video on demand consisted of renting a VHS tape at
the local video rental outlet. The user’s communication experience was frag-
mented but clear: one medium was used for each type of communication.
Now most consumers still have cable television and wireline telephone,
but they can use either for internet access and will shortly be able to use the
phone line for television and the cable TV line for telephony. Their PCS wire-
less phone is reliable enough and almost cheap enough to use as their main
phone, but then they would not be in the phone book (and if everyone did
this, current PCS would be overloaded and would quickly become unreli-
able). In addition, the consumer can choose direct-broadcast satellite or
LMDS for television, perhaps internet access, and, in the case of LMDS, for
telephone service as well. This is not even to mention the bewildering choice
of local and long-distance telephone companies and internet service provid-
ers that can face the consumer once he or she has decided on the medium to
use. To some this sounds like convergence, but not to this writer.
Part of the convergence idea is that equipment will also converge. People
will watch television on their computer, and send e-mail with their televi-
sion. Indeed this is already possible. Whether it will become popular is an-
other question, however. If more entertainment becomes interactive, the
idea of a combined big-screen television and computer will make sense. Oth-
erwise people are more likely to watch movies in the living room and com-
municate by e-mail in the study, as now. This implies two separate pieces of
equipment, so it matters little how much technology they share.
' 14.7 Safety and Esthetic Concerns

The proliferation of wireless systems has caused some concerns about safety
and esthetics. At this time it does not look like either type of concern will
have too much effect on the proliferation of these systems, but we should
look at them. Safety, in particular, should be kept in mind by anyone work-
ing on wireless systems.
Safety Safety concerns surrounding wireless equipment are of three kinds: distrac-
tion of drivers by mobile-phone use, causing accidents; health effects due to
radiation; and possible adverse effects due either to radiation or possible
sparks, in dangerous environments.
The first problem is easily dealt with. Portable phones are a distraction to
drivers, of course, and must be dealt with like any other distraction. Some ju-
risdictions have made it illegal to use a portable phone when driving; in oth-
ers, users should exercise caution. Hands-free setups and speed dialing can
obviously help.
The second problem is more complex. Certainly wireless equipment is
not harmful in the same way that X rays and gamma rays are. It is well
known that ionizing radiation, including X rays and gamma rays, can cause
cancer by breaking molecular bonds and altering cell structure. Television
receivers and computer monitors emit low levels of X rays, and household
smoke detectors emit gamma rays. There is no safe level for this radiation,
but traditionally the danger has been assumed to be negligible if the user’s
total radiation dose is not increased significantly above the background
level that is present all the time.
Ionization depends on the energy per photon of radiation, which in-
creases with frequency. Radiation from radio equipment is at far too low a
frequency to cause ionization, regardless of its power level.
On the other hand, there are other physiological effects due to RF radia-
tion. Localized heating is the best documented. Such heating effects are par-
ticularly dangerous to the eyes, which have insufficient circulation to
remove heat. Radar technicians have become blind from working near oper-
ating radar equipment. However, the power density levels from wireless
equipment are much lower than that, and as long as reasonable precautions
are taken when working on base-station transmitting antennas, there
should be no danger on that score.
Whether there is any danger at all from exposures at levels below those
which cause damage from heating is not clear. There are many anecdotal re-
ports of people who use cell phones getting brain cancer for instance, but
this kind of juxtaposition hardly constitutes proof. After all, people who do
not use cell phones also get cancer.
Studies of the effects of radiation, or of any other environmental factor,
are of two basic types. There are epidemiological studies, which try to link
rates of various diseases to the exposure of the people to radiation. These use
statistical methods in an attempt to determine whether people exposed to
the environmental factor have more illness than those whose lives are simi-
lar except for this factor. Epidemiological studies are not concerned with the
way in which the damage is done.
There have also been laboratory studies made on mice that have been ex-
posed to radiation. To date no serious epidemiological studies have shown
any correlation between RF fields at the levels encountered by wireless
equipment users and illness. There has been some evidence that quite high
levels of radiation (well above current safety standards) may contribute to
cancer in mice. Whether the effect on humans would be similar is not clear.
536 ! CHAPTER 14
The second major way to study radiation, or other environmental fac-

tors, is to look for physiological changes in the presence of radiation. Some
studies have looked at the cellular level and claim to have found some effect
on cellular biology; again, quite high levels of radiation were used.
The problem with research into radiation safety is that it is impossible, or
nearly so, to prove that a particular radiation level is safe at all times and for
all people. A consensus has to be developed, subject to change as more data is
received. There are many variables, including power level, frequency and dura-
tion of exposure, distance of the source from the person, frequency of the radia-
tion, and even the type of modulation in use (for instance, pulsed radiation
may have more damaging effects than continuous radiation.)
Most countries have standards that limit exposure to RF radiation. In the
United States and Canada, the limit at cellular frequencies is 0.57 mW/cm2;
at PCS frequencies it is 1.2 mW/cm 2 . These standards are set at 2% of the
level at which it appears there may be some effects.
Certainly studies must be continued to see whether there are any prob-
lems with radiation that meets current standards. Obviously the question of
danger from RF radiation at low levels is of great concern to the whole wire-
less industry, but at present the attitude of the industry seems to be to keep
quiet and hope the problem goes away. There may be some logic to this: as
the industry gradually moves to systems that use smaller cells and lower
transmitted power levels, the danger due to RF radiation, is there is any, is
likely to diminish.
Sometimes cellular providers have trouble establishing base stations at
good locations because of neighbors’ concerns about radiation. However,
the effects of radiation, if any, will undoubtedly be greater from using a
phone in close proximity to one’s body than from a tower-mounted antenna
hundreds of meters away. Even though base-station transmitters are more
powerful than those in portable phones, the field strength on the ground
from such transmitters is very much lower than that from a phone held at
one’s ear, because of the square-law attenuation of free space. Most such con-
cerns are probably more esthetic than safety-oriented.
Interaction of radio transmitters with the nonhuman environment is
much easier to understand. For many years highway construction crews
have posted notices for motorists to turn off radio transmitters in the neigh-
borhood of blasting operations. This is because of the possibility that the
wiring from the detonator to the blasting cap may pick up radiation, creat-
ing an electrical current in the wire that could prematurely set off an explo-
sive charge. This is much less likely with cell phones and PCS than with the
higher-powered transmitters found in taxicabs, police cars and utility vehi-
cles, but it would still be wise for the cell phone user to comply.
Other situations pose interference dangers. Cell phone use is not al-
lowed on airplanes for two reasons: the signals could possibly affect aircraft
navigation or operation, but more likely a cell phone signal at an altitude of

several kilometers could easily be received by many cell sites, placing a
greater than normal load on the system.
Cell phones can also cause problems with sensitive electronic equip-
ment in hospitals. The best solution to that problem would be better shield-
ing of the affected equipment, but in the meantime it would be prudent for
wireless phone users to obey admonitions to turn off their phones in hospi-
tals and similar locations.
On the other hand, there are some alleged dangers that are more urban
myth than reality. For instance, some gasoline stations demand that people
refrain from using cell phones while their cars are being refilled. This is sup-
posed to be due to the danger of a spark setting off an explosion. The RF radi-
ation from cell phones is certainly not sufficient to cause arcing in external
devices and the chances of an internal spark setting off an explosion are less
from portable phones than from simpler devices, such as flashlights, since
once the phone is turned on, all the switching in cell phones is done elec-
tronically. Still, it could do no harm to avoid using anything electrical while
refilling a vehicle.
In all of these situations, it should be remembered that cell phones, PCS
phones and two-way pagers may transmit even while no one is making a
call. The only way to make sure they do not is to turn them off completely.
Esthetics Esthetic considerations are not subject to objective measurement and evalu-
ation. The proliferation of wireless devices using many different standards
has caused an equivalent proliferation of towers and antennas. This situa-
tion will become worse as cell sites become smaller. PCS base stations must
be located in residential areas, for instance, if people are to have coverage
in their homes. This causes problems for people who are used to seeing an-
tenna towers in industrial and commercial areas but not in their own neigh-
borhood.
Various techniques are available to minimize the impact of antenna in-
stallations. Antennas can be installed on, and sometimes even inside, build-
ings (church steeples are a favorite location, though sometimes the
parishioners complain about an imagined danger to their health). Antenna
masts have even been disguised as trees. See Figure 14.8 for some disguised
antennas.
As cellular and PCS antennas become more ubiquitous and as more peo-
ple use them, the chances are good that people will stop noticing them, in
much the same way they tend to ignore utility poles or street lights (in fact,
antennas for microcells are often fastened to street light poles). It certainly
appears that wireless communication is here to stay, and people who want to
get its benefits will have to put up with some minor nuisance.
538 ! CHAPTER 14
FIGURE 14.8
Camouflaged
antennas
(Courtesy of Valmont
Microflect)
' 14.8 Some Rash Predictions

