EC3354 Signals and Systems Lecture Notes 2

`
EnggTree.com
Unit 1: Classification of signals and systems

Signal
Signal is one that carries information and is defined as a physical quantity that varies
with one or more independent variable.
Example: Music, speech
Classification of signals
Analog and Digital signal
Analog signal:
A signal that is defined for every instants of time is known as analog signal. Analog
signals are continuous in amplitude and continuous in time. It is denoted by x(t). It is also called
as Continuous time signal. Example for Continuous time signal is shown in Fig 1.1
Amplitude Sine Wave
time 0 1 0 1 0 0 1 0
Fig 1.2 Digital Signal

Fig 1.1 Continuous time signal
www.EnggTree.com
Digital signal:
The signals that are discrete in time and quantized in amplitude is called digital signal
(Fig 1.2)
Continuous time and discrete time signal
Continuous time signal:

A signal that is defined for every instants of time is known as continuous time signal.
Continuous time signals are continuous in amplitude and continuous in time. It is denoted by
x(t) and shown in Fig 1.1
Discrete time signal:

A signal that is defined for discrete instants of time is known as discrete time signal.
Discrete time signals are continuous in amplitude and discrete in time. It is also obtained by
sampling a continuous time signal. It is denoted by x(n) and shown in Fig 1.3
1
Downloaded from EnggTree.com
EnggTree.com
Fig 1.3 Discrete time signal
Even (symmetric) and Odd (Anti-symmetric) signal

Continuous domain:
Even signal:
www.EnggTree.com
A signal that exhibits symmetry with respect to t=0 is called even signal
Even signal satisfies the condition 𝑥(𝑡) = 𝑥(−𝑡)
Odd signal:
A signal that exhibits anti-symmetry with respect to t=0 is called odd signal
Odd signal satisfies the condition 𝑥(𝑡) = −𝑥(−𝑡)
Even part 𝒙𝒆(𝒕) and Odd part 𝒙𝟎(𝒕) of continuous time signal 𝒙 𝒕 :
Even part 𝑥𝑒 𝑡 = 1 [𝑥 2𝑡 + 𝑥 −𝑡 ]
Odd part 𝑥𝑜 𝑡 = 1 [𝑥 2𝑡 − 𝑥 −𝑡 ]
Discrete domain:
Even signal:
A signal that exhibits symmetry with respect to n=0 is called even signal
Even signal satisfies the condition 𝑥(𝑛) = 𝑥(−𝑛).
Odd signal:
A signal that exhibits anti-symmetry with respect to n=0 is called odd signal Odd signal
satisfies the condition 𝑥(𝑛) = −𝑥(−𝑛).
Even part 𝒙𝒆(𝒏) and Odd part 𝒙𝟎(𝒏) of discrete time signal 𝒙 𝒏 :
Even part 𝑥𝑒 𝑛 = 1 [𝑥 2𝑛 + 𝑥 −𝑛 ]
Odd part 𝑥𝑜 𝑛 = 1 [𝑥 2𝑛 − 𝑥 −𝑛 ]
2
EnggTree.com
Periodic and Aperiodic signal

Periodic signal:
A signal is said to periodic if it repeats again and again over a certain period of time.
Aperiodic signal:
A signal that does not repeat at a definite interval of time is called aperiodic signal.
Continuous domain:
A Continuous time signal is said to periodic if it satisfies the condition
𝑥 𝑡 = 𝑥 𝑡 + 𝑇 𝑤𝑕𝑒𝑟𝑒 𝑇 𝑖𝑠 𝑓𝑢𝑛𝑑𝑎𝑚𝑒𝑛𝑡𝑎𝑙 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑
If the above condition is not satisfied then the signal is said to be aperiodic
Fundamental time period 𝐓 = 𝟐𝛑, where Ω is fundamental angular frequency in rad/sec
Ω
Discrete domain:
A Discrete time signal is said to periodic if it satisfies the condition
𝑥 𝑛 = 𝑥 𝑛 + 𝑁 𝑤𝑕𝑒𝑟𝑒 𝑁 𝑖𝑠 𝑓𝑢𝑛𝑑𝑎𝑚𝑒𝑛𝑡𝑎𝑙 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑
If the above condition is not satisfied then the signal is said to be aperiodic
Fundamental time period 𝐍 = 𝟐𝛑𝐦, where ω is fundamental angular frequency in rad/sec, 𝑚 is
𝛚
www.EnggTree.com
smallest positive integer that makes N as positive integer
Energy and Power signal
Energy signal:
The signal which has finite energy and zero average power is called energy signal. The
non periodic signals like exponential signals will have constant energy and so non periodic
signals are energy signals.
i.e., For energy signal, 0 < 𝐸 < ∞ 𝑎𝑛𝑑 𝑃 = 0
For Continuous time signals,
𝑇
𝐸𝑛𝑒𝑟𝑔𝑦 𝐸 = lim |𝑥 𝑡 |2𝑑𝑡
𝑇→∞ −𝑇
For Discrete time signals,

