Chapter I

10/5/2012
Course Handled by
Dr. S. Thai Subha
Assistant Professor (Sr. G)
Dept. of EEE
SRM University
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
DISCRETE TIME SIGNAL AND SYSTEMS

Characteristics and classification of signalsdiscrete time signal-basic definitions
Representation of signals, discrete time systems
Linear time invariant systems-properties of LTI
systems-linear constant coefficient difference
equations
Fourier transform of discrete time signals
Sampling techniques Nyquist rate, aliasing
effect.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
10/5/2012
Digital Signal Processing the processing

or manipulation of signals using digital
techniques
Input
Signal
ADC
Analogue
to Digital
Converter
Slide
10/5/2012
3
Digital
Signal
Processor
S.THAI SUBHA
DAC
Output
Signal
Digital to
Analogue
Converter
DSP, CHAPTER-I
Digital signal processing (DSP) is the mathematical manipulation

of an information signal to modify or improve it in some way.
It is characterized by the representation of discrete time, discrete
frequency, or other discrete domain signals by a sequence of
numbers or symbols and the processing of these signals.
Digital signal processing and analog signal processing are
subfields of signal processing.
DSP includes subfields like:
Audio and speech signal processing,
Sonar and radar signal processing,
Sensor array processing,
Spectral estimation,
Statistical signal processing,
10/5/2012
S.THAI SUBHA
Digital image processing,

Signal processing for communications,
Control of systems,
Biomedical signal processing,
Seismic data processing, etc.
DSP, CHAPTER-I
10/5/2012
Advantages of DSP
Programmable system allows flexibility
Tolerance
Storage
Implementation
Cost
Disadvantages of DSP
System complexity
Bandwidth
Power
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
Signals:
A signal is defined as a function of one or more variables which
conveys information.
A signal is a physical quantity that varies with time in general, or any
other independent variable.
It can be dependent on one or more independent variables.
Ex.
x1(t) = 20t
One-dimensional signal
x2(t) = 20t2
g(x,y) = x+3xy+2y2 Two-dimensional signal
N
x(t) = Ai(t)sin[2fi(t)t + i(t)];

Multi-dimensional
i=1
Ai(t) = time varying ith amplitude
fi(t) = time varying ith frequency
i(t) = time varying ith phase
*Dimensions- Based on the number of independent variables.

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
10/5/2012
Based on dimensions:
One- dimensional signal
Two- dimensional signal
Multi- dimensional signal
Based on channels used:

Single channel
Multi-channel
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
One dimensional signals
Two dimensional signal
x
Multi dimensional signal
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
Ir (x, y, t)
I(x, y, t) = Ig (x, y, t)
Ib (x, y, t)
8
10/5/2012
s1(t)
Sk (t) = s2 (t)
s3(t)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
Based on the fundamental properties;

Continuous-time signals & Discrete-time signals
Continuous-valued signals & Discrete-valued signals
Deterministic & Random signal
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
10
10/5/2012
Continuous-time signal: A signal x(t) is said to

be continuous if it is defined for all time t.
The amplitude of the signal varies with time.
All signals by nature are continuous time signals.
Discrete-time signal: A signal x(n) is said to be

discrete-time signal if it can be defined for a
discrete instant of time.
Amplitude of the signal varies at every discrete
values of n, which is generally uniformly spaced.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
11
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
12
10/5/2012
How to convert a continuous-time

signal to a discrete-time signal?
A continuous-time signal x(t) can be converted
to discrete-time signal by substituting t = nT
i.e., x(t) = x(nT)
x(n)
t = nT
where n constant integer, n = 0, 1, 2, 3, .
T sampling period
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
13
P. No.1:
The continuous time signal x(t) = 5sint for the
interval 3 t 0. Plot the corresponding discretetime signal with a sampling period T = 0.1s.
Solution:
x(t) = 5sint
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
14
10/5/2012
Continuous time signal
x(t) = 5sint
x(t) = 5sin( t)
5
4
3
2
x (t)
1
0
-1
-2
-3
-4
-5
0.5
1.5
2.5
time (t)
10/5/2012
S.THAI SUBHA
t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
DSP, CHAPTER-I
x(t)
0
2.9399
4.7561
4.7541
2.9348
-0.0063
-2.9451
-4.758
-4.7521
-2.9297
0.0126
2.9502
4.7599
4.7502
2.9246
-0.019
15
Discrete-time signal
x(t) = 5sin(t)
5
x(t)
-1
-2
-3
-4
-5
10/5/2012
0.5
S.THAI SUBHA
1.5
time (t)
DSP, CHAPTER-I
2.5
16
10/5/2012
Steps for converting

The discrete-time signal can be obtained as below;
x(t) = x(nT)
t = nT
x(t) = x(0.1n)
t = 0.1n
x(n) = 5sin(nT) = 5sin(0.1n) where n constant
integer, n = 0, 1, 2, 3, .
T sampling period
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
17
Discrete-time signal with a sampling period T = 0.1s

x(n) = 5sin(0.1n)
5
4
3
2
x (n )
1
0
-1
-2
-3
-4
-5
10
15
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
n
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
x(n)
0
1.5457
2.9399
4.0462
4.7561
5
4.7541
4.0425
2.9348
1.5397
-0.0063
-1.5517
-2.9451
-4.0499
-4.758
-5
18
10/5/2012
P. No.2: The continuous time signal x(t) = e-2t for

the interval 2 t -2. Plot the corresponding
discrete-time signal with a sampling period T =
0.1s.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
19
Continuous time signal

x(t) =exp(-2t)
x(t) = e-2t
60
t
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
50
x (t)
40
30
20
x(t)
54.5982
20.0855
7.3891
2.7183
1
0.3679
0.1353
0.0498
0.0183
10
0
-2
10/5/2012
-1.5
-1
-0.5
0.5
time (t)
S.THAI SUBHA
1.5
DSP, CHAPTER-I
20
10
10/5/2012
Steps for converting
The discrete-time signal can be obtained as below;

x(t) = x(nT)
t = nT
x(t) = x(0.1n)
t = 0.1n
x(n) = e-2(0.1n) = e-0.2n where n constant integer, n =
0, 1, 2, 3, .
T sampling period
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
21
Discrete-time signal with a sampling period T = 0.1s