Let us conclude with a few predictions. First, the future looks pretty chaotic,
at least in the near term. As we have seen, the third generation of personal
wireless communication will have almost as many conflicting standards
as the second. In addition, there will be competition between different
methods of delivering almost identical services; this is bound to become
confusing.
Perhaps the future will eventually look something like this. For voice
communication, wireless makes sense. Bandwidth demands for voice are
small and shrinking, and voice communication devices are easy to miniatur-
ize. People want to be able to call and be called wherever they are (provided
they have the option to send calls to voice mail instead of answering them
when they need privacy). At some point in the fairly near future people will
become used to carrying a miniature phone with them and using it wherever
they are. This is possible even today and will shortly become ubiquitous.
Data is more problematic. Quite a few people need mobile e-mail, which
can be accomplished quite easily even with current PCS, but high-speed
connectivity for internet access and streaming video is a different story. For
easy browsing or viewing of video, the portable device needs to be larger
than pocket size, which means most people will not want to carry it around
with them. Also, outside some niche markets, few people have a real need for
high-speed data everywhere they go. (They may find it a fun idea, but let’s
define “need” as being willing to pay for the extra convenience.) Even
third-generation wireless data will be slow and expensive compared with
wired solutions using cable modems or ADSL on telephone lines. Therefore,
for the foreseeable future, it seems likely that most people will continue to
rely on wired services (with perhaps a wireless network confined to the
home or office) for high-speed data. The possible exception is that fixed
wireless services like LMDS may be competitive with wired services in some
markets.
It also seems likely that some of the more radical developments will take
place outside the major cities on which most of our attention tends to be fo-
cused. Such cities already have extensive wired infrastructure, most of which
is already paid for. That lowers the cost for wired systems and makes it harder
for wireless to compete. The use of cellular and PCS phones instead of wired
phones is already common in developing countries that have little installed
copper, and this trend seems likely to continue.
( Until recently wireless local loops have been used only where the cost or
difficulty of installing wire is prohibitive.
( Modern CDMA equipment makes wireless local loops practical in devel-
oping countries, in rural areas, and sometimes even for extra lines where
wired service is already provided.
( The third generation of personal wireless systems will feature higher max-
imum data rates, greater capacity for voice calls, and the ability to work
with a wide range of cell sizes and types. It may also be somewhat more
standardized than the second generation.
( Both CDMA and TDMA systems appear likely to be part of the third-
generation specifications.
( Terrestrial microwave systems at 28 GHz are beginning to be used to de-
liver television, internet, and telephone services to individual residences.
They will probably supplant an older one-way microwave system.
( Many people are expecting that the digitization of practically all commu-
nication systems will lead to a gradual convergence of systems, but it is
doubtful whether this will actually happen in the near future.
540 ! CHAPTER 14
( Safety and esthetic concerns could slow the development of wireless tech-
nology.
( The future of wireless seems assured, particularly in the areas of voice and
low-speed data. Truly high-speed data may have to wait some time for the
fourth generation of wireless.
( Key Terms
local multipoint distribution system (LMDS) network using microwaves
for two-way transmission of telephony, television, and high-speed
data
multichannel multipoint distribution system (MMDS) terrestrial
microwave system for the distribution of television, internet, and
telephone services to businesses and residences
time-division duplexing (TDD) transmission of data in two directions on
a channel by using different time slots for each direction
( Questions
1. Under what circumstances are wireless local loops more economical
than ordinary twisted-pair wiring?
2. What is the difference between simply using a PCS phone at home and
using a modern wireless local loop?
3. What advantages do wireless local loops have in areas where phone
services are just being introduced?
4. What are the major improvements to be expected from a third-
generation PCS standard?
5. Why do third-generation PCS standards specify three different maxi-
mum data rates?
6. What are the variations in mobility specified for the third PCS genera-
tion? Why does the speed at which a user is moving make a difference?
7. What are the data rates envisioned by the IMT-2000 specifications, and
how do they compare with current systems?
8. The second generation of PCS uses three incompatible standards in
North America. Does the third generation improve on this situation?
9. Explain time-division duplexing. How does it compare with traditional
duplexing methods?
10. How does the cdma2000 system maintain backward compatibility with
IS-95? Under what circumstances would this be a useful feature?
11. Explain the difference between MMDS and LMDS.
12. How does an individual user communicate with the headend in an
MMDS? Are there any obvious problems with this method?
13. Why do maximum propagation distances at 28 GHz tend to be less than
at lower frequencies?
14. Why is a direct line of sight not always required for LMDS?
15. Why can LMDS, but not MMDS, be used for local telephony?
16. Compare the radiation from cell phones with that from X-ray machines
and smoke detectors, in terms of known hazards.
17. Why is there no “safe” level for gamma radiation?
18. What is the major known danger from high-power microwave sources
like radar transmitters and microwave ovens?
19. Compare the reasons for restricting the operation of wireless equipment
in aircraft, hospitals, gas stations, and movie theatres.
( Problems
1. Compare the use of wireless and wired technologies for each of the fol-
lowing services:
(a) Local telephone service
(b) Internet access
(c) Broadcast television
(d) Interactive video
2. List all of the ways by which the following can be accomplished in your
area, and discuss each method, including advantages, disadvantages,
and price.
(a) Local telephone service
(b) Reception of broadcast television stations
(c) Reception of television signals that are not broadcast over the air
(for instance, news and sports stations, movie stations)
(d) Music service (radio broadcasting or equivalent)
(e) Internet access (Be sure to get approximate data rates, and find out,
in the case of non-telephone services, whether a telephone modem
is still required for uploads.)
542 ! CHAPTER 14
(f) E-mail (Remember that this is not necessarily the same as full in-
ternet access.)
(g) Paging services
3. Suppose a wireless system operates at 200 kb/s using BPSK. What data
rate would be possible in the same bandwidth using
(a) QPSK?
(b) 8-DPSK?
4. Find out what internet service providers are available in your area, and
their approximate cost, for
(a) Telephone modems
(b) Cable modems, if available
(c) ADSL on telephone lines, if available
(d) LMDS, if available
(e) Satellite service (Don’t forget to include the cost of the uplink using
a telephone line and internet service provider.)
5. Suppose that a line-of-sight radio path has a path loss of 50 dB at 1 GHz.
(a) What will be the free-space path loss, in dB at 28 GHz?
(b) What accounts for the increase in loss?
(c) Are there any factors that are likely to make the path loss at 28 GHz
higher than that calculated in part (b)? If so, what are they?
6. What zoning restrictions, if any, exist in your area to restrict either:
(a) towers for base stations?
(b) antennas attached to houses, for direct-broadcast satellites or
LMDS?
7. Compare the data rate achievable with 3G PCS with that available from
telephone modems, cable modems, and ADSL using telephone lines.
8. Make some rash predictions of your own concerning the future of wire-
less (and wired) communication.
Appendix A Impedance Matching
In Chapter 6 we noted the deleterious effects of impedance mismatches on

transmission lines. Mismatches result in power being reflected back to the
source and in higher-than-normal voltages and currents that can stress the line
and connected equipment. In general, best results are obtained when the load
is matched to the characteristic impedance of the line. When this is not the
case, it is often useful to connect some sort of matching network to correct
the mismatch. Similarly, when either the load or the line is unbalanced,
and the other is balanced, it is highly desirable to install a device called a balun
network to convert one to the other. Matching networks can be constructed
using lumped constants (inductors, capacitors, and transformers), transmis-
sion line sections, or waveguide components. Matching networks are some-
times broadband, so as to pass a wide range of frequencies, but often they are
made narrow in bandwidth to allow them to act as bandpass filters as well.
We noted earlier that the impedance looking into a transmission line
varies with its length. For a lossless line the impedance repeats every one-
half wavelength. If it is acceptable to have a mismatch on a short section of
line near the load, it is often possible to simplify a matching problem by
backing off a short distance (less than one-half wavelength) from the load
and installing the matching device at a point on the line where the imped-
ance is easier to work with. It should be obvious that this is a narrowband
technique, since the electrical length (in degrees) of a particular physical
length of line (in meters) varies with frequency.
The Smith Chart