𝑁
𝐸𝑛𝑒𝑟𝑔𝑦 𝐸 = lim 𝑥(𝑛) 2
𝑁→∞
𝑛=−𝑁
Power signal:
The signal which has finite average power and infinite energy is called power signal. The
periodic signals like sinusoidal complex exponential signals will have constant power and so
periodic signals are power signals.
i.e., For power signal, 0 < 𝑃 < ∞ 𝑎𝑛𝑑 𝐸 = ∞
For Continuous time signals,
3
EnggTree.com
𝑇 2
1
𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑝𝑜𝑤𝑒𝑟 𝑃 = lim |𝑥 𝑡 | 𝑑𝑡
𝑇→∞ 2𝑇 −𝑇
For Discrete time signals,
1 𝑁
𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑝𝑜𝑤𝑒𝑟 𝑃 = lim 𝑥(𝑛) 2
𝑁→∞ 2𝑁 + 1
𝑛=−𝑁
Deterministic and Random signals
Deterministic signal:
A signal is said to be deterministic if there is no uncertainity over the signal at any
instant of time i.e., its instantaneous value can be predicted. It can be represented by
mathematical equation.
Example: sinusoidal signal
Random signal (Non-Deterministic signal):

A signal is said to be random if there is uncertainity over the signal at any instant of time
i.e., its instantaneous value cannot be predicted. It cannot be represented by mathematical
equation.
Example: noise signal
Sinusoidal Signal
www.EnggTree.com Random Signal
Random signal
Deterministic signal
Causal and Non-causal signal
Continuous domain:
Causal signal:
A signal is said to be causal if it is defined for t≥0.
𝑖. 𝑒., 𝑥 𝑡 = 0 𝑓𝑜𝑟 𝑡 < 0
Non-causal signal:
A signal is said to be non-causal, if it is defined for t< 0 or for both
𝑡 < 0 and 𝑡 ≥ 0
𝑖. 𝑒., 𝑥 𝑡 ≠ 0 𝑓𝑜𝑟 𝑡 < 0
When a non-causal signal is defined only for t<0, it is called as anti-causal signal
4
EnggTree.com
Discrete domain:
Causal signal:
A signal is said to be causal, if it is defined for n≥0.
𝑖. 𝑒., 𝑥 𝑛 = 0 𝑓𝑜𝑟 𝑛 < 0
Non-causal signal:
A signal is said to be non-causal, if it is defined for n< 0
or for both n < 0 𝑎𝑛𝑑 𝑛 ≥ 0
𝑖. 𝑒., 𝑥 𝑛 ≠ 0 𝑓𝑜𝑟 𝑛 < 0
When a non-causal signal is defined only for n<0, it is called as anti-causal signal
Basic(Elementary or Standard) continuous time signals

Step signal
Unit Step signal is defined as
u (t)
𝑢 𝑡 = 1 𝑓𝑜𝑟 𝑡 ≥ 0 1
= 0 𝑓𝑜𝑟 𝑡 < 0
www.EnggTree.com0 t
Unit step signal
Ramp signal
Unit ramp signal is defined as

.
.
r (t)
𝑟 𝑡 = 𝑡 𝑓𝑜𝑟 𝑡 ≥ 0
3
= 0 𝑓𝑜𝑟 𝑡 < 0
2
0 1 2 3. . t
Unit ramp signal
Parabolic signal
Unit Parabolic signal is defined as
5
EnggTree.com
p(t)
:
4.5
𝑡2
𝑥 𝑡 = 𝑓𝑜𝑟 𝑡 ≥ 0 2
2
= 0 𝑓𝑜𝑟 𝑡 < 0 0.5
0 1 2 3...t
Unit Parabolic signal
Relation between Unit Step signal, Unit ramp signal and Unit Parabolic signal:
• Unit ramp signal is obtained by integrating unit step signal

𝑖. 𝑒. , 𝑢 𝑡 𝑑𝑡 = 1𝑑𝑡 = 𝑡 = 𝑟(𝑡)
• Unti Parabolic signal is obtained by integrating unit ramp signal
𝑡2
𝑖. 𝑒. , 𝑟 𝑡 𝑑𝑡 = 𝑡𝑑𝑡 = = 𝑝(𝑡)
2
• Unit step signal is obtained by differentiating unit ramp signal

𝑑 𝑑
𝑖. 𝑒. , 𝑟 𝑡 = 𝑡 = 1 = 𝑢(𝑡)
𝑑𝑡 𝑑𝑡
• Unit ramp signal is obtained by differentiating unit Parabolic signal
www.EnggTree.com
𝑑
𝑖. 𝑒. , 𝑝 𝑡 =
𝑑 𝑡2 1
2
𝑡 = 𝑡 = 𝑟(𝑡)
𝑑𝑡 2 =
𝑑𝑡 2
Unit Pulse signal is defined as

∏ (t)
1
∏ 𝑡 = 1 𝑓𝑜𝑟 𝑡 ≤ 1
2
= 0 𝑒𝑙𝑠𝑒𝑤𝑕𝑒𝑟𝑒
-1/2 1/2 t
Unit Pulse signal
Impulse signal
Unit Impulse signal is defined as
𝛿 𝑡 = 0 𝑓𝑜𝑟 𝑡 ≠ 0
∞
𝛿 𝑡 𝑑𝑡 = 1
−∞
Unit Impulse signal
6
EnggTree.com
Properties of Impulse
signal: Property 1:
∞
𝒙(𝒕)𝜹 𝒕 𝒅𝒕 = 𝒙(𝟎)
−∞
Proof:
∞
𝑥(𝑡)𝛿 𝑡 𝑑𝑡 = 𝑥 0 𝛿 0 = 𝑥 0 [∵ 𝛿 𝑡 𝑒𝑥𝑖𝑠𝑡𝑠 𝑜𝑛𝑙𝑦 𝑎𝑡 𝑡 = 0 𝑎𝑛𝑑 𝛿 0 = 1]