3
x(n) = e-0.2n
n
-5
-4
-3
-2
-1
0
1
2
3
4
5
2.5
x(n )
1.5
0.5
0
-5
-4
-3
-2
-1
x(n)
2.7183
2.2255
1.8221
1.4918
1.2214
1
0.8187
0.6703
0.5488
0.4493
0.3679
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
22
11
10/5/2012
Ex.1: Sketch the signal x(t) = sin7t + sin10t for an

interval 0 t 2. Sample the signal with a
sampling period T = 0.2sec and sketch the
discrete-time signal.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
23
Continuous-Time sinusoidal signals:

xa(t) = Acos(t+),
< t <
A- amplitude of the sinusoid
- frequency in radian/s
- phase in radians
= 2F
xa(t) = Acos(2Ft+)
< t <
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
24
12
10/5/2012
Properties:
For every fixed value of the frequency F, xa(t) is
periodic, i.e., xa(t+T) = xa(t)
where T = 1/F is the fundamental period of the
sinusoidal signal.
Continuous-Time sinusoidal signals with distinct
frequencies are themselves distinct.
Increasing the frequency F results in an increase in
the rate of oscillation of the signal, i.e., more periods
are included in a given time interval.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
25
Discrete- Time Sinusoidal signals:

Discrete time signals may be expressed as
x(n) = Acos(n + ),
< t <
where n is an integer, called sample number,
A- Amplitude of the sinusoid,
- Frequency in radians/sample,
- Phase in radians.
Also,
x(n) = Acos(2fn + ),
where, f = F
Fs
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
26
13
10/5/2012
Properties:
1. A Discrete-time sinusoid is periodic only if its
frequency f is a rational number.
By definition, a discrete-time signal x(n) is periodic with
period N(N > 0) iff,
x(n+N) = x(n)
Smallest value of N Fundamental period.
Proof:
x(n) = Acos(n + )
x(n+N) = cos[2f0(N+n) + ] = cos[2f0n+ ]
2f0N = 2k
f0 = Nk
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
27
2. Discrete-time sinusoids whose frequencies

are separated by an integer multiple of 2 are
identical.
3. In Discrete-time sinusoids, high rate of
oscillation is attained when = (or) (-) or
equivalently f = 12 (or) 12
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
28
14
10/5/2012
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
29
t = nT = n
Fs
xa(t) = A cos(2Ft+)
x(n) = A cos(2fn+); where f =
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
F
Fs
30
15
10/5/2012
Relationship of frequency variable
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
31
Digital signal processing system
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
32
16
10/5/2012
1. Sampler
2. Quantizer
Sampler
3. DSP
4. DAC
Continuous-time sinusoid is sampled at regular

intervals & it is converted into the discrete-time
sinusoid.
Sequence of sampled values of original analog signal.
Amplitude is not restricted.
Quantizer
Sampled values are represented by a certain no. of
bits depending on the accuracy (only finite amplitude
level is possible)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
33
Periodic sampling of an analog signal
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
34
17
10/5/2012
Illustration of quantization
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
35
Sampling and quantization of a sinusoidal signal
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
36
18
10/5/2012
Consider two sinusoids,

x1(t) = cos210t
x2(t) = cos250t which are sampled at
a rate of Fs = 40Hz.
The corresponding discrete-time signals are
10

x1(n) = cos2 40 n = cos 2 n

x2(n) = cos250
40 n = cos 2 n = cos 2 n
x2(t) yields exactly the same values as x1(t) at Fs = 40Hz.

F2 = 50Hz is an alias of the frequency F1 = 10Hz at Fs=40Hz.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
37
Aliasing effect
Aliasing effect
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
38
19
10/5/2012
Nyquist rate relative to sampling

The Nyquist rate is the minimum sampling rate
required to avoid aliasing, equal to twice the
highest frequency contained within the signal.
Fs = 2F
where F is the highest frequency at which the signal
can have nonzero energy.
To avoid aliasing, the sampling rate must exceed

the Nyquist rate:
Fs > 2F
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
39
Consider the analog signal, xa(t) = 3 cos100t.

(a). Determine the minimum sampling rate
required to avoid aliasing.
(b). Suppose that the signal is sampled at the rate
Fs = 200Hz. What is the discrete time signal
after sampling.
(c). Suppose that the signal is sampled at the rate
Fs = 75Hz. What is the discrete time signal
after sampling.
F
(d). What is the frequency 0 < F < 2s of a
sinusoid that yields samples identical to those
obtained in part (c).
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
40
20
10/5/2012
Solution:
xa(t) = A cos(2Ft+)
(a). xa(t) = 3 cos(2*50)t
F = 50Hz
The minimum sampling rate required to avoid
aliasing is Fs = 100Hz.
(b). The discrete time signal after sampling at Fs
= 200Hz is,
100
x a ( n ) = 3 cos
n = 3 cos n
200
2
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
41
(c). The discrete time signal after sampling at Fs =

75Hz is,
100
4
x ( n ) = 3 cos
n = 3 cos
n
75
3
2
x (n ) = 3 cos n
3
(d). For the sampling rate of Fs = 75Hz, we have

F = fFs = 75f
The frequency of the sinusoid in part (c) is f = 1/3.
Hence, F = 25Hz.
The sinusoidal signal
y a ( t ) = 3 cos(2Ft ) = 3 cos 50t
Hence, F = 50Hz is an alias of F = 25Hz when sampled
at Fs = 75Hz.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
42
21
10/5/2012
Representation of signals
Discrete-time sequence may be represented in many
ways:
(i). Functional representation:
2 for n = 1, 3, 5
x(n) = 1 for n = -1, -2, 4, 7
0 , otherwise
(ii). Tabular representation

- Represents origin
(iii). Sequence representation
Finite duration, x(n) = {-2, -1, 3, 1, 0, 4, 2}
Infinite duration, x(n) ={0, 0, 1, -1, 2, 4, 1, 0, 0..}
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
43
Basic sequences
Unit sample sequence (Unit Impulse sequence)
contains only one non-zero valued element & it is defined
as,
Plays role in digital filters
(n)= 1 , for n = 0
0 , otherwise
Unit step sequence

u(n) = 1, n 0
otherwise
0,
used to define the starting point of the sequence in analytic
expressions. Ex.
x(n) = a n , n 0
otherwise
0,
can be written as x(n)=anu(n), for all n
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
44
22
10/5/2012
Unit sample sequence
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
45
Unit step sequence
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
46
23
10/5/2012
Unit ramp sequence:

ur(n) = n, for n 0
0, otherwise
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
47
Classification of discrete-time signals