For many years (since 1944 in fact), a special circular graph called a Smith
chart has been used to indicate complex impedances and admittances and
the way in which they vary along a line. It is more common nowadays to do
transmission line calculations with the aid of a computer than to use pencil
and compass, but many of the computer programs use the Smith chart to dis-
play their output. Test instruments, called vector network analyzers, that
measure complex functions, also tend to use this display format. Therefore
it is important to understand the layout of the Smith chart and to be able to
interpret it—the actual calculations can be safely left to a computer.
Figure A.1 shows a conventional paper Smith chart. A distance of one-
half wavelength on the line corresponds to one revolution around the chart.
Clockwise rotation represents movement toward the generator, and coun-
terclockwise rotation represents progress toward the load, as shown on the
543
544 ! APPENDIX A
FIGURE A.1 Smith Chart (Courtesy of Analog Instruments Company)

IMPEDANCE MATCHING ! 545
chart itself. For convenience there are two scales (in decimal fractions of a
wavelength) around the outside of the chart, one in each direction. Each
scale runs from zero to 0.5 wavelength.
The body of the chart is made up of families of orthogonal circles; that
is, they intersect at right angles. The impedance or admittance at any point
on the line can be plotted by finding the intersection of the real component
(resistance or conductance), which is indicated along the horizontal axis,
with the imaginary component (reactance or susceptance), shown above the
axis for positive values and below for negative.
Because of the wide variation of transmission line impedances, most
paper Smith charts use normalized impedance and admittance to reduce the
range of values that have to be shown. The value of 1 in the center of the
chart represents an impedance equal to Z 0 . Paper charts are also available
with an impedance of 50 Ω in the center, and computerized displays can
have any value for Z0, with no normalization required.
To normalize an impedance, simply divide it by the characteristic
impedance of the line.
Z
z = (A.1)
Z0
where
z = normalized impedance at a point on the line
Z = actual impedance at the same point
Since z is actually the ratio of two impedances, it is dimensionless.
EXAMPLE A.1 Y
Normalize and plot a load impedance of 100 + j25 Ω on a 50-Ω line.
SOLUTION
From Equation (A.1),
Z
z =
Z0
100 + j 25 Ω
=
50 Ω
= 2 + j0.5
Figure A.2 shows how this impedance would be plotted on a normalized

chart.
546 ! APPENDIX A
FIGURE A.2 Use of Smith Chart (Courtesy of Analog Instruments Company)
X
Once the normalized impedance at one point on the line has been plotted,
the impedance at any other point can be found very easily. Draw a circle with
its center in the center of the chart, which is at the point on the horizontal axis
where the resistive component is equal to one. Set the radius so that the circle
passes through the point just plotted. Then draw a radius through that point,
right out to the outside of the chart, as shown in Figure A.2. Move around the
outside in the appropriate direction, using the wavelength scale as a guide. Just
follow the arrows. If the first point plotted is the load impedance, then move in
the direction of the generator. Once the new location on the line has been
found, draw another radius. The normalized impedance at the new position
is the intersection of the radius with the circle. In other words, the circle is
the locus of the impedance of the line. Every point on the circle represents the
impedance at some point on the line.
The radius of the circle represents the SWR on the line. In fact, it is usu-
ally referred to as the SWR circle. The SWR can easily be found by reading the
normalized resistance value where the circle crosses the horizontal axis to
the right of the center of the chart. Another way is to mark the radius of the
circle on the “standing wave voltage ratio” scale at the bottom of the chart. If
required, the SWR can also be found in decibels by using the adjoining scale.
For this example, the SWR is read as 2.16.
Figure A.3 on page 548 is a computer printout from a program called
WinSmith  , which is one of many that perform Smith chart calculations.
The output is provided on a Smith chart, as well as in tabular form.
We noted before that the SWR depends on the magnitude of the reflec-
tion coefficient. There is another scale on the chart that allows this magni-
tude to be found directly. See the reflection coefficient voltage scale at the
lower right of the chart. Marking off the radius of the SWR circle on this scale
will give the magnitude of Γ, which can be read as 0.37, on both the paper
and computerized charts. (On the computer chart, Γ is represented by G and
VSWR by V.)
It is possible to use the Smith chart to find the impedance at any point
along a line. The next example illustrates this.
EXAMPLE A.2 Y
A 50-Ω line operating at 100 MHz has a velocity factor of 0.7. It is 6 m long
and is terminated with a load impedance of 50 + j50 Ω. Find the input imped-
ance for the line.
SOLUTION
Now that we have seen how much the computer display resembles the paper
chart, let us use the computer. See Figure A.4.
548 ! APPENDIX A
FIGURE A.3
Printout from
winSmith software
(Courtesy of Noble
Publishing)
We need to know the length of the line in degrees. First find the wave-
length using Equation (6.1).
vp
λ =
ƒ
v ƒc
=
ƒ
0.7 × 300 × 106

=
100 × 106
= 2.1 m
Now we can find the length in degrees with Equation (6.7).
360L
φ =
λ
360 × 6
=
2.1
= 1029°
From the plot in Figure A.4, we can see that the input impedance (repre-
sented by the small circle) is 19.36 + j5.44 Ω.
FIGURE A.4
Printout from
winSmith software
(Courtesy of Noble
Publishing)
Now that we understand how impedances and lines are plotted on the
Smith chart, it should be possible to use the chart as an aid for impedance
matching. Since the center of the chart always represents the characteristic
impedance of the system, matching a line involves moving its input imped-
ance into the center of the chart. The progress of the solution to a matching
problem can be monitored by observing the input impedance: the closer it is
to the center of the chart, the better the match.
Probably the best way to see how this works is to try a few examples
using various matching techniques. Each technique will be described,
then followed with an example. Smith charts will be used to illustrate the
technique.
550 ! APPENDIX A
Matching Using a Transformer

A transformer can be used to match impedances, provided the load imped-
ance is real at the point where the transformer is inserted. RF transformers
are usually toroidal, using ferrite or powdered iron cores as illustrated in Fig-
ure A.5, although air-core transformers are also common. Recall from basic
electrical theory that the impedance ratio is the square of the turns ratio,
that is:
2
Z1 N 
=  1 (A.2)
Z2  N2 
FIGURE A.5
Toroidal
transformer
Transformers are also useful for connecting balanced lines to unbal-

anced lines, since there is no electrical connection between windings. A
transformer used in this way is called a balun transformer or just a balun.
A good example of a balun transformer can be found on the back of many
television sets, where it adapts 75-Ω coaxial (unbalanced) cable-television
cable to an antenna input designed for 300-Ω twin-lead (balanced). Fig-
ure A.6 is a photo of a typical TV balun. Since the required impedance ratio is
4:1, the turns ratio must be 2:1.
FIGURE A.6
TV balun
A transformer can be used to match loads with complex impedances by

installing it a little distance from the load in the direction of the source. The
next example illustrates this.
EXAMPLE A.3 Y
Find the correct turns ratio and location for a transformer that is required to
match a 50-Ω line to a load impedance of 75 + j25 Ω.
SOLUTION
We solve the problem by moving along the line in the direction of the gener-
ator until the impedance looking into the line is resistive. The arc represent-
ing the line crosses the horizontal axis of the chart at a point 16.8° from the
load. See Figure A.7(a).
FIGURE A.7(a)
Printout from
winSmith software
(Courtesy of Noble
Publishing)
Now the requirement is simply to match the 50-Ω line to the impedance at
this point on the line, which is 88.38 Ω, resistive. The required turns ratio
can be found from Equation (A.2):
552 ! APPENDIX A
2
Z1 N 
=  1
Z2  N2 
N1 Z1
=
N2 Z2
50
=
88.38
= 0.752
Figure A.7(b) shows what happens when we add a transformer with

this ratio at the point found above. The line is matched. Note, however,
that a change in frequency will change the electrical length of the line
and destroy the match. This can be seen by changing the frequency from
30 MHz to 40 MHz while leaving everything else alone. See the result in
Figure A.7(c) on page 553. There is no longer a perfect match; however, a
closer look shows that the SWR is still a very reasonable 1.12, so the setup is
still usable.
FIGURE A.7(b)
Printout from
winSmith software
(Courtesy of Noble
Publishing)
FIGURE A.7(c)
Printout from
winSmith software
(Courtesy of Noble
Publishing)
A one-quarter wavelength of transmission line can also be used as a

transformer. A quarter wavelength along a line represents one-half a rota-
tion around the chart, so a one-quarter wavelength of line of the right im-
pedance can transform one real impedance into another. The characteristic
impedance Z ′0 of the line for the transformer can be found from the follow-
ing equation:
Z 0′ = Z0ZL (A.3)
EXAMPLE A.4 Y
Solve the problem of the previous example using a quarter-wave trans-
former.
SOLUTION
Since a quarter-wave transformer, like a conventional transformer, can only
match real impedances, it is necessary to place the transformer at the same
554 ! APPENDIX A
distance from the load as in the previous example. From the previous prob-
lem we see that the load impedance to be matched at this point is 88.38 Ω.
This can be matched with a quarter-wave section of line having a character-
istic impedance, from Equation (A.3), of
Z 0′ = Z0ZL
= 50 × 88.38
= 66.48 Ω
Figure A.8 shows that this does indeed result in a match.
FIGURE A.8
Printout from
winSmith software
(Courtesy of Noble
Publishing)
X
Series Capacitance and Inductance