−∞
Thus proved
Property 2:
∞
𝒙(𝒕)𝜹 𝒕 − 𝒕𝟎 𝒅𝒕 = 𝒙(𝒕𝟎)
−∞
Proof:
∞
www.EnggTree.com
𝑥(𝑡)𝛿 𝑡 − 𝑡0 𝑑𝑡 = 𝑥 𝑡0 𝛿 𝑡0 − 𝑡0 = 𝑥 𝑡0 𝛿 0 = 𝑥 𝑡0
−∞
∵ 𝛿 𝑡 − 𝑡0 𝑒𝑥𝑖𝑠𝑡𝑠 𝑜𝑛𝑙𝑦 𝑎𝑡 𝑡 = 𝑡0 𝑎𝑛𝑑 𝛿 0 = 1
Thus proved
Sinusoidal signal
Cosinusoidal signal is defined as Sinusoidal signal is defined as
𝑥 𝑡 = 𝐴𝑐𝑜𝑠 𝛺𝑡 + 𝛷 𝑥 𝑡 = 𝐴𝑠𝑖𝑛 𝛺𝑡 + 𝛷
where Ω = 2πf = 2π and Ω is angular frequency in rad/sec

T
f is frequency in cycles/sec or Hertz and
A is amplitude
T is time period in seconds
𝛷 is phase angle in radians
Cosinusoidal signal Sinusoidal signal

𝑤𝑕𝑒𝑛 𝜙 = 0, 𝑥 𝑡 = 𝐴𝑐𝑜𝑠 𝛺𝑡 𝑤𝑕𝑒𝑛 𝜙 = 0, 𝑥 𝑡 = 𝐴𝑠𝑖𝑛 𝛺𝑡
x(t) φ=0
x(t) φ=0
A
A
t
t
-A
-A
7
EnggTree.com
Sinusoidal signal
Cosinusoidal signal
Exponential signal
Real Exponential signal is defined as 𝑥 𝑡 = 𝐴𝑒𝑎𝑡
where A is amplitude
Depending on the value of ‘a’ we get dc signal or growing exponential signal or decaying
exponential signal
x(t) a=0 x(t) a>0 x(t) a<0
A t A t A t
DC signal Exponentially Exponentially

growing signal decaying signal
www.EnggTree.com
Complex exponential signal is defined as 𝑥 𝑡 = 𝐴𝑒𝑠𝑡
where 𝐴 is amplitude, s is complex variable and 𝑠 = 𝜎 + 𝑗Ω
𝑥 𝑡 = 𝐴𝑒𝑠𝑡 = 𝐴𝑒 𝜎+𝑗Ω 𝑡 = 𝐴𝑒𝜎𝑡 𝑒𝑗Ω𝑡 = 𝐴𝑒𝜎𝑡 (𝑐𝑜𝑠Ω𝑡 + 𝑗𝑠𝑖𝑛Ω𝑡)
𝑤𝑕𝑒𝑛 𝜎 = +𝑣𝑒, 𝑡𝑕𝑒𝑛 𝑥 𝑡 = 𝐴𝑒 𝜎𝑡 (𝑐𝑜𝑠Ω𝑡 + 𝑗𝑠𝑖𝑛Ω𝑡),
𝑤𝑕𝑒𝑟𝑒 𝑥𝑟 𝑡 = 𝐴𝑒 𝜎𝑡 𝑐𝑜𝑠Ω𝑡 𝑎𝑛𝑑 𝑥𝑖 𝑡 = 𝐴𝑒 𝜎𝑡 𝑠𝑖𝑛Ω𝑡
xr(t) xi(t)
A A
t t
Exponentially growing sinusoidal signal

Exponentially growing Cosinusoidal signal
𝑤𝑕𝑒𝑛 𝜎 = −𝑣𝑒, 𝑡𝑕𝑒𝑛 𝑥 𝑡 = 𝐴𝑒 −𝜎𝑡 (𝑐𝑜𝑠Ω𝑡 + 𝑗𝑠𝑖𝑛Ω𝑡),

𝑤𝑕𝑒𝑟𝑒 𝑥𝑟 𝑡 = 𝐴𝑒 −𝜎𝑡 𝑐𝑜𝑠Ω𝑡 𝑎𝑛𝑑 𝑥𝑖 𝑡 = 𝐴𝑒 −𝜎𝑡 𝑠𝑖𝑛Ω𝑡
xr(t) xi(t)
A A
t t
8
EnggTree.com
Exponentially decaying Cosinusoidal signal Exponentially decaying sinusoidal signal
Basic(Elementary or Standard) Discrete time signals
Step signal
Unit Step signal is defined as
u(n)
𝑢 𝑛 = 1 𝑓𝑜𝑟 𝑛 ≥ 0
= 0 𝑓𝑜𝑟 𝑛 < 0
1
. . .
0 1 2 3 4 n
www.EnggTree.comUnit step signal

Unit Ramp signal
Unit Ramp signal is defined as
r(n)
:
4
𝑟 𝑛 = 𝑛 𝑓𝑜𝑟 𝑛 ≥ 0 3
= 0 𝑓𝑜𝑟 𝑛 < 0 2 ...
1
0 1 2 3 4 n
Unit Ramp signal
Pulse signal (Rectangular pulse function)