Energy & Power signals:
The energy of a signal x(n) is defined as
E =
[x ( n ) ]
n =
If E is finite, [0 E < ] then x(n) is called an

energy signal. (P = 0)
Many signals that posses infinite energy have finite
average power.
Pavg =
Lim
N
10/5/2012
S.THAI SUBHA
N
1
2
x (n )
2 N + 1 n = N
DSP, CHAPTER-I
48
24
10/5/2012
The average power of periodic sequence with

period N is given by
Pavg =
1
N
N 1
x (n )
n =0
If P is finite, then the signal is known as power

signal. (E= )
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
49
Problem:
Determine the power & energy of the unit step
sequence.
The average power of the unit step signal is
1
P = Lim
u (n )
2N + 1
N
avg
Pavg =
Lim
N
n =0
N
1
1 =
2 N + 1 n =0
1+
Pavg =
1
N
1
N
Lim 2 N + 1 ( N + 1)
N
Lim 2 + = 2
Pavg is finite. Unit step signal is a power signal.
For power signal, E= , prove it.
N
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
50
25
10/5/2012
Problem:
Test whether the given signal is an energy signal
or a power signal. x(n) = (-0.5)nu(n)
E =
x (n )
= ( 0 . 5) n u ( n )
n =
n =0
1
4
= (0.5) =
= <
1 0.25 3
n =0
2n
a
n =0
1
1 a
x(n) is energy signal.
10/5/2012
S.THAI SUBHA
P=
Lim
N
DSP, CHAPTER-I
51
N
1
[x ( n ) ]2
2 N + 1 n= N
2
1 N
P = Lim
(0.5) n u(n)
N 2N + 1 n =0
N
1
1
= Lim
(0.25) n = = 0
N 2 N + 1 n =0
x(n) is an energy signal. P = 0.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
52
26
10/5/2012
A continuous-time signal is said to be periodic if

x(t) = x(t+T), T > 0, for all values of T.
where T= period of cycle, which is an integer value
x(t) = x(t+T) = x(t+2T) = x(t+3T) = .= x(t+T) &
n = integer.
Hence a periodic signal with period T > 0 is also
periodic with period nT.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
53
Problem: Test whether the given signals are

periodic or not. (i) x(t) = esin(t) (ii) x(t) = tesin(t)
Solution:
x(t) = esin(t)
From the definition of periodicity, x(t) = x(t+T) for T > 0.
Substitute t = t+T, (T = 2)
sin(t+2)=sint
x(t+T) = esin(t+T) = esin(t) = x(t)
Hence, the signal x(t) = esin(t) is periodic
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
54
27
10/5/2012
(ii). x(t) = tesin(t)

From the definition of periodicity, x(t) = x(t+T) for T > 0.
Substitute t = t+T, (T = 2),

(sin(t+2)=sint)
x(t+T) = (t+T)esin(t+T) = (t+T)esin(t) 0
Hence, the signal x(t) = tesin(t) is aperiodic.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
55
Pr.: Determine whether or not each of the

following signals is periodic. If a signal is
periodic, specify its fundamental 3period.
1
j n +
5 2
j6n
(i). x(n) = e
(ii). x(n) =e
(iii). x(n) = cos 23 n

(iv). x(n) = cos 3 n+cos 34 n
Solution:
(i). x(n) = ej6n - Periodic
0N = 2k
N-integer
N = 2 k = 26k
k-integer
Minimum value of N is 1.
0
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
56
28
10/5/2012
3
5
(ii). x(n) is not periodic

(iii). x(n) = cos 23 n
N =
is not a multiple of .
2 k
2k * 3
=
0
2
Minimum value of N is 3.
2 k
2k * 3
N =
=
(iv). N1= 6
2 k
2 k * 4
N2= 8
N =
=
1
N
N
1
2
x(n) = cos 3 n+cos3 n

4
6
3
=
8
4
N = 4N1 = 3N2 = 24
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
57
For an Even signal x(n) = x(-n)

For an odd signal x(-n) = -x(n)
Any signal can be written as x(n) = xe(n) + xo(n)
xe(n)= 12 [x(n) + x(-n)]
xo(n)= 12 [x(n) - x(-n)]
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
58
29
10/5/2012
Example of even and odd signal
(a) and (b) are even

(c) is odd
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
59
Pr.: Find the odd and even component on x(t) = ej2t

x(t) = xe(t) + xo(t)= ej2t
xe(t)= 12 [x(t) + x(-t)] = 12 [ej2t + e-j2t ]= cos2t
xo(t)= 12 [x(t) - x(-t)] = 2jj [ej2t - e-j2t]= jsin2t
Pr.: Find the odd and even components of x(n) = {3,2,1,4,5}
Position -2
-1
X(n)
xe(n)= 12 [x(n) + x(-n)]

For n = -2, xe(-2)= 12 [x(-2) + x(-(-2))]=4
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
60
30
10/5/2012
For n = -1, xe(-1)= 12 [x(-1) + x(-(-1))]=3

For n = 0, xe(0)= 12 [x(0) + x(0)]= 1
For n = 1, xe(1)= 12 [x(1) + x(-1)]=3
For n = 2, xe(2)= 12 [x(2) + x(-2)]=4
xe(n)= {4, 3, 1, 3, 4}
For odd part, xo(n)= 12 [x(n) - x(-n)]
xo(n)= {-1, -1, 0, 1, 1}

Adding xe(n) & xo(n) position-wise we obtain the
origin signal.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
61
Simple manipulations of discrete-time

signals
Operations performed on dependent variables:
Amplitude modifications include addition,
multiplication, and scaling of discrete-time
signals.
Amplitude scaling of a signal by a constant A is
accomplished by multiplying the value of every
signal sample by A, i.e.,
< n <
y(n) = A * x(n),
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
62
31
10/5/2012
Simple manipulations of discrete-time signals
Addition & Subtraction, y(n) = x1(n)x2(n) < n <

Multiplying two signals, y(n) = x1(n) * x2(n)
Any operations listed above does not affect the
period of the signal. It only acts upon the
magnitude of the signal.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
63
Operations performed on independent variables:

Time reversal or Folding,
y(n) = FD[x(n)] = x(-n)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
64
32
10/5/2012
Time shifting of signals:

A sequence x(n) may be shifted in time by
replacing independent variable n by n-k, where k
is an integer.
If k is +ve, time shift results in a delay of the
sequence by k units of time. (shifting right)
If k is -ve, time shift results in an advance of the
sequence by k units of time. (shifting left)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
65
Pr.: A sequence is given by

x(n) = {0 0 1 1 2 1 3 1 0 0 } determine x(n-2)
& x(n+1).
Solution:
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
66
33
10/5/2012
x(n-2) = {0 0 1 1 2 1 3 1 0 0 }
x(n+1)= {0 0 1 1 2 1 3 1 0 0 }
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
67
Pr.: Represent x(-n) & x(-n+3) graphically for the

x(n) shown below.
FD [ x ( n )] = x ( n k )
FD [ x ( n )] = x ( ( n k )) = x ( n + k )
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
68
34
10/5/2012
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
69
Time scaling of signals:

Replace n by n.
This modifications is known as Time scaling or
down-sampling.
Pr.: Show the graphical representation of

the signal y(n) = x(2n), where x(n) is shown
in figure.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
70
35
10/5/2012
Given, y(n) = x(2n)

y(0) = x(0)
y(1) = x(2)
y(2) = x(4) & so on.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
71
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
72
36
10/5/2012
y ( n ) = x ( n )
Transformation operator
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
73
Linear & non-linear systems

Time-variant & Time-Invariant systems
Static & Dynamic systems
Causal & noncausal sytems
Stable & unstable systems
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
74
37
10/5/2012

Linear systems satisfies superposition
principle
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
75
The statement of superposition principle can be

decomposed into two parts. If a2=0 & a1, x1 are
simply replaced by a, x respy. then
[ax ( n ) ] = a[x ( n ) ] (Scaling property)
Also, if a1=a2=1, then
[x 1 (n) + x 2 (n)] = [x 1 (n )] + [x 2 (n)]
(Additive property)
Also, if a1= a2= 0, for x(n) = 0, if y(n) = 0
Relaxed system.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
76
38
10/5/2012
Pr.: Determine whether the system given below

are linear or non-linear.
(1). y(n) = [x(n)] = x(-n)
[a 1 x 1 (n) + a 2 x 2 (n)] = [a 1 x 1 (n )] + [a 2 x 2 (n )]
x(n)
System
y(n) = x(-n)
Linear combination of the inputs:

x1(n)
x2(n)
10/5/2012
a1
System
a2
yI(n) = a1x1(-n)+a2x2(-n)
S.THAI SUBHA
DSP, CHAPTER-I
77
Linear combination of the outputs:

x1(n)
x2(n)
System
y1(n) = x1(-n)
a1
a2
System
yO(n) = a1x1(-n)+a2x2(-n)
y2(n) = x2(-n)
yI(n) = yO(n) indicating the given system is linear.

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
78
39
10/5/2012
(2). y(n) = Ax(n)+B

x(n)
System
y(n) = Ax(n)+B

x1(n)
x2(n)
10/5/2012
a1
a2
System
yI(n) = A[a1x1(n)+a2x2(n)]+B
S.THAI SUBHA
DSP, CHAPTER-I
79

x1(n)
x2(n)
System
System
y1(n) = Ax1(n)+B
a1
yO(n) = a1[Ax1(n)+B]+
a2[Ax2(n)+B]
y2(n) = Ax2(n)+B
a2
yI(n) yO(n) indicating the given system is non- linear.

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
80
40
10/5/2012
(3). y(n) = [x(n)] = x (k )

k =n 0
x(n)
System
y(n) = x ( k )
k =n 0

x1(n)
x2(n)
a1
a2
System
n
yI(n) = [a 1 x 1 (k ) + a 2 x 2 (k )]
k =n 0
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
81

n
x1(n)
x2(n)
System
System
y1(n) =
x
k =n 0
(k )
a1
n
k =n 0
k=n0
yO(n) = a 1 x 1 ( k ) + a 2 x 2 ( k )
a2
y2(n) = x 2 ( k )
k =n 0

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
82
41
10/5/2012
EX.: (4). y(n) = x(n) + u(n+1)
(5). y(n) = x(2n)

Ans: (4). Non-linear
(5). Linear
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
83
Time-variant & Time- Invariant (shift variant)

systems:
y(n k)
If x ( n )
y ( n ) , then x(n k)
If the output y(n, k) = y(n-k) for all possible value
of k, then the system is called Time-Invariant.
If the output y(n, k) y(n-k) for all possible value
of k, then the system is called Time-variant.
x(n)
Delay
System
10/5/2012
System
y(n)
S.THAI SUBHA
Delay
DSP, CHAPTER-I
y(n, k)
y(n-k)
84
42
10/5/2012
Ex.: Examine the following system whether

time-variant or time-invariant.
(1). y(n) = x(2n)
x(n)
x(n-k)
Delay
System
System
y(n)=
x(2n)
x(2n-k)
Delay
y(n, k)
x(2n-2k)
y(n-k)
y(n, k) y(n-k), the system is time-variant.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
85
(2). y(n) = x(n)sin0n

x(n)
Delay
System
x(n-k)
x(n)sin 0n
System
Delay
y(n, k)= x(n-k)sin 0n
y(n-k)= x(n-k)sin 0(n-k)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
86
43
10/5/2012
(3). y(n) = [x(n)] = x(-n)

x(n)
Delay
System
x(n-k)
x(-n)
System
Delay
y(n, k)= x(-n-k)
y(n-k)= x(-(n-k))=x(-n+k)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
87
(4). y(n) = [x(n)] = ax(n) + b

x(n)
Delay
System
x(n-k)
ax(n)+b
System
Delay
y(n, k)= ax(n-k)+b
y(n-k)= ax(n-k)+b
y(n, k) = y(n-k), the system is time-invariant.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
88
44
10/5/2012
(5). y(n) = [x(n)] =

x(n)
Delay
x (L )
L=n 0
x(n-k)
System
n
y(n, k)= x ( L k )
x (L )
System
L=n 0
L=n0
Delay
y(n-k)= x ( L k )
L=n 0
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
89
DSP, CHAPTER-I
90
DSP 4th ed. by P. Ramesh Babu

Ex. 1.4
Practice problem 1.3
Ex. 1.5
Ex. 1.8
Ex. 1.9
Ex. 1.11
10/5/2012
S.THAI SUBHA
45
10/5/2012
Static & Dynamic system

Output depends only on the present I/P
sample (n) & not on past (n-1) or future (n+1)
samples of I/P- Static (memory less)
Output depends on past, present & future
samples of I/P- Dynamic (with memory)
Ex.: (1). y(n) = ax(n)+bnx2(n) Static
(2). y(n) = x(n-1)+x(n)+3x(n+1) Dynamic
(3). y(n) = x (n k ) Dynamic
k =0
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
91
Causal & non causal system

Causal system:
y(n) = f[x(n), x(n-1), x(n-2),]
Non causal system:
The system does not satisfy the above condition.
Ex.: (1). y(n) = x(-n)