Where the resistive part of the load impedance is correct, the reactive part
can be canceled by adding a series reactance of the opposite type. If the resis-
tive part is not correct, the reactance can be installed at the correct distance
from the load to bring the resistive part of the impedance to the cor-
rect value.
EXAMPLE A.5 Y
Use a series capacitor or inductor to match a 50-Ω line to each of the follow-
ing loads at a frequency of 100 MHz:
(a) 50 + j75 Ω
(b) 150 + j75 Ω
SOLUTION
(a) Since the real part of the impedance is correct, we need only add a capac-
itive impedance of −j75 Ω, that is, a capacitor with 75-Ω reactance. Recall
from electrical fundamentals that
1
Xc =
2πƒC
where
Xc = capacitive reactance in ohms
C = capacitance in farads
In this case, we know f and Xc , so we rearrange the equation:
1
C =
2πƒ X c
1
=
2 π × 100 × 106 × 75
= 21.2 pF
It was not necessary to use the Smith chart, whether on paper or on

a computer, to solve this problem, but the computerized version of the chart
verifies the match. See Figure A.9 on page 556, and notice that the arc on the
chart follows one of the circles of constant resistance to the center of
the chart.
556 ! APPENDIX A
FIGURE A.9
Printout from
winSmith software
(Courtesy of Noble
Publishing)
(b) Since the real part of the load impedance is not equal to 50 Ω, it is neces-
sary to add enough line to reach a point on the circle representing a resis-
tance of 50 Ω, that is, the circle that passes through the center of
the chart. Figure A.10(a) on page 557 shows that this takes a length of
35 electrical degrees. At this point the resistive component of the imped-
ance is correct but there is a capacitive reactance of 72.6 Ω. Fig-
ure A.10(b) shows that the system can be matched by using a series
inductance of 115 nH.
FIGURE A.10(a)
Printout from
winSmith software
(Courtesy of Noble
Publishing)
FIGURE A.10(b)
Printout from
winSmith software
(Courtesy of Noble
Publishing)
X
558 ! APPENDIX A
Stub Matching
As noted earlier, shorted transmission line stubs are often used instead of
capacitors or inductors at VHF and above. Usually these are placed in parallel
with the main line, rather than in series. In this case it is easier to handle the
problem from an admittance rather than an impedance point of view, since
parallel admittances add. Shorted transmission lines can compensate only
for the imaginary component of the load admittance, so if the load admit-
tance is complex it is once again necessary to install the matching compo-
nent at some distance from the load. In this case we back off until the real
component of the admittance is equal to the characteristic admittance of
the line.
Smith charts are available with admittance as well as impedance coor-
dinates. It is also possible to use a conventional chart and convert from
impedance to admittance. Computer programs invariably offer a choice of
coordinates. The WinSmith program introduced earlier allows for either
or both to be displayed. An example will demonstrate transmission line
matching with a single shorted stub.
EXAMPLE A.6 Y
Match a line with a characteristic impedance of 72 Ω to a load impedance of
120 − j100 Ω using a single shorted stub.
SOLUTION
The stub must be inserted at a point on the line where the real part of the
load admittance is correct. This value is
1
= 0.0139 S
72 Ω
Actually, we can simply find the conductance circle that passes through
the center of the chart. Figure A.11(a) on page 559 shows a chart with
both impedance and admittance circles. This is a little confusing, so in
Figure A.11(b) the impedance circles have been removed. It can be seen
from either figure that a distance of 42° is required between the load and the
stub.
Next we add a shorted stub and adjust its length to bring the input im-
pedance to the center of the chart. It can be seen from Figure A.11(c) on page
560 that this requires a stub length of 40°.
FIGURE A.11(a)
Printout from
winSmith software
(Courtesy of Noble
Publishing)
FIGURE A.11(b)
Printout from
winSmith software
(Courtesy of Noble
Publishing)
560 ! APPENDIX A
FIGURE A.11(c)
Printout from
winSmith software
(Courtesy of Noble
Publishing)
The foregoing is only a sample of the ways in which lines can be

matched to loads.
The problems that follow can be solved using a paper Smith chart or a
computer program, if available.
( Problems
1. A 75-Ω transmission line is terminated with a load having an imped-
ance of 45 − j30 Ω. Find:
(a) the distance (in wavelengths) from the load to the closest place at
which a quarter-wave transformer could be used to match the line
(b) the characteristic impedance that should be used for the quarter-
wave transformer
2. A 50-Ω coaxial transmission line with a velocity factor of 0.8 is con-

nected to a load of 89 − j47 Ω. The system operates at 300 MHz. Find the
proper position and component value to match this line using:
(a) a series capacitor
(b) a series inductor
(c) a transformer
3. Find the SWR for each setup of problem 2, at a frequency of 350 MHz.
4. A transmitter supplies 100 W to a 50-Ω lossless line that is 5.65 wave-
lengths long. The other end of the line is connected to an antenna with
a characteristic impedance of 150 + j25 Ω.
(a) Find the SWR and the magnitude of the reflection coefficient.
(b) How much of the transmitter power reaches the antenna?
(c) Find the best place at which to insert a shorted matching stub on
the line. Give the answer in wavelengths from the load.
(d) Find the proper length for the stub (in wavelengths).
Appendix B Frequencies for Common Wireless Systems
CITIZENS’ BAND RADIO: 27-MHz Band
Frequency (MHz). Frequency (MHz).

(Same channel used for (Same channel used for
Channel both directions.) Channel both directions.)
1 26.965 21 27.215
2 26.975 22 27.225
3 26.985 23 27.255
4 27.005 24 27.235
5 27.015 25 27.245
6 27.025 26 27.265
7 27.035 27 27.275
8 27.055 28 27.285
9 27.065 29 27.295
10 27.075 30 27.305
11 27.085 31 27.315
12 27.105 32 27.325
13 27.115 33 27.335
14 27.125 34 27.345
15 27.135 35 27.355
16 27.155 36 27.365
17 27.165 37 27.375
18 27.175 38 27.385
19 27.185 39 27.395
20 27.205 40 27.405
563
564 ! APPENDIX B
CORDLESS TELEPHONE FREQUENCIES: 43–49 MHz Band
Channel Base Handset
1 43.720 48.760
2 43.740 48.840
3 43.820 48.860
4 43.840 48.920
5 43.920 49.000
6 43.960 49.080
7 44.120 49.100
8 44.160 49.160
9 44.180 49.200
10 44.200 49.240
11 44.320 49.280
12 44.360 49.360
13 44.400 49.400
14 44.460 49.480
15 44.480 49.500
16 46.610 49.670
17 (B) 46.630 49.845
18 (C) 46.670 49.860
19 46.710 49.770
20 (D) 46.730 49.875
21 (A) 46.770 49.830
22 (F) 46.830 49.890
23 46.870 49.930
24 46.930 49.990
25 46.970 49.970
The letters A through F denote channels that are also used for baby monitors.
FREQUENCIES FOR COMMON WIRELESS SYSTEMS ! 565
NORTH AMERICAN CELLULAR RADIO FREQUENCIES
Base Frequencies Mobile Frequencies Type of

(forward channels) (reverse channels) Channel Carrier
869.040–879.360 824.040–834.360 Voice A
879.390–879.990 834.390–834.990 Control A
880.020–880.620 835.020–835.620 Control B
880.650–889.980 835.650–844.980 Voice B
890.010–891.480 845.010–846.480 Voice A*
891.510–893.970 846.510–848.970 Voice B*
Table denotes transmit carrier frequencies. Mobile transmits 45 MHz below base.
A = non-wireline carrier (RCC) B = wireline carrier (Telco)
* = frequencies added in 1986
BROADBAND PCS BAND PLAN
Base transmit (Forward Mobile transmit (Reverse

Allocation Channel or Downlink) Channel or Uplink)
A 1850–1865 1930–1945
B 1870–1885 1950–1965
C 1895–1910 1975–1990
D 1865–1870 1945–1950
E 1885–1890 1965–1970
F 1890–1895 1970–1975
Frequencies are in MHz.