Pulse signal is defined as
x(n)
𝑥 𝑛 = 𝐴 𝑓𝑜𝑟 𝑛1 ≤ 𝑛 ≤ 𝑛2 A
... ...
= 0 𝑒𝑙𝑠𝑒𝑤𝑕𝑒𝑟𝑒
n1 -1 0 1 n2 n
Pulse signal
Unit Impulse signal

Unit Impulse signal is defined as
9
EnggTree.com
𝛿 𝑛 = 1 𝑓𝑜𝑟 𝑛 = 0
𝛿 𝑛 = 0 𝑓𝑜𝑟 𝑛 ≠ 0
Unit Impulse signal
Sinusoidal signal
Cosinusoidal signal is defined as Sinusoidal signal is defined as
𝑥 𝑛 = 𝐴𝑐𝑜𝑠(𝜔𝑛) 𝑥 𝑛 = 𝐴𝑠𝑖𝑛(𝜔𝑛)
where 𝜔 = 2πf = 2π m and 𝜔 is frequency in radians/sample

N
m is smallest integer
f is frequency in cycles/sample, A is amplitude
Cosinusoidal signal Sinusoidal signal
www.EnggTree.com
Sinusoidal signal
Cosinusoidal signal
Exponential signal
Real Exponential signal is defined as 𝑥 𝑛 = 𝑎𝑛 𝑓𝑜𝑟 𝑛 ≥ 0
Decreasing exponential signal Increasing exponential signal
Complex Exponential signal is defined as 𝑥 𝑛 = 𝑎𝑛 𝑒𝑗 (𝜔 0𝑛) = 𝑎𝑛 [𝑐𝑜𝑠𝜔0𝑛 + 𝑗𝑠𝑖𝑛𝜔0𝑛]

𝑤𝑕𝑒𝑟𝑒 𝑥𝑟 𝑛 = 𝑎𝑛 𝑐𝑜𝑠𝜔0 𝑛 𝑎𝑛𝑑 𝑥𝑖 𝑛 = 𝑎𝑛 𝑠𝑖𝑛𝜔0 𝑛
10
EnggTree.com
Exponentially decreasing Cosinusoidal signal

Exponentially decreasing sinusoidal signal
www.EnggTree.com
Exponentially growing Cosinusoidal signal
Exponentially growing sinusoidal signal
Classification of System
• Continuous time and Discrete time system
• Linear and Non-Linear system
• Static and Dynamic system
• Time invariant and Time variant system
• Causal and Non-Causal system
• Stable and Unstable system
Continuous time and Discrete time system

Continuous time system:
Continuous time system operates on a continuous time signal (input or excitation) and
produces another continuous time signal (output or response) which is shown in Fig 1.84. The
signal 𝑥(𝑡) is transformed by the system into signal 𝑦(𝑡), this transformation can be expressed
as,
11
EnggTree.com
Response 𝑦 𝑡 = 𝑇 𝑥 𝑡
where 𝑥(𝑡) is input signal, 𝑦(𝑡) is output signal, and T denotes transformation
𝑥(𝑡) T 𝑦(𝑡)
Fig 1.84 Representation of continuous time system

Discrete time system:
Discrete time system operates on a discrete time signal (input or excitation) and
produces another discrete time signal (output or response) which is shown in Fig 1.85.
The signal 𝑥(𝑛) is transformed by the system into signal 𝑦(𝑛), this transformation can be
expressed as,
Response 𝑦 𝑛 = 𝑇 𝑥 𝑛
where x(n) is input signal, y(n) is output signal, and T denotes transformation
𝑥(𝑛) T 𝑦(𝑛)
Fig 1.85 Representation of discrete time system
Linear system and Non Linear system

Continuous time domain:
Linear system: www.EnggTree.com
A system is said to be linear if it obeys superposition theorem. Superposition theorem
states that the response of a system to a weighted sum of the signals is equal to the
corresponding weighted sum of responses to each of the individual input signals.
Condition for Linearity:
𝑇 𝑎𝑥1 𝑡 + 𝑏𝑥2 𝑡 = 𝑎𝑦1 𝑡 + 𝑏𝑦2(𝑡)
𝑤𝑕𝑒𝑟𝑒 𝑦1 𝑡 𝑎𝑛𝑑 𝑦2 𝑡 𝑎𝑟𝑒 𝑡𝑕𝑒 𝑟𝑒𝑠𝑝𝑜𝑛𝑠𝑒𝑠 𝑜𝑓 𝑥1 𝑡 𝑎𝑛𝑑 𝑥2 𝑡
𝑟𝑒𝑠𝑝𝑒𝑐𝑡𝑖𝑣𝑒𝑙𝑦 Non Linear system:
A system is said to be Non linear if it does not obeys superposition theorem.
𝑖. 𝑒. , 𝑇 𝑎𝑥1 𝑡 + 𝑏𝑥2 𝑡 ≠ 𝑎𝑦1 𝑡 + 𝑏𝑦2(𝑡)
where y1 t and y2 t are the responses of x1 t and x2 t respectively
Discrete time domain:

Linear system:
A system is said to be linear if it obeys superposition theorem. Superposition theorem
states that the response of a system to a weighted sum of the signals is equal to the
corresponding weighted sum of responses to each of the individual input signals.
Condition for Linearity:
𝑇 𝑎𝑥1 𝑛 + 𝑏𝑥2 𝑛 = 𝑎𝑦1 𝑛 + 𝑏𝑦2(𝑛)
where y1 n and y2 n are the responses of x1 n and x 2 n respectively
Non Linear system:
A system is said to be Non linear if it does not obeys superposition theorem.
𝑖. 𝑒. , 𝑇 𝑎𝑥1 𝑛 + 𝑏𝑥2 𝑛 ≠ 𝑎𝑦1 𝑛 + 𝑏𝑦2(𝑛)
12
EnggTree.com
where y1 n and y2 n are the responses of x1 n and x2 n respectively
Static (Memoryless) and Dynamic (Memory) system