For n=-1, y(-1)=x(1)(future sample) - noncausal system
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
92
46
10/5/2012
(2). y(n) = ax(n)+b Causal

(3). y(n) = x(n)+u(n+1) Causal
n
(4). y(n) = x ( k ) Non-Causal

For n < n0; i.e., n=1, n0=4, y(1)= x ( k )
y(1)=x(4)+x(3)+x(2)+x(1)
k=n0
k=4
Future samples
(5). y(n) = x(2n) Non-Causal

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
93
Stable & Unstable systems

An arbitrary relaxed system is said to be BIBO
stable, iff every bounded input produces a
bounded output.
If x(n) is bounded, there exists a constant Mx
such that x ( n ) M < for all n.
If y(n) is bounded, there exists a constant My
such that y ( n ) M < for all n.
Any LTI system is stable if the impulse
response is absolutely summable.
x
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
94
47
10/5/2012
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
95
Test the stability of the following systems:

(a). y(n) = cos[x(n)]
For stability of a system the impulse response
should be absolutely summable.
h (k ) <
k =
Impulse response is obtained when the input is unit

impulse signal (n).
i.e., if x(n) = (n), then y(n) = h(n)
(n)= 1, for n = 0
0 , otherwise
h(n) = cos [(n)]
h(k) < h() + .....+ h(2) + h(1) + h(0) + h(1) + h(2) + ......+ h()
k =
Since, response tends to infinity the system is unstable

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
96
48
10/5/2012
(b). y(n) = x(-n-2)

x(n) = (n), then y(n) = h(n)
h(n) = (-n-2)
h(k) < h() + .....+ h(2) + h(1) + h(0) + h(1) + h(2) + ......+ h()
k =
Summation will be zero in all the places, except

when n=-2.
h ( k ) = 1 < Hence the system is stable.
k =
Ex: y(n) = nx(n).

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
97
Impulse response
The response to the unit sample sequence
{(n)} is called unit sample sequence or
impulse response denoted by {h(n)}.
Ex. If y(n, k) is a response of the system to the
I/P, unit sample sequence at n = k, then h(n, k)
= y(n, k) = [(n-k)].
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
98
49
10/5/2012
LTI systems are characterized in the time domain simply

by their response to a unit sample sequence.
Any arbitrary input signal can be decomposed &
represented as a weighted sum of unit sample sequence.
In terms of unit sample response of the system.
The general form of the expression that relates

the unit sample response of the system & the
arbitrary Input signal to the Output signal, is
called as Convolution sum of Convolution
formula.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
99
Any DTS can be represented as a sequence of

individual impulses.
x ( n ) = x ( k ) ( n k )
k =
Ex.: A finite duration sequence x(n) is given by

x(n) = {0.5, 1.5, 0, -1, 1, 0.75, 2}
Resolve the above sequence into a sum of
weighted impulse response.
Solution:
x(n) = 0.5(n+2)+1.5(n+1)-(n-1)+(n2)+0.75(n-3)+2(n-4)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
100
50
10/5/2012
Let h(n, k) denote the response of the linear

system to the shifted unit impulse (n-k), i.e.,
h(n, k) = [(n-k)]
From superposition property of a linear
system, the response y(n, k) of the linear
system to the input x(n) is the weighted linear
combination of the basic responses.

y( n ) = x ( k ) ( n k )
k =
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
101
y(n) = x(k)(n k)
k=
y(n ) =
x (k )[(n k )]
k =
y(n ) = x (n ) * h (n )
x (k )h (n , k )
Thus, the response of a linear system to the
set of shifted unit impulses, we can construct
the response to an arbitrary I/P.
y(n ) =
k =
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
102
51
10/5/2012
Properties of convolution:
(i). Commutative law: x(n)*h(n) = h(n)*x(n)
(ii). Associative law: [x(n)*h1(n)]*h2(n) =
x(n)*[h1(n)*h2(n)]
(iii). Distributive law: x(n)*[h1(n)*h2(n)] =
x(n)*h1(n)+x(n)*h2(n)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
103
Causality
y(n ) =
x (k )h (n, k )
k =
An LTI is causal iff its impulse response is zero for

negative values of n
n
y(n ) =
x (k )h (n, k )
k =0
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
104
52
10/5/2012
Test the stability of the system of the system

whose impulse response is h (n ) = 1 u (n )
2
Solution:
Condition: h ( k ) <
n
k =
1
h (n ) = u (n )
n =
n = 2
n =
10/5/2012
n =0
1
1 a
h (n ) = 1
n =0
1 1
1
= 1 + + ....
2 2
n =0 2
h (n ) =
1
2
=2
S.THAI SUBHA
Stable
DSP, CHAPTER-I
105
Steps for convolution process

Choose the initial value of n. If x(n) starts at n = n1
& h(n) starts at n = n2, then n1 + n2 can be the
initial value of n.
Represent the input sequence x(k) & the impulse
response h(k) using k as time index.
Fold h(k) to obtain h(-k).
Shift by n to the right if n is positive & left if n is
negative to obtain h(n-k).
Multiply x(k) & h(n-k) to obtain y(n).
Increment n & obtain y(n), till y(n) becomes zero.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
106
53
10/5/2012
Pr.: Determine the output y(n) of a LTI system

with impulse response h(n)={6, 5, 4, 3, 2, 1}
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
107
y ( n ) = x ( n )
Transformation operator
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
108
54
10/5/2012

Time-variant & Time-Invariant systems
Static & Dynamic systems
Causal & noncausal sytems
Stable & unstable systems
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
109

Linear systems satisfies superposition
principle
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
110
55
10/5/2012
The statement of superposition principle can be

decomposed into two parts. If a2=0 & a1, x1 are
simply replaced by a, x respy. then
[ax ( n ) ] = a[x ( n ) ] (Scaling property)
Also, if a1=a2=1, then
[x 1 (n) + x 2 (n)] = [x 1 (n )] + [x 2 (n)]
(Additive property)
Also, if a1= a2= 0, for x(n) = 0, if y(n) = 0
Relaxed system.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
111
Pr.: Determine whether the system given below

are linear or non-linear.
(1). y(n) = [x(n)] = x(-n)
[a 1 x 1 (n) + a 2 x 2 (n)] = [a 1 x 1 (n )] + [a 2 x 2 (n )]
x(n)
System
y(n) = x(-n)

x1(n)
x2(n)
10/5/2012
a1
a2
System
yI(n) = a1x1(-n)+a2x2(-n)
S.THAI SUBHA
DSP, CHAPTER-I
112
56
10/5/2012

x1(n)
x2(n)
System
y1(n) = x1(-n)
a1
a2
System
yO(n) = a1x1(-n)+a2x2(-n)
y2(n) = x2(-n)

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
113
(2). y(n) = Ax(n)+B

x(n)
System
y(n) = Ax(n)+B

x1(n)
x2(n)
10/5/2012
a1
a2
System
yI(n) = A[a1x1(n)+a2x2(n)]+B
S.THAI SUBHA
DSP, CHAPTER-I
114
57
10/5/2012

x1(n)
x2(n)
System
System
y1(n) = Ax1(n)+B
a1
yO(n) = a1[Ax1(n)+B]+
a2[Ax2(n)+B]
y2(n) = Ax2(n)+B
a2
yI(n) yO(n) indicating the given system is non- linear.