Answers to Odd-Numbered Problems
Chapter 1
1. 62.8 × 103 rad/s

3. 16 dB
5. See Figure 1. FIGURE 1
9. 750 THz, 429 THz
Chapter 2
1. (a) 7.4 MHz

(b) 7.0 kHz
(c) 0.267
(d) 15 V
567
568 ! ANSWERS TO ODD-NUMBERED PROBLEMS
(e) See Figure 3. FIGURE 3
3. (a) 8 V
(b) 50%
(c) See Figure 4. FIGURE 4
(d) v(t) = (8 + 4 sin(31.4 × 103t))sin(62.8 × 106t)V

5. (a) 0.1
(b) 10 V
(c) 1 V
7. 30 kHz
11. (a) 15.8 mW

(b) 0.796
CHAPTER 2 ! 569
15. (a) 16 W
(b) 8 dB
17. 10 kHz
19. (a) No effect

(b) Doubling Vm doubles the modulation index.
21. (a) 2.5 kHz
(b) 484 mW
(c) 96.6%
23. 154 W
25. (a) 316 µV
(b) 0.0316 rad
(c) 50 dB
27. For full-carrier AM, sideband power is in addition to carrier power, which remains constant.
For SSBSC AM, the carrier power is zero.
Chapter 3
1. (a) 7 kHz
(b) Practical filters need some distance between passband and stopband.
3. See Figure 3.4. Make the time intervals 250 µs and the peak voltage 1 V in part (a).
5. 75.2%
7. 11101111
9. 11110011
Chapter 4
1. (a) 60 kb/s
(b) 120 kb/s
3. 0.083 b/s/Hz; No
5. (a) The number of states (angles, in
this case) is 23 = 8
(b) See Figure 7.
(c) Advantage: higher data rate for a
given bandwidth. Disadvantage: less
robust in the presence of noise or
interference. FIGURE 7
CHAPTER 6 ! 571
7. (a) 8 17. (a) 320 kHz

(b) 640 Mb/s (b) 40 kHz
9. 30 (c) 60 kHz
11. (a) 12 Mb/s (d) 30 kHz
(b) 187 (e) 60 kHz
(c) 8 (f) 300 kHz
13. 500
15. (a) 1 Mb/s
(b) 667 kHz
(c) 8 dB
Chapter 5
1. (a) 50 mA 7. 33.3%, 0.52% (assuming no bit robbing)

(b) 760 Ω 9. (a) upstream 10, downstream 23.4
3. 852 Hz, 1336 Hz (b) upstream 5, downstream 11.7
5. (a) 16.128 Mb/s
(b) 5.88%
Chapter 6
1. (a) 1.875 m 3. 429 THz to 750 THz

(b) 37.5 cm 5. 47.6 Ω
(c) 15 cm 7. 240 m
9. See Figure 8.
FIGURE 8
11. (a) 1.5

(b) −0.2
(c) 400 µW
13. Series resonant
15. 0.594 µV
17. 6.55 GHz, 6.55−13.1 GHz
19. 6.6 ns
CHAPTER 7 ! 573
21. (a) See Figure 9(a). FIGURE 9

(b) See Figure 9(b).
(c) The circulator does
not require a load.
Chapter 7
1. 107 × 106 m/s 11. (a) 150 dB

3. 2
35.4 nW/m , 3.65 mV/m (b) 160 dB
5. 135 Ω 13. 3.6 min.
7. (a) 49 dBW 15. 167 µs, 3.33 ms, 240 ms

(b) 178 pW 17. 581 fW/m2
9. (a) 12.5 km
(b) 25.6 km
(c) 51.2 km
Chapter 8
1. 950 mm 5. 15.7 dBW

3. 2.71 mV/m 7. 71.25 cm
9. (a) 750 MHz
(b) 6.76 dBi
(c) 33°
11. θH = 24° θE = 19.3°
13. (a) parasitic, end-fire, to the right
(b) phased, end-fire, to the left
(c) phased, broadside, all directions except left and right
(d) phased, broadside, into and out of the page
(e) phased, end-fire, to the left
15. (a) 3 dB increase in gain
(b) cancellation
17. (a) 39 dBi (b) 39 dBi
19. (a) 262 µW (b) 8.28 µW
21. 7.68 dBi
Chapter 9
1. (a) 2000
(b) 2000 to 2400
3. See Figure 10.
FIGURE 10
CHAPTER 9 ! 575
5. (a) 220 MHz, 20 MHz

(b) 20 MHz, 100 MHz, 120 MHz, 200 MHz, 220 MHz, 240 MHz
7. 2.5 µW
9. (a) 1.65 MHz
(b) High-side
11. 108 dB
13. (a) 17.5 MHz
(b) 14 MHz
15. 15 kHz
17. (a) High-side
(b) 750 kHz
(c) ƒLO1 = 25.5 MHz, ƒLO2 = 9.25 MHz
19. (a) See Figure 11.
FIGURE 11
(b) 115.9 MHz

(c) Improve bandpass filter before mixer; raise IF
21. (a) 156.7 MHz, 139.8 MHz
(b) Direct-FM using PLL
(c) 16 kHz (from Carson’s rule)
Chapter 10
1. (a) 427 nW
(b) 380 µW; effective area of receiving antenna is smaller
3. 2 W; suitable for mobile but not portable cell phone
5. (a) 70.7 m
(b) (i) Less interference
(ii) No change
(c) (i) Improvement
(ii) No change
7. 112.5 m
9. 40 000
11. (a) 34 cells with a radius of 920 m
(b) better for digital phones, much worse for analog
(c) leave more channels analog
13.
User Action Channel Data Sent
Key in no.; Press send reverse control MIN, ESN, number called
forward control no. of voice channel, CNAC, DCC
forward voice confirmation, SAT
reverse voice confirmation, SAT
Talk and listen forward and reverse voice conversation, SAT
15. (a) 12.5 km

(b) 9.3 km
(c) Shadowing by hills or buildings, multipath propagation, attenuation due to operating
portable inside a car or building, sensitivity of cell phone receiver, etc.
17. 192 voice channels
CHAPTER 12 ! 577
19.
System P (mW) P (dBm)
Class IV digital cell phone (Minimum) 0.4 −4
49-MHz cordless phone 0.03 −15.2
900-MHz cordless phone 16 +12
Chapter 11
1. (a) 82.8 mm, 37.3 mm

(b) same
(c) 800 MHz, by 6.7 dB
3. (a) 2.53 kb/s
(b) 506 b/s
5. (a) TDMA 1.62 b/s/Hz; GSM 1.35 b/s/Hz; CDMA 0.98 b/s/Hz
(b) TDMA 0.81 symbol/s/Hz; GSM 1.35 symbol/s/Hz; CDMA 0.49 symbol/s/Hz
7. (a) 52.1 µs
(b) 7.3 years
9. 6.67 µs, 3.33 µs, 0
Chapter 12
1. 7.35 km/s; 1.76 hours

3. 7.91 km/s; air resistance
5. Increase antenna gain to 55 dBi or reduce noise temperature to 6.6 K.
7. 290 K
9. LEO by 38 dB
11. (a) 154 dB
(b) 5.2 ms
13. (a) 7464 m/s, 1.67 hours
(b) 7155 m/s, 1.91 hours
(c) 7173 m/s, 1.89 hours
(d) 7461 m/s, 1.68 hours
Chapter 13
1. 0.02 b/s/Hz, 0.048 b/s/Hz, 0.096 b/s/Hz