Static system:
A system is said to be memoryless or static if the response of the system is due to
present input alone.
Example: 𝑦(𝑡) = 2𝑥(𝑡)
𝑦(𝑡) = 𝑥2 (𝑡) + 𝑥(𝑡)
Dynamic system:
A system is said to be memory or dynamic if the response of the system depends on
factors other than present input also.
Example: 𝑦(𝑡) = 2𝑥(𝑡) + 𝑥(−𝑡)
𝑦(𝑡) = 𝑥2 (𝑡) + 𝑥(2𝑡)

Static system:
A system is said to be memoryless or static if the response of the system is due to
present input alone.
Example: 𝑦(𝑛) = 𝑥(𝑛) www.EnggTree.com
𝑦(𝑛) = 𝑥2 (𝑛) + 3𝑥(𝑛)
Dynamic system:
A system is said to be memory or dynamic if the response of the system depends on
factors other than present input also.
Example: 𝑦(𝑛) = 2𝑥(𝑛) + 𝑥(−𝑛)
𝑦(𝑛) = 𝑥2 (1 − 𝑛) + 𝑥(2𝑛)
Time invariant (Shift invariant) and Time variant (Shift variant) system
Time invariant system:
A system is said to time invariant if the relationship between the input and output does
not change with time.
If 𝑦 𝑡 = 𝑇 𝑥 𝑡
Then 𝑇 𝑥 𝑡 − 𝑡0 = 𝑦(𝑡 − 𝑡0) should be satisfied for the system to be time invariant
Time variant system:

A system is said to time variant if the relationship between the input and output changes
with time.
If 𝑦 𝑡 = 𝑇 𝑥 𝑡
Then 𝑇 𝑥 𝑡 − 𝑡0 ≠ 𝑦(𝑡 − 𝑡0) should be satisfied for the system to be time variant
13
EnggTree.com
Time invariant system:

A system is said to time invariant if the relationship between the input and output does
not change with time.
If 𝑦 𝑛 = 𝑇 𝑥 𝑛
Then 𝑇 𝑥 𝑛 − 𝑛0 = 𝑦(𝑛 − 𝑛0) should be satisfied for the system to be time invariant
Time variant system:
A system is said to time variant if the relationship between the input and output changes
with time.
If 𝑦 𝑛 = 𝑇 𝑥 𝑛
Then 𝑇 𝑥 𝑛 − 𝑛0 ≠ 𝑦(𝑛 − 𝑛0) should be satisfied for the system to be time variant
Causal and Non-Causal system

Causal system:
A system is said to be causal if the response of a system at any instant of time depends
only on the present input, past input and past output but does not depends upon the future
input and future output.
Example: 𝑦(𝑡) = 3𝑥(𝑡) + 𝑥(𝑡 − 1)
www.EnggTree.com
A system is said to be causal if impulse response 𝑕(𝑡) is zero for negative values
of 𝑡 i.e., 𝑕(𝑡) = 0 𝑓𝑜𝑟 𝑡 < 0
Non-Causal system:
A system is said to be Non-causal if the response of a system at any instant of time
depends on the future input and also on the present input, past input, past output.
Example: 𝑦(𝑡) = 𝑥(𝑡 + 2) + 𝑥(𝑡 − 1)
𝑦(𝑡) = 𝑥(−𝑡) + 𝑥(𝑡 + 4)
A system is said to be non-causal if impulse response 𝑕(𝑡) is non-zero for negative values
of 𝑡 i.e., 𝑕(𝑡) ≠ 0 𝑓𝑜𝑟 𝑡 < 0

Causal system:
A system is said to be causal if the response of a system at any instant of time depends
only on the present input, past input and past output but does not depends upon the future
input.
Example: 𝑦(𝑛) = 3𝑥(𝑛) + 𝑥(𝑛 − 1)
A system is said to be causal if impulse response h(n) is zero for negative values of n
i.e., 𝑕(𝑛) = 0 𝑓𝑜𝑟 𝑛 < 0
Non-Causal system:
A system is said to be Non-causal if the response of a system at any instant of time
depends on the future input and also on the present input, past input, past output.
Example: 𝑦(𝑛) = 𝑥(𝑛 + 2) + 𝑥(𝑛 − 1)
14
EnggTree.com
𝑦(𝑛) = 𝑥(−𝑛) + 𝑥(𝑛 + 4)

A system is said to be non-causal if impulse response 𝑕(𝑛) is non-zero for negative values
of n i.e., 𝑕(𝑛) ≠ 0 for 𝑛 < 0
Stable and Unstable system

A system is said to be stable if and only if it satisfies the BIBO stability criterion.
BIBO stable condition:
• Every bounded input yields bounded output.
𝑖. 𝑒. , if 0 < 𝑥 t < ∞ 𝑡𝑕𝑒𝑛 0 < 𝑦 t < ∞ should be satisfied for the system to be stable
• Impulse response should be absolutely integrable
∞
𝑖. 𝑒. , 0 < 𝑕(𝜏) 𝑑𝜏 < ∞