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
115
(3). y(n) = [x(n)] = x (k )

k =n 0
x(n)
System
y(n) = x ( k )
k =n 0

x1(n)
x2(n)
a1
a2
System
n
yI(n) = [a 1 x 1 (k ) + a 2 x 2 (k )]
k =n 0
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
116
58
10/5/2012

n
x1(n)
x2(n)
System
System
y1(n) =
x
k =n 0
(k )
a1
n
k =n 0
k=n0
yO(n) = a 1 x 1 ( k ) + a 2 x 2 ( k )
a2
y2(n) = x 2 ( k )
k =n 0

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
117
EX.: (4). y(n) = x(n) + u(n+1)
(5). y(n) = x(2n)

Ans: (4). Non-linear
(5). Linear
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
118
59
10/5/2012
Time-variant & Time- Invariant (shift variant)

systems:
y(n k)
If x ( n )
y ( n ) , then x(n k)
If the output y(n, k) = y(n-k) for all possible value
of k, then the system is called Time-Invariant.
If the output y(n, k) y(n-k) for all possible value
of k, then the system is called Time-variant.
x(n)
Delay
System
System
10/5/2012
y(n)
S.THAI SUBHA
y(n, k)
Delay
y(n-k)
DSP, CHAPTER-I
119
Ex.: Examine the following system whether

time-variant or time-invariant.
(1). y(n) = x(2n)
x(n)
x(n-k)
Delay
System
y(n)=
x(2n)
System
x(2n-k)
Delay
y(n, k)
x(2n-2k)
y(n-k)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
120
60
10/5/2012
x(n)
x(n-k)
Delay
System
10/5/2012
System
y(n)=
x(n2)
S.THAI SUBHA
Delay
x(n2-k)
y(n, k)
x(n2+k2-2kn)
y(n-k)
DSP, CHAPTER-I
121
(2). y(n) = x(n)sin0n

x(n)
Delay
System
x(n-k)
x(n)sin 0n
System
Delay
y(n, k)= x(n-k)sin 0n
y(n-k)= x(n-k)sin 0(n-k)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
122
61
10/5/2012
(3). y(n) = [x(n)] = x(-n)

x(n)
Delay
System
x(n-k)
x(-n)
System
Delay
y(n, k)= x(-n-k)
y(n-k)= x(-(n-k))=x(-n+k)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
123
(4). y(n) = [x(n)] = ax(n) + b

x(n)
Delay
System
x(n-k)
ax(n)+b
System
Delay
y(n, k)= ax(n-k)+b
y(n-k)= ax(n-k)+b
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
124
62
10/5/2012
(5). y(n) = [x(n)] = x ( L )

L=n0
x(n)
Delay
x(n-k)
System
n
y(n, k)= x ( L k )
x(L)
System
L=n 0
L=n0
Delay
y(n-k)= x ( L k )
L=n 0
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
125
Static & Dynamic system

Output depends only on the present I/P
sample (n) & not on past (n-1) or future (n+1)
samples of I/P- Static (memory less)
Output depends on past, present & future
samples of I/P- Dynamic (with memory)
Ex.: (1). y(n) = ax(n)+bnx2(n) Static
(2). y(n) = x(n-1)+x(n)+3x(n+1) Dynamic
(3). y(n) = x (n k ) Dynamic
k =0
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
126
63
10/5/2012
Causal & non causal system

Causal system:
y(n) = f[x(n), x(n-1), x(n-2),]
Non causal system:
The system does not satisfy the above condition.
Ex.: (1). y(n) = x(-n)

For n=-1, y(-1)=x(1)(future sample) - noncausal system
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
127
p
(2). y(n) = ax(n)+b Causal
(3). y(n) = x(n)+u(n+1) Causal
n
(4). y(n) = x ( k ) Non-Causal

For n < n0; i.e., n=1, n0=4, y(1)= x ( k )
y(1)=x(4)+x(3)+x(2)+x(1)
k=n0
k=4
Future samples
(5). y(n) = x(2n) Non-Causal

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
128
64
10/5/2012
Stable & Unstable systems

An arbitrary relaxed system is said to be BIBO
stable, iff every bounded input produces a
bounded output.
If x(n) is bounded, there exists a constant Mx
such that x ( n ) M < for all n.
If y(n) is bounded, there exists a constant My
such that y ( n ) M < for all n.
Any LTI system is stable if the impulse
response is absolutely summable.
x
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
129
Recall:
Impulse response:
The response to the unit sample sequence
{(n)} is called unit sample sequence or
impulse response denoted by {h(n)}.
Ex. If y(n, k) is a response of the system to the
I/P, unit sample sequence at n = k, then h(n, k)
= y(n, k) = [(n-k)].
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
130
65
10/5/2012
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
131
Test the stability of the following systems:

(a). y(n) = cos[x(n)]
For stability of a system the impulse response should
be absolutely summable.
h (k ) <
k =
Impulse response is obtained when the input is unit

impulse signal (n).
i.e., if x(n) = (n), then y(n) = h(n)

(n)= 1, for n = 0
0 , otherwise
h(n) = cos [(n)]
h(k) < h() + .....+ h(2) + h(1) + h(0) + h(1) + h(2) + ......+ h()
k =
Since, response tends to infinity the system is unstable

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
132
66
10/5/2012
(b). y(n) = x(-n-2)

x(n) = (n), then y(n) = h(n)
h(n) = (-n-2)
h(k) < h() + .....+ h(2) + h(1) + h(0) + h(1) + h(2) + ......+ h()
k =
Summation will be zero in all the places, except

when n = -2.
h ( k ) = 1 < Hence the system is stable.
k =
Ex: y(n) = nx(n).