3. approximately 5.3 times
5. 0.38 b/s/Hz
7. 19 dB
9.
Format Data Rate Network Type
Mobitex 8 kb/s Packet-switched
ARDIS 4.8 kb/s standard; 19.2 kb/s in some areas Packet-switched
CDPD 19.2 kb/s Packet-switched
Analog Cellular with Modem 9.6 kb/s approx., depending on connection Circuit-switched
TDMA PCS 9.6 kb/s Circuit-switched
GSM PCS 9.6 kb/s Circuit-switched
CDMA PCS 14.4 kb/s Circuit-switched
Chapter 14
1. (a) Wireless is more expensive in built-up areas with high telephone density, often less
expensive in rural areas or areas with few customers. Wireless has the advantage of
portability and the promise of a single portable phone number. Wired technology
has arguably better security, though new digital wireless technologies are better than
older ones.
(b) Wired technologies are currently faster and cheaper, especially cable modems and ADSL.
Wireless service offers portability. Fixed wireless using LMDS/LMCS is competitive with
cable modems and ADSL; fixed wireless is not portable.
(c) Wired service (CATV) is economical in high-density areas; in rural areas LMCS/LMDS
and satellite service are competitive. Conventional terrestrial broadcasting is cost-
effective but offers a limited number of channels.
(d) This can be supported on CATV, LMDS/LMCS, and in the near future, on telephone
lines. The three systems appear to be competitive, with CATV having the advantage of a
large installed base. Low-quality interactive video is possible on the internet, though at
present wireless internet speeds are barely adequate even for this.
APPENDIX A ! 579
3. (a) 400 kb/s

(b) 600 kb/s
5. (a) 78.9 dB
(b) decrease in effective area for receiving antenna of equivalent gain
(c) attenuation by trees, rain, etc.
7. All are variable. Let us assume telephone modems operate at 64 kb/s and both ADSL and
cable modems at about 1 Mb/s; for 3G wireless, maximum data rates will be 144 kb/s for
mobile users in high-speed vehicles, about 384 kb/s for pedestrians and perhaps for
slow-moving vehicles, and up to 2 Mb/s for stationary users. All types of 3G are faster than
telephone modems, while only stationary 3G PCS users will have data rates comparable with
and perhaps exceeding cable modems and ADSL.
Appendix A
1. (a) 49°
(b) 44.6 Ω
3. (a) 2.07
(b) 1.27
(c) 1.52
Glossary
Advanced Mobile Phone Service (AMPS) North baseband information signal

American first-generation cellular radio baud rate speed at which symbols are
standard using analog FM transmitted in a digital communication
air interface in wireless communication, the system
radio equipment and the propagation path beamwidth angle between points in an
aliasing distortion created by using too low a antenna pattern at which radiation is 3 dB
sampling rate when coding an analog signal down from its maximum
for digital transmission bent-pipe configuration a design of satellite
amplitude-shift keying (ASK) data transmission transponder in which signals are amplified
by varying the amplitude of the transmitted and shifted in frequency but do not undergo
signal any other processing
angle modulation term that applies to both bit rate speed at which data is transmitted in a
frequency modulation (FM) and phase digital communication system
modulation (PM) of a transmitted signal bit robbing use of bits that normally carry
angle of elevation angle measured upward from payload information for other purposes, such
the horizon. Used to describe antenna as controlling the communication system
patterns and directions bit stuffing addition of bits to a bitstream to
apogee point in a satellite orbit that is farthest compensate for timing variations
from the earth blank-and-burst signaling in cellular
array combination of several antenna elements communication, interrupting the voice
channel to send control information
asymmetrical digital subscriber line (ADSL)
method of providing high-speed data bridge a device to connect two segments of a
transmission on twisted-pair telephone loops network
by using high-frequency carriers
audio frequency-shift keying (AFSK) use of an call blocking failure to connect a telephone
audio tone of two or more different call because of lack of system capacity
frequencies to modulate a conventional capcode unique address for a pager
analog transmitter for data transmission capture effect tendency of an FM receiver to
receive the strongest signal and reject others
B (bearer) channels in ISDN, channels that carrier high-frequency signal which is
carry subscriber communication modulated by the baseband signal in a
Bandwidth portion of frequency spectrum communication system
occupied by a signal carrier-sense multiple-access with collision
base station controller (BSC) in cellular and avoidance (CSMA/CA) same as CSMA/CD
PCS systems, the electronics that controls base except used in situations where a station
station transmitters and receivers cannot receive and transmit at the same
581
582 ! GLOSSARY
time. With CSMA/CA, a collision cannot be companding combination of compression at

detected once a transmission has started. the transmitter and expansion at the receiver
of a communication system
carrier-sense multiple-access with collision
detection (CSMA/CD) method of reducing compression deliberate reduction of the
contention in a network, involving each dynamic range of a signal by applying more
station checking continuously for inter- gain to lower-amplitude than to higher-
ference before and during transmissions amplitude components
constellation in satellite telephony, a group
cellular digital packet data (CDPD) method of
of satellites coordinated in such a way as to
transmitting data on AMPS cellular telephone
provide continuous communication
voice channels that are temporarily unused
constellation diagram in digital
central office switch in a telephone system that communication, a pattern showing all the
connects to local subscriber lines possible combinations of amplitude and
characteristic impedance ratio between voltage phase for a signal
and current on a transmission line, on a wave- control mobile attenuation code (CMAC)
guide, or in a medium information sent by the base station in a
cellular radio system to set the power level
chip extra bit used to spread the signal in a
of the mobile transmitter
direct-sequence spread-spectrum system
crosslink a radio or optical connection directly
circuit-switched network communication between satellites, without going through an
system in which a dedicated channel is set earth station
up between parties for the duration of the
crosstalk interference between two signals
communication
multiplexed into the same channel
citizens’ band (CB) radio short-distance
unlicensed radio communication system D (data) channel in ISDN, a communication
codec (coder-decoder) device that converts channel used for setting up calls and not for
sampled analog signal to and from its PCM user communication
or delta modulation equivalent decoding conversion of a PCM or delta
modulation bitstream to analog samples
code-division multiple access (CDMA) system
to allow multiple users to use the same delta modulation a coding scheme that
frequency using separate PN codes and a records the change in signal level since the
spread-spectrum modulation scheme previous sample
delta phase-shift keying (DPSK) digital
coding conversion of a sampled analog signal
modulation scheme that represents a bit
into a PCM or delta modulation bitstream
pattern by a change in phase from the
collision attempt by two transmitters to use the previous state
same channel simultaneously demodulation recovery of a baseband signal
common-channel signaling use of a separate from a modulated signal
signaling channel in a telephone system, deviation in FM, the peak amount by which
so that voice channels do not have to carry the instantaneous signal frequency differs
signaling information from the carrier frequency in each direction
GLOSSARY ! 583
dibit system any digital modulation scheme density in a given direction as a given
that codes two bits of information per transmitting installation
transmitted symbol effective radiated power (ERP) transmitter
digital color code (DCC) signal transmitted by a power that would, if used with a lossless
cell site to identify that site to the mobile user dipole oriented for maximum gain, produce
digital signal processing (DSP) filtering of the same power density in a given direction
signals by converting them to digital form, as a given transmitting installation
performing arithmetic operations on the data electronic serial number (ESN) number
bits, then converting back to analog form assigned to a cell phone by the manufacturer
dipole any antenna with two sections as a security feature. It is transmitted with
the phone’s telephone number to authorize
direct-sequence spread spectrum technique for a call.
increasing the bandwidth of a transmitted
signal by combining it with a pseudo-random element an antenna used as part of an array
noise signal with a higher bit rate end office see central office
directional coupler device that allows signals to envelope imaginary pattern formed by
travel between ports in one direction only connecting the peaks of individual RF
waveforms in an amplitude-modulated signal
directivity gain of an antenna with losses
ignored equatorial orbit a satellite orbit that is entirely
above the equator
dispersion pulse spreading caused by variation
of propagation velocity with frequency Ethernet form of local-area network using
CSMA/CD and a logical bus structure
downlink signal path from a base station or
a satellite to a mobile station or a ground expansion restoration of the original dynamic
station range to a previously-compressed signal by
applying more gain to higher-amplitude
driver amplifier immediately preceding the
components than to lower-amplitude
power amplifier stage in a transmitter
components
dropped call a telephone connection that is
unintentionally terminated while in progress
fading reduction in radio signal strength,
dual-tone multi-frequency (DTMF) dialing usually caused by reflection or absorption
signaling using combinations of two audio of the signal
tones transmitted on the voice channel far-field region distance from an antenna
duplexer combination of filters to separate great enough to avoid local magnetic or
transmit and receive signals when both use electrical coupling, and great enough for
the same antenna simultaneously the antenna to resemble a point source
dynamic range ratio, usually expressed in fast associated control channel (FACCH) in a
decibels, between the strongest and weakest digital cellular system or PCS, control infor-
signals that can be present in a system mation that is transmitted by “stealing” bits
that are normally used for voice information
effective isotropic radiated power (EIRP) fiber-in-the-loop (FITL) use of optical fiber
transmitter power that would, if used with an for telephone connections to individual
isotropic radiator, produce the same power customers
584 ! GLOSSARY
flat-topped sampling sampling of an analog frequency-division multiple access (FDMA)