−∞
If the BIBO stable condition is not satisfied, then the system is said to be unstable system
A system is said to be stable if and only if it satisfies the BIBO stability criterion.
BIBO stable condition:
• Every bounded input yields bounded output.
• Impulse response should be absolutely summable
∞
𝑖. 𝑒. , 0 < 𝑕(𝑘) < ∞

www.EnggTree.com 𝑘=−∞
If the BIBO stable condition is not satisfied, then the system is said to be unstable system
Solved Problems
1. Draw 𝑟(𝑡 + 3), where r(t) is ramp signal

Solution:
𝑟 𝑡 = 𝑡; 𝑡 ≥ 0
r(t) r(t+3)
: :
: 4
3
3
2
2
1
1
0 1 2 3 ... t
-3 -2 -1 0 1...t
15
EnggTree.com
2. Sketch 𝑥(𝑡) = 3𝑟(𝑡 − 1) + 𝑟(−𝑡 + 2)
www.EnggTree.com
𝑥 𝑡 = 3𝑟 𝑡 − 1 + 𝑟(−𝑡 + 2)
= 0 + 4 − 𝑡 𝑓𝑜𝑟 − 2 ≤ 𝑡 ≤ −1
= 0 + 3 − 𝑡 𝑓𝑜𝑟 − 1 ≤ 𝑡 ≤ 0
= 0 + 2 − 𝑡 𝑓𝑜𝑟 0 ≤ 𝑡 ≤ 1
= 3𝑡 + 1 − 𝑡 𝑓𝑜𝑟 1 ≤ 𝑡 ≤ 2
= 3 + 3𝑡 + 0 𝑓𝑜𝑟 2 ≤ 𝑡 ≤ 3
= 6 + 3𝑡 + 0 𝑓𝑜𝑟 3 ≤ 𝑡 ≤ 4
and so on
Fig 1.163
3. Draw time reversal signal of unit step signal

Solution:
𝑢 𝑛 = 1; 𝑛 ≥ 0
16
EnggTree.com
4. Check whether the following is periodic or not. If periodic, determine fundamental time
period
a. 𝒙 𝒕 = 𝟐 𝐜𝐨𝐬 𝟓𝒕 + 𝟏 − 𝐬𝐢𝐧 𝟒𝒕
Here Ω1 = 5, Ω2 = 4
2π 2π 2π
𝑇1 = Ω1 = 5 = 5
2π 2π π
𝑇2 = = =
Ω2 4 2
𝑇1 2π5
= 4
𝑇2 π = 5 (It is rational number)
2
Hence 𝑥(𝑡) is periodic
www.EnggTree.com
𝑇 = 5𝑇 = 4𝑇 = 2𝜋
1 2
∴ 𝑥(𝑡) is periodic with period 𝟐𝝅
b. 𝑥 𝑛 = 3 cos 4𝜋𝑛 + 2 sin 𝜋𝑛

Here ω1 = 4π, ω2 = π
𝑁 2πm 2πm m
1 = = =
ω1 4π 2
𝑁1 = 1 (taking m = 2)
2πm 2πm
𝑁2 = ω2 = π = 2m
𝑁2 = 2 (taking m = 1)
𝑁 = 𝐿𝐶𝑀 1,2 = 2
Hence x(n) ∴ 𝑥(𝑛) is periodic with period 2
5. Determine whether the signals are energy or power signal

𝒙 𝒕 = 𝒆−𝟑𝒕𝒖(𝒕)
𝑇 𝑇 𝑇
𝑇
𝑬𝒏𝒆𝒓𝒈𝒚 𝑬∞ = lim |𝑥 𝑡 |2𝑑𝑡 = lim |𝑒−3𝑡 |2𝑑𝑡 = lim 𝑒−6𝑡 𝑑𝑡 = lim 𝑒−6𝑡
𝑇→∞ 𝑇→∞ 𝑇→∞ 𝑇→∞ −6 0
−𝑇 0 0
−6𝑇 −0 𝟏
= lim 𝑒 − 𝑒 = <∞ ∵ 𝑒−∞ = 0, 𝑒−0 = 1
𝑇→∞ −6 −6 𝟔
17
EnggTree.com
1 𝑇 1 𝑇 1 𝑇
𝑷𝒐𝒘𝒆𝒓 𝑷 = lim 𝑥 𝑡 2𝑑𝑡 = lim 𝑒−3𝑡 2𝑑𝑡 = lim 𝑒−6𝑡 𝑑𝑡
∞
𝑇→∞ 2𝑇 𝑇→∞ 2𝑇 𝑇→∞ 2𝑇
−𝑇 0 0
1 𝑇 −0 1 1 1
= lim 𝑒−6𝑡 = lim 1 𝑒
−6𝑇
−𝑒 = lim =𝟎 =0 ∵ 𝑒 −∞ = 0, 𝑒 −0 = 1,
𝑇→∞ 2𝑇 −6 0 𝑇→∞ 2𝑇 −6 −6 𝑇→∞ 2𝑇 6 ∞
Since energy value is finite and average power is zero, the given signal is an energy signal.
6. Determine whether the signals are energy or power signal