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
133
LTI systems are characterized in the time domain simply

by their response to a unit sample sequence.
Any arbitrary input signal can be decomposed &
represented as a weighted sum of unit sample sequence.
In terms of unit sample response of the system.
The general form of the expression that relates

the unit sample response of the system & the
arbitrary Input signal to the Output signal, is
called as Convolution sum of Convolution
formula.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
134
67
10/5/2012
Any DTS can be represented as a sequence of

individual impulses.
x ( n ) = x ( k ) ( n k )
k =
Ex.: A finite duration sequence x(n) is given by x(n)

= {0.5, 1.5, 0, -1, 1, 0.75, 2}
Resolve the above sequence into a sum of
weighted impulse response.
Solution:
x(n) = 0.5(n+2)+1.5(n+1)-(n-1)+
(n-2)+0.75(n-3)+2(n-4)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
135
Let h(n, k) denote the response of the linear

system to the shifted unit impulse (n-k), i.e.,
h(n, k) = [(n-k)]
From superposition property of a linear
system, the response y(n, k) of the linear
system to the input x(n) is the weighted linear
combination of the basic responses.

y( n ) = x ( k ) ( n k )
k =
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
136
68
10/5/2012
y(n) = x(k)(n k)
k=
y(n ) =
x (k )[(n k )]
k =
y(n ) = x (n ) * h (n )
x (k )h (n , k )
Thus, with the response of a linear system to
the set of shifted unit impulses, we can
construct the response to an arbitrary I/P.
y(n ) =
k =
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
137
Properties of LTI systems (convolution):

(i). Commutative law: x(n)*h(n) = h(n)*x(n)
(ii). Associative law: [x(n)*h1(n)]*h2(n) =
x(n)*[h1(n)*h2(n)]
(iii). Distributive law: x(n)*[h1(n)+h2(n)] =
x(n)*h1(n)+x(n)*h2(n)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
138
69
10/5/2012
Causality
x (k )h (n, k )
y(n ) =
k =
An LTI is causal iff its impulse response is zero for

negative values of n
n
y(n ) =
x (k )h (n, k )
k =0
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
139
Test the stability of the system of the system

whose impulse response is h (n ) = 1 u (n )
2
Solution:
Condition: h ( k ) <
n
k =
1
h (n ) = u (n )
n =
n = 2
n =0
1
1 a
1 1
1
h ( n ) = = 1 + + ....
2 2
n =
n =0 2
h (n ) = 1
n =0
10/5/2012
1
2
=2
S.THAI SUBHA
Stable
DSP, CHAPTER-I
140
70
10/5/2012
Choose the initial value of n. If x(n) starts at n = n1 & h(n)

starts at n = n2, then n1 + n2 can be the initial value of n.
Represent the input sequence x(k) & the impulse response
h(k) using k as time index.
Fold h(k) to obtain h(-k).
Shift by n to the right if n is positive & left if n is negative
to obtain h(n-k).
Multiply x(k) & h(n-k) & sum all the values to obtain
y(n).
Increment n & obtain y(n), till y(n) becomes zero.
Check: No. of samples in the output: N1 + N2 - 1
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
141
Pr.: Determine the output y(n) of LTI system with

impulse response h(n) = {6, 5, 4, 3, 2, 1} when
the input is x(n) = {1, 1, 1, 1}
Solution:
y(n ) = x (k )h (n k )
k =
h(-k) = {1, 2, 3, 4, 5, 6}; x(k) = {1, 1, 1, 1}
y (0) = x ( k ) h ( k ) = x (0) h (0) + x (1) h ( 1) + .... = 6

k =0
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
142
71
10/5/2012
h(n) = {6, 5, 4, 3, 2, 1}; x(n) = {1, 1, 1, 1}
y(1) = x (k )h (1 k ) = x (0)h (1) + x (1)h (0) + .... = 5 + 6 = 11

k =0
y(2) = x (k )h (2 k ) = x (0)h (2) + x (1)h (1) + x (2)h (0) + ...

k =0
= 4 + 5 + 6 = 15
y(3) = x (k )h (3 k ) = x (0)h (3) + x (1)h (2) + x (2)h (1) + x (3)h (0) + ...
k =0
= 3 + 4 + 5 + 6 = 18
y(4) = x (k )h (4 k ) = x (0)h (4) + x (1)h (3) + x (2)h (2) + x (3)h (1) + ...
k =0
= 2 + 3 + 4 + 5 = 14
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
143
y(5) = x (k )h (5 k ) = x (0)h (5) + x (1)h (4) + x (2)h (3) + x (3)h (2) + ...
k =0
= 1 + 2 + 3 + 4 = 10
y(6) = x (k )h (6 k ) = x (0)h (6) + x (1)h (5) + x (2)h (4) + x (3)h (3) + ...
k =0
= 1+ 2 + 3 = 6
y(7) = x (k )h (7 k ) = x (2)h (5) + x (3)h (4) + ... = 1 + 2 = 3

k =0
y(8) = x (k )h (8 k ) = x (3)h (5) + ... = 1

k =0
y(9) = y(10) = 0
y(n) = {6, 11, 15, 18, 14, 10, 6, 3, 1, 0}
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
144
72
10/5/2012
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
145
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
146
73
10/5/2012
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
147
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
148
74
10/5/2012
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
149
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
150
75
10/5/2012
Ex.: Determine the convolution y(n) of the

signals x(n) = 1, 3 n 6
0, elsewhere.
h(n) = 1, -4 n -3
0, elsewhere.
Ans: y(n) = {1, 2, 2, 2, 1, 0, 0}
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
151
DSP 4th ed. by P. Ramesh Babu (To be done in CW)
Ex. 1.4
Ex. 1.5
Ex. 1.8
Ex. 1.9
Ex. 1.11
Ex. 1.12
Ex.1.14
Ex. 1.15
Ex. 1.16
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
152
76
10/5/2012
Recursive & non-recursive systems

Recursive System:
A discrete time system in which output y(n)
depends on inputs as well as previous outputs
is called a recursive system.
y(n) = x(n) + ay (n -1)

10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
153
Non-recursive system :
It do not require any past output sample to
calculate the present output.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
154
77
10/5/2012
Linear Constant Coefficient Difference Equation

LTI systems are characterised
by input output relations called as difference
equation with constant co-efficients.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
155
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
156
78
10/5/2012
It consists of two parts as follows:

(1) The first term y(-l) is called initial condition of
the system.
(2) The second term is the response of system to the
input x(n).
Now depending upon initial conditions, two types
of responses are obtained.
1. Forced response : Initial condition y(-1) of
system is zero.
2. Zero input response or natural response: Input is
not applied to the system and initial condition.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
157
Total response can be expressed as the sum of the

natural response and forced response of the system.
The natural response is not the same as the zero input
response of the system.
The natural response is defined as the component of the
total response, which consists of all terms displaying
the natural frequencies of the system.
Natural frequencies are also known as the
characteristic roots of the system, eigenvalues of the
system determinant, and poles of the transfer function.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
158
79
10/5/2012
Solution of Linear Constant Coefficient Equation
Aim is to determine the output y(n), n 0, of the

system for the given specific input x(n), n 0 &
set of initial conditions.
Obtain two sets of solutions to the difference
equation, a homogeneous solution & a particular
solution.
Homogeneous Solution:
Obtained by setting terms involving x(n) = 0 &
finding outputs that are possible with zero inputs.
The initial conditions are used to determine the
arbitrary coefficients in the H.S.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
159
The particular solution is obtained by guessing

a sequence, with respect to x(n).
x(n) Input signal
A (step input)
AMn
AnM
Acos0n/Asin0n
yp(n) Particular solution

K
KMn
K0nM+K1nM-1++KM
K1cos0n+k2sin0n
Total solution can be written as

y(n) = yh(n)+yp(n)
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
160
80
10/5/2012
Pr.: Determine the homogeneous solution of the system described by the first
order difference equation y(n) +3y(n-1) = x(n), with initial condition y(-1) = 1.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
161
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
162
81
10/5/2012
Pr.: Find the response of the system described by the

difference equation,
y(n) + 2y(n-1) + y(n-2) = x(n) +x(n-1)
1
for the input x(n) = 2 u(n) with initial conditions, y(-1) =
y(-2) = 1.
Solution:
y(n) = yh(n)+yp(n)
Obtain yh(n) with x(n) = 0, Let yh(n) = n
n + 2 n -1+ n-2 = 0 -------- Char. Eqn.
= -1, -1
H. S: yh(n) = A1(-1)n + A2n(-1)n
n
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
163
To obtain Particular solution: [y(n) + 2y(n-1) + y(n-2) = x(n) +x(n-1)]

n
1
y p (n) = k u(n)
2
n
1
1
k + 2k
2
2
n 1
1
+ k
2
n2
n 1
1 1
= + u(n 1)
2 2
Sub., n 2, such that none of the term gets vanishes,

2
1 1
1
1
1
k + 2k + k = + k =
4 2
3
2
2
n
1 1
y p ( n ) = u ( n )
3 2
n
1 1
y(n) = A1 (1) + A2 n(1) + u(n)
3 2
n
(1)
Using (1), for n = 0 & n = 1, we get,

10/5/2012
y(0) = A1 +
S.THAI SUBHA
1
3
DSP, CHAPTER-I
y(1) = A1 A2 +
1
6
164
82
10/5/2012
Using the given D.E,

y(n) + 2y(n-1) + y(n-2) = x(n) +x(n-1)
9
When n = 0, y(0) = -2 & when n = 1, y(1) = 2
A1 =
7
3
A 2 = 2
The response is given by,

n
7
1 1
y(n) = (1) n 2n(1) n + u(n)
3
3 2
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
165
Ex.: Find the forced response of the system

described by the difference equation
y(n) + 2y(n-1) + y(n-2) = x(n) +x(n-1)
for input, x ( n ) = 1 u ( n )
n
[Hint: Assume initial conditions as zero]

* Forced response is the solution of difference
equation for the given input with zero initial
conditions.
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
166
83
10/5/2012
1 1
y(n) = A1 (1) + A2 n(1) + u(n)
3 2
n
Using (1), we get,
(1)
y(0) = A1 +
1
3
y(1) = A1 A2 +
Using the given D.E,

y(n) + 2y(n-1) + y(n-2) = x(n) +x(n-1)
1
6
When n = 0, y(0) = 1 & when n = 1, y(1) = 2

2
A1 =
A2 = 0
3
The forced response is given by,
n
2
1 1
y(n) = (1) n + u(n)
3
3 2
10/5/2012
S.THAI SUBHA
DSP, CHAPTER-I
167
84

Chapter I

Uploaded by

Copyright:

Available Formats

Chapter I

Uploaded by

Document Information

Original Description:

Original Title

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

Chapter I

Uploaded by

Copyright:

Available Formats

10/5/2012

DISCRETE TIME SIGNAL AND SYSTEMS

Digital Signal Processing the processing

Digital signal processing (DSP) is the mathematical manipulation

Digital image processing,

x(t) = Ai(t)sin[2fi(t)t + i(t)];

*Dimensions- Based on the number of independent variables.

Based on channels used:

One dimensional signals

Two dimensional signal

Based on the fundamental properties;

Continuous-time signal: A signal x(t) is said to

Discrete-time signal: A signal x(n) is said to be

How to convert a continuous-time

Continuous time signal

Steps for converting

Discrete-time signal with a sampling period T = 0.1s

P. No.2: The continuous time signal x(t) = e-2t for

Continuous time signal

Steps for converting

The discrete-time signal can be obtained as below;

Discrete-time signal with a sampling period T = 0.1s

Ex.1: Sketch the signal x(t) = sin7t + sin10t for an

Continuous-Time sinusoidal signals:

Discrete- Time Sinusoidal signals:

2. Discrete-time sinusoids whose frequencies

Relationship of frequency variable

Digital signal processing system

Continuous-time sinusoid is sampled at regular

Periodic sampling of an analog signal

Sampling and quantization of a sinusoidal signal

Consider two sinusoids,

x2(t) yields exactly the same values as x1(t) at Fs = 40Hz.

Nyquist rate relative to sampling

To avoid aliasing, the sampling rate must exceed

Consider the analog signal, xa(t) = 3 cos100t.

(c). The discrete time signal after sampling at Fs =

(d). For the sampling rate of Fs = 75Hz, we have

(ii). Tabular representation

Unit step sequence

Unit sample sequence

Unit step sequence

Unit ramp sequence:

Classification of discrete-time signals

If E is finite, [0 E < ] then x(n) is called an

The average power of periodic sequence with

If P is finite, then the signal is known as power

x(n) is energy signal.

x(n) is an energy signal. P = 0.

A continuous-time signal is said to be periodic if

Problem: Test whether the given signals are

(ii). x(t) = tesin(t)

Substitute t = t+T, (T = 2),

Pr.: Determine whether or not each of the

(iii). x(n) = cos 23 n

(ii). x(n) is not periodic

x(n) = cos 3 n+cos3 n

For an Even signal x(n) = x(-n)