signal using a sample-and-hold circuit, such sharing of a communication channel among
that the sample has the same amplitude for multiple users by assigning each a different
its whole duration carrier frequency
foldover distortion see aliasing frequency-division multiplexing (FDM) com-
bining of several signals into one communi-
footprint depiction of the signal strength cation channel by assigning each a different
contours from a satellite transmitter on the carrier frequency
earth
frequency-shift keying (FSK) digital modula-
forward channel communication from a cell tion scheme using two or more different
site or repeater to the mobile unit output frequencies
Fourier series expression showing the structure full-duplex communication two-way
of a signal in the frequency domain communication in which both terminals can
framing bits bits added to a digital signal to transmit simultaneously
help the receiver to detect the beginning and
end of data frames
Gaussian minimum-shift keying (GMSK)
frequency diversity use of two or more RF variant of FSK which uses the minimum
channels to transmit the same information, possible frequency shift for a given bit rate
usually done to avoid fading geostationary orbit satellite orbit in which
frequency domain method of analyzing signals the satellite appears to remain stationary at a
by observing them on a power-frequency point above the equator
plane geosynchronous orbit satellite orbit in which
frequency hopping form of spread-spectrum the satellite’s period of revolution is equal to
communication in which the RF carrier con- the period of rotation of the earth
tinually moves from one frequency to another group velocity speed at which signals move
according to a prearranged pseudo-random down a transmission line or waveguide
pattern
frequency modulation modulation scheme in half-duplex communication two-way commu-
which the transmitted frequency varies in nication system in which only one station
accordance with the instantaneous amplitude can transmit at a time
of the information signal
handoff transfer of a call in progress from one
frequency modulation index peak phase shift in cell site to another
frequency-modulated signal, in radians
hemispheric beam antenna beam on a
frequency multiplier circuit whose output geostationary satellite that is adjusted to
frequency is an integer multiple of its input cover the whole earth
frequency
hybrid coil a specialized transformer (or its
frequency synthesizer device to produce electronic equivalent) that allows telephone
programmable output frequencies that are voice signals to travel in both directions
accurate and stable simultaneously on a single twisted-pair loop
GLOSSARY ! 585
image frequency reception of a spurious local area network a small data network,
frequency by a superheterodyne receiver, usually confined to a building or cluster of
resulting from mixing of the unwanted signal buildings
with the local oscillator signal to give the local loop in a telephone system, the wiring
intermediate frequency from the central office to an individual
Improved Mobile Telephone Service (IMTS) a customer
mobile telephone service, now obsolescent, local multipoint distribution system (LMDS)
using trunked channels but not cellular in network using microwaves for two-way
nature transmission of telephony, television, and
in-band signals control signals sent in a voice high-speed data
channel at voice frequencies low-earth-orbit (LEO) satellite an artificial
inbound channel communication channel from satellite orbiting the earth at an altitude less
mobile to base station than about 1500 kilometers
in-channel signals control signals using the

same channel as a voice signal mark in digital communication, a logic one
integrated services digital network (ISDN) medium earth orbit (MEO) satellite a satellite
telephone system using digital local loops for in orbit at a distance above the earth’s
both voice and data, with the codec in the surface of between approximately 8,000 and
telephone equipment 20,000 km
microcell in cellular radio, a small cell
intelligence information to be communicated
designed to cover a high-traffic area
interleaving changing the time order of digital
microstrip transmission line consisting of a
information before transmission to reduce the
circuit-board trace on one side of a dielectric
effect of burst errors in the channel
substrate (the circuit board) and a ground
International Mobile Subscriber Identification plane on the other side
(IMSI) in the GSM system, a telephone mixer nonlinear device designed to produce
number that is unique to a given user, sum and difference frequencies when
worldwide provided with two input signals
isotropic radiator an antenna that radiates all mobile identification number (MIN) number
power applied to it, equally in all directions. that identifies a mobile phone in a cellular
It is a theoretical construct and not a practical system; the mobile telephone number
possibility.
mobile switching center (MSC) switching
facility connecting cellular telephone base
justification addition of bits to a digital signal stations to each other and to the public
to compensate for differences in clock rates; switched telephone network
informally known as bit stuffing. mobile telephone switching office (MTSO) see
mobile switching center (MSC)
local access and transport area (LATA) in a modem acronym for modulator-demodulator;
telephone system, the area controlled by one device to enable data to be transmitted via
central office switch an analog channel
586 ! GLOSSARY
modulating signal the information signal that noise power density the power in a one-hertz
is used to modulate a carrier for transmission bandwidth due to a noise source
modulation process of using an information number assignment module (NAM) in a
signal to vary some aspect of a cellular phone, a memory location that
higher-frequency signal for transmission stores the telephone number(s) to be used on
the system
modulation index number indicating the degree
to which a signal is modulated
orbital satellite any satellite that is not in a
monopole an antenna with only one geostationary orbit
conductor, generally using ground or a ground
plane to represent a second conductor outbound channel a radio channel used
for communication from a base station to
multi-channel multipoint distribution system mobile stations
(MMDS) terrestrial microwave system for
out-of-band in telephone signaling, a control
the distribution of television, internet, and
signal that is outside the voice frequency
telephone services to businesses and
range
residences
overmodulation modulation to an extent
multipath distortion distortion of the informa- greater than that allowed for either technical
tion signal resulting from the difference in or regulatory reasons
arrival time in signals arriving via multiple
paths of different lengths
packet-switched network a communication
multiple access use of a single channel by more system that works using data divided into
than one transmitter relatively short transmissions called packets;
multiplexing use of a single channel by more these are routed through the system without
than one signal requiring a long-term connection between
sender and receiver
perigee the point in a satellite orbit that is
natural sampling sampling of an analog signal
closest to the earth
so that the sample amplitude follows that
of the original signal for the duration of the personal communication system (PCS) a
sample cellular telephone system designed mainly
for use with portable (hand-carried)
near/far effect in a spread-spectrum system, the
telephones
tendency for stronger signals to interfere with
the reception of weaker signals phase modulation communication system in
which the phase of a high-frequency carrier
near-field region the region of space close to is varied according to the amplitude of the
an antenna, where the radiation pattern is baseband (information) signal
disturbed by induced, as well as radiated,
phase velocity the apparent velocity of waves
electric and magnetic fields
along the wall of a waveguide
network an organized system for
phase-shift keying (PSK) digital modulation
communicating among terminals
scheme in which the phase of the transmit-
noise an unwanted random signal that extends ted signal is varied in accordance with the
over a considerable frequency spectrum baseband data signal
GLOSSARY ! 587
picocells very small cells in a cellular radio quadrature phase-shift keying (QPSK) digital
system modulation scheme using four different
point of presence (POP) place where one transmitted phase angles
telephone network connects to another quantizing representation of a continuously
polar orbit a satellite orbit passing over the varying quantity as one of a number of
north and south poles discrete values
private branch exchange (PBX) small telephone quantizing errors inaccuracies caused by the
switch located on customer premises representation of a continuously varying
processing gain improvement in interference quantity as one of a number of discrete
rejection due to spreading in a values
spread-spectrum system quantizing noise see quantizing errors
pseudo-random noise (PN) sequence a
transmitted series of ones and zeros that
radiation resistance representation of energy
repeats after a set time, and which appears
lost from an antenna by radiation as if it
random if the sequence is not known to the
were dissipated in a resistance
receiver
public switched telephone network (PSTN) the radio common carrier (RCC) a company that
ordinary public wireline phone system acts as a carrier of radiotelephone signals
pulse-amplitude modulation (PAM) a series of rake receiver a radio receiver that is capable
pulses in which the amplitude of each pulse of combining several received signals with
represents the amplitude of the information different time delays into one composite
signal at a given time signal
pulse-code modulation (PCM) a series of pulses Rayleigh fading variation in received signal
in which the amplitude of the information strength due to multipath propagation
signal at a given time is coded as a binary repeater a transmitter-receiver combination
number used to receive and retransmit a signal
pulse-duration modulation (PDM) a series of
reverse channel communication channel from
pulses in which the duration of each pulse
mobile station to base station
represents the amplitude of the information
signal at a given time ringback signal in telephony, a signal
pulse-position modulation (PPM) a series of generated at the central office and sent
pulses in which the timing of each pulse to the originating telephone to indicate
represents the amplitude of the information that the destination telephone is
signal at a given time ringing
pulse-width modulation (PWM) see roamer a cellular customer using a network
pulse-duration modulation (PDM) other than the subscriber’s local cellular
network
quadrature AM (QAM) modulation scheme in run-length encoding method of data compres-
which both the amplitude and phase of the sion by encoding the length of a string of
transmitted signal are varied by the baseband ones or zeros instead of transmitting all the
signal one or zero bits individually
588 ! GLOSSARY
sectorization division of a cell into several soft handoff in a cellular or PCS system,
sections all radiating outward from a cell site connecting a mobile to two or more base
stations simultaneously
selectivity ability of a receiver to discriminate
against unwanted signals and noise space binary zero
sensitivity ability of a receiver to detect weak spatial diversity use of two or more physically
signals with a satisfactory signal-to-noise ratio separated antennas at one end of a commu-
nication link
short messaging service (SMS) transmission of
brief text messages, such as pages or e-mail, by spectrum analyzer test instrument that typi-
a cellular radio or PCS system cally displays signal power as a function of
frequency
side frequencies frequency components
produced above and below the carrier splatter frequency components produced by a
frequency by the process of modulation transmitter that fall outside its assigned
channel
sideband a group of side frequencies above or
below the carrier frequency spot beam in a satellite system, a focused
beam of energy that covers a relatively small
sidetone in telephony, the presence in the re-
area on the earth, produced by a high-gain
ceiver of sounds picked up by the transmitter
antenna on the satellite
of the same telephone
spreading gain improvement in interference
signaling system seven (SS7) system used in
rejection due to spreading in a spread-
telephony which transmits all call setup
spectrum system
information on a packet-data network that is
separate from the voice channels used for spurious responses in a receiver, reception of
telephone conversations frequencies other than that to which it is
tuned
signal-to-noise ratio ratio between the signal
power and noise power at some point in a spurious signals unwanted signals accidentally
communication system produced by a transmitter
simplex a unidirectional communication star network a computer network topology in
system, for example, broadcasting which each terminal is connected to a
central mainframe or server
slope overload in delta modulation, an error
condition that occurs when the analog signal station class mark (SCM) code which describes
to be digitized varies too quickly for the the maximum power output of a cellular
system to follow phone
slow associated control channel (SACCH) in a store-and-forward repeater in digital commu-
digital cellular system or PCS, control infor- nication, a device which receives one or
mation that is transmitted along with the more data packets, stores them, and retrans-
voice mits them at a later time
smart card a card with an embedded integrated stripline transmission line consisting of a
circuit, that can be used for functions such circuit board having a ground plane on each
as storing subscriber information for a PCS side of the board, with a conducting trace in
system the center
GLOSSARY ! 589
subscriber ID module (SIM) in the GSM system, directional antenna on the ground, so that it
a “smart card” containing all user always points at the satellite
information, which is inserted into the phone transceiver a combination transmitter and
before use receiver
superheterodyne principle use of a mixer and transponder in satellite communication, a
local-oscillator combination to change the repeater located on a satellite
frequency of a signal
trunk lines transmission line carrying many
supervisory audio tone (SAT) in the AMPS signals, either on multiple pairs or multi-
system, a sine wave above the voice frequency plexed together on a single twisted-pair,
range, transmitted on the voice channel along coaxial cable, or optical fiber
with the voice, used by the base station to
detect loss of signal
uplink transmission channel to a satellite or
symbol in digital communication, the state of
base station
the signal at a sampling time
system identification number (SID) in the
AMPS system, a number transmitted by the variable-frequency oscillator (VFO) an
base station to identify the system operator oscillator whose frequency can be changed
easily, usually by means of a variable
capacitor or inductor
tandem office telephone switch that connects
only to other switches, and not to individual vocoder circuit for digitizing voice at a low
customers data rate by using knowledge of the way in
which voice sounds are produced
telepoint a very small cell used with some
cordless phones to allow their use in public voltage standing-wave ratio (VSWR) ratio
areas between maximum and minimum peak or
rms voltage on a transmission line
time-division duplexing (TDD) transmission of
data in two directions on a channel by using voltage-controlled oscillator (VCO) oscillator
different time slots for each direction whose frequency can be changed by
adjusting a control voltage
time-division multiple access (TDMA) system to
allow several transmissions to use a single
channel by assigning time slots to each Walsh code class of orthogonal spreading
time-division multiplexing (TDM) system to codes used in CDMA communication
combine several data streams onto a single waveguide metallic tube down which waves
channel by assigning time slots to each propagate
time domain representation of a signal as a wavelength distance a wave travels in one
function of time and some other parameter, period
such as voltage white noise noise containing all frequencies
tracking in satellite communication, with equal power in every hertz of
continuously adjusting the position of a bandwidth