𝝅𝒏 𝝅
𝒋 𝟒 +𝟐
𝒙𝒏=𝒆
𝑁 𝑁
𝑁 𝝅𝒏 𝝅 2
𝑬𝒏𝒆𝒓𝒈𝒚 𝑬∞ = lim 𝑥(𝑛) 2 = lim 𝒆 𝒋( 𝟒 +𝟐) = lim 12 = lim 2𝑁 + 1 = ∞
𝑁→∞ 𝑁→∞ 𝑁→∞ 𝑁→∞
𝑛=−𝑁 𝑛=−𝑁 𝑛=−𝑁
𝑁
∵ 𝑒𝑗 (𝜔𝑛 +𝜃) = 1 𝑎𝑛𝑑 1 = 2𝑁 + 1

𝑛=−𝑁
𝑁 𝑁
1 1 𝝅𝒏 𝝅 2
𝑨𝒗𝒆𝒓𝒂𝒈𝒆 𝒑𝒐𝒘𝒆𝒓 𝑷 ∞ = lim 𝑥(𝑛) 2 = lim 𝒆 𝒋( 𝟒 +𝟐)
www.EnggTree.com
𝑁→∞ 2𝑁 +1
𝑛=−𝑁
𝑁
𝑁→∞ 2𝑁 + 1
𝑛=−𝑁
1 1
lim 12 = lim 2𝑁 + 1 = 𝟏
=𝑁→∞ 2𝑁 + 1 𝑁→∞ 2𝑁 + 1
𝑛=−𝑁
Since energy value is infinite and average power is finite, the given signal is power signal
7. Determine whether the following systems are linear or not
𝒅𝒚(𝒕)
+ 𝒕𝒚 𝒕 = 𝒙𝟐(𝒕)
𝒅𝒕
Output due to weighted sum of inputs:
𝑑[𝑎𝑦1 𝑡 + 𝑏𝑦2 𝑡 ]
+ 𝑡 𝑎𝑦 𝑡 + 𝑏𝑦 𝑡 = 𝑎𝑥 𝑡 + 𝑏𝑥 𝑡 2 … (1)
1 2 1 2
𝑑𝑡
Weighted sum of outputs:

For input 𝑥1 𝑡 :
𝑑𝑦1 𝑡
+ 𝑡𝑦 𝑡 = 𝑥 2 𝑡 … (2)
1 1
𝑑𝑡
For input 𝑥2 𝑡 :
𝑑𝑦2 𝑡
+ 𝑡𝑦 𝑡 = 𝑥 2 𝑡 … (3)
2 2
𝑑𝑦1 𝑡 𝑑𝑡𝑑𝑦 𝑡
2
2×𝑎+3×𝑏⇒𝑎 + 𝑎𝑡𝑦 𝑡 + 𝑏 + 𝑏𝑡𝑦 𝑡 = 𝑎𝑥 2 𝑡 + 𝑏𝑥 2 𝑡 … (4)
2 1 2
𝑑𝑡 1 𝑑𝑡
18
EnggTree.com
1 ≠ (4)
The given system is Non-Linear
8. Determine whether the following systems are linear or not

𝒚 𝒏 = 𝒙 𝒏 − 𝟐 + 𝒙(𝒏𝟐)
Output due to weighted sum of inputs:
𝑦3 𝑛 = 𝑎𝑥1 𝑛 − 2 + 𝑏𝑥2 𝑛 − 2 + 𝑎𝑥1 𝑛2 + 𝑏𝑥2(𝑛2)
Weighted sum of outputs:
For input 𝑥1 𝑛 :
𝑦1 𝑛 = 𝑥1 𝑛 − 2 + 𝑥1 𝑛 2
For input 𝑥2 𝑛 :
𝑦2 𝑛 = 𝑥2 𝑛 − 2 + 𝑥2(𝑛2 )
𝑎𝑦1 𝑛 + 𝑏𝑦2 𝑛 = 𝑎𝑥1 𝑛 − 2 + 𝑎𝑥1 𝑛2 + 𝑏𝑥2 𝑛 − 2 + 𝑏𝑥2(𝑛2 )
∵ 𝑦3 𝑛 = 𝑎𝑦1 𝑛 + 𝑏𝑦2 𝑛
9. Determine whether the following systems are static or dynamic

𝒚 𝒕 = 𝒙 𝟐𝒕 + 𝟐𝒙 𝒕
𝑦 0 = 𝑥 0 + 2𝑥 0 ⇒ present inputs
𝑦 −1 = 𝑥 −2 + 2𝑥 −1 ⇒ past and present inputs
𝑦 1 = 𝑥 2 + 2𝑥 1 ⇒ future and present inputs
Since output depends on past and future inputs the given system is dynamic system
www.EnggTree.com
10. Determine whether the following systems are static or dynamic
𝒚(𝒏) = 𝒔𝒊𝒏𝒙(𝒏)
𝑦 0 = 𝑠𝑖𝑛𝑥(0) ⇒ present input
𝑦 −1 = 𝑠𝑖𝑛𝑥(−1) ⇒ present input
Since output depends on present input the given system is Static system
11. Determine whether the following systems are time invariant or not
𝒚(𝒕) = 𝒙(𝒕)𝒔𝒊𝒏𝒘𝒕
Output due to input delayed by T seconds
𝑦(𝑡, 𝑇) = 𝑥(𝑡 − 𝑇)𝑠𝑖𝑛𝑤𝑡
Output delayed by T seconds
𝑦(𝑡 − 𝑇) = 𝑥(𝑡 − 𝑇)𝑠𝑖𝑛𝑤(𝑡 − 𝑇)
∵ 𝑦 𝑡, 𝑇 ≠ 𝑦 𝑡 − 𝑇
The given system is time variant
12. Determine whether the following systems are time invariant or not
𝒚 𝒏 = 𝒙(−𝒏 + 𝟐)
Output due to input delayed by k seconds
𝑦 𝑛, 𝑘 = 𝑥(−𝑛 + 2 − 𝑘)
Output delayed by k seconds
𝑦 𝑛 − 𝑘 = 𝑥 −(𝑛 − 𝑘 + 2) = 𝑥(−𝑛 + 𝑘 + 2)
∵ 𝒚 𝒏, 𝒌 ≠ 𝒚 𝒏 − 𝒌
The given system is time variant
19
EnggTree.com
13. Determine whether the following systems are causal or not