Wireless Communication Technology

Uploaded by

Copyright:

Available Formats

Wireless Communication Technology

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Wireless Communication Technology

Uploaded by

Copyright:

Available Formats

Fundamental Concepts 1

' 1.1 Introduction

' 1.2 Brief History of Wireless Telecommunication

' 1.3 Elements of a Wireless Communication System

FIGURE 1.1 Elements of a communication system

Simplex Figure 1.1 represents a simplex communication system. The communica-

FIGURE 1.2 Full-duplex communication system

Some two-way communication systems do not require simultaneous

FIGURE 1.3 Half-duplex communication system

' 1.4 Signals and Noise

Modulated Given the necessity for modulating a higher-frequency signal with a

e(t) = E c sin(ω c t + θ) (1.1)

Noise It is not sufficient to transmit a signal from transmitter to receiver if the

T(K) = T(°C) + 273 (1.4)

T(K) = T(°C) + 273

Now substitute into Equation (1.3),

NF 2 = noise figure of the second stage

NF(dB) = 10 log NF (1.7)

Stage Power Gain Noise Figure

AT = A1A2A3 = 10 × 25 × 30 = 7500 = 38.8 dB

The noise figure is found from

The equivalent noise temperature is:

' 1.5 The Frequency Domain

FIGURE 1.5 Sine wave in time and frequency domains

When we move into the study of individual radio signals, frequency-

ƒ(t) = any well-behaved function of time. For our purposes, ƒ(t)

The radian frequency can be found from the time-domain representa-

The simplest ac signal is a sinusoid. The frequency-domain representa-

v(t) = sin (2000πt) V

TABLE 1.1 Fourier Series for Common Repetitive Waveforms

1. Half-wave rectified sine wave

V V 2V  cos 2ωt cos 4ωt cos 6ωt 

2. Full-wave rectified sine wave

2V 4V  cos 2ωt cos 4ωt cos 6ωt 

(b) With time zero at voltage peak

2V  2 cos 2ωt 2 cos 4ωt 2 cos 6ωt 

TABLE 1.1 (continued)

(b) Even function

Vτ 2Vτ  sin πτ/Τ sin 2 πτ/Τ sin 3πτ/Τ 

TABLE 1.1 (continued)

For this example, the above equation becomes:

The component at three times the fundamental frequency (3 kHz) has an

FIGURE 1.8 Addition of square-wave Fourier components with wrong phase

v(t) = 2.5 − 1.59 (sin(2π × 103t) + 0.5 sin (4π × 103t)

The first four voltage components are:

square wave can be obtained with a band-limited signal. In general, the

N 0 = noise power density in watts per hertz

The spectrum is shown in Figure 1.10(a). Note that the spectrum is a

' 1.6 The Radio-Frequency Spectrum

TABLE 1.2 The Radio-Frequency Spectrum

Frequency Frequency Wavelength Wavelength

Extremely High 30–300 GHz 1 mm–1 cm Millimeter Waves

Super High Frequency 3–30 GHz 1–10 cm Microwaves

Ultra High Frequency 300 Mhz–3 GHz 10 cm–1 m

Very High Frequency 30–300 MHz 1–10 m

High Frequency (HF) 3–30 MHz 10–100 m Short Waves

Medium Frequency 300 kHz–3 MHz 100 m–1 km Medium Waves

v = velocity of propagation of the wave in meters per second

Substituting the numbers from the problem we get

Obviously the type of baseband signal to be transmitted makes a great