𝒅𝒙(𝒕)
𝒚𝒕= + 𝟐𝒙(𝒕)
𝒅𝒕
The given equation is differential equation and the output depends on past input. Hence the
given system is 𝐂𝐚𝐮𝐬𝐚𝐥
14. Determine whether the following systems are causal or not

𝒚(𝒏) = 𝒔𝒊𝒏𝒙(𝒏)
𝑦 −1 = 𝑠𝑖𝑛𝑥(−1) ⇒ present input
Since output depends on present input the given system is Causal system
15. Determine whether the following systems are stable or not
𝒉 𝒕 = 𝒆−𝟒𝒕𝒖(𝒕)
∞
www.EnggTree.com
Condition for stability 𝑕(𝜏) 𝑑𝜏 < ∞
−∞
∞ ∞ ∞ ∞
𝑒−4𝜏 1
𝑕(𝜏) 𝑑𝜏 = 𝑒 −4𝜏 𝑢(𝜏) 𝑑𝜏 = 𝑒 −4𝜏 𝑑𝜏 = =
−40 4
−∞ −∞ 0
∞
∵ 𝑕(𝜏) 𝑑𝜏 < ∞ the given system is 𝐬𝐭𝐚𝐛𝐥𝐞

−∞
16. Determine whether the following systems are stable or not

𝒚 𝒏 = 𝟑𝒙(𝒏)
Let 𝑥 𝑛 = 𝛿 𝑛 , 𝑦 𝑛 = 𝑕(𝑛)
⇒ 𝑕 𝑛 = 3𝛿(𝑛)
∞
Condition for stability 𝑕(𝑘) < ∞

𝑘=−∞
∞ ∞ ∞
𝑕(𝑘) = 3𝛿(𝑘) = 3𝛿(𝑘) = 3

𝑘=−∞ 𝑘=0 𝑘=0
∵ 𝛿 𝑘 = 0 𝑓𝑜𝑟 𝑘 ≠ 0 𝑎𝑛𝑑 𝛿 𝑘 = 1 𝑓𝑜𝑟 𝑘 = 0
∞
∵ 𝑕(𝑘) < ∞ the given system is 𝐬𝐭𝐚𝐛𝐥𝐞

𝑘=−∞
20
EnggTree.com
Unit 2: Analysis of continuous time signals

Fourier series analysis
The Fourier representation of signals can be used to perform frequency domain analysis of
signals in which we can study the various frequency components present in the signal, magnitude
and phase of various frequency components.
Conditions for existence of Fourier series:

The Fourier series exist only if the following Dirichlet’s conditions are satisfied.
• The signal 𝑥(𝑡) must be single valued function.
• The signal 𝑥(𝑡) must possess only a finite number of discontinuous in the period T.
• The signal must have a finite number of maxima and minima in the period T.
𝑇
• 𝑥(𝑡) must be absolutely integrable. i.e., ∫ |𝑥(𝑡)|𝑑𝑡 < ∞
0
Types of Fourier series:

• Trigonometric Fourier series
• Exponential Fourier series
• Cosine Fourier series
Trigonometric Fourier series
www.EnggTree.com
The trigonometric form of Fourier series of a periodic signal, 𝑥(𝑡) with period 𝑇 is defined
as
∞ ∞
𝑥(𝑡) = 𝑎0 + ∑ 𝑎𝑛 cos 𝑛Ω0𝑡 + ∑ 𝑏𝑛 sin 𝑛Ω0𝑡 … … … (1)

𝑛=1 𝑛=1
𝑎0, 𝑎𝑛 , 𝑏𝑛 → Fourier coefficients of trigonometric form of Fourier series
𝑡0+𝑇
1
𝑎0 = ∫ 𝑥(𝑡)𝑑𝑡
𝑇
𝑡0
𝑡0+𝑇
2
𝑎𝑛 = ∫ 𝑥(𝑡) cos 𝑛Ω0𝑡 𝑑𝑡
𝑇
𝑡0
𝑡0+𝑇
2
𝑏𝑛 = ∫ 𝑥(𝑡) sin 𝑛Ω0𝑡 𝑑𝑡
𝑇
𝑡0
1
EnggTree.com
Example 1 Find the trigonometric Fourier series for the periodic signal x(t) as shown in Figure
x(t)
1 x(t)
1
-7 -5 -3 -1 0 1 3 5 7 9 t
-1 --1 0 1 3 t
T -1
Solution:
𝑇 = 3 − (−1) = 4 𝑎𝑛𝑑 Ω0 = 2𝜋 = 𝜋
𝑇 2
Evaluation of 𝒂𝟎 1 1
[[ ]1 [ ]3] [ )
1 𝑡0 +𝑇 1 1 3
( )
𝒂𝟎 = ∫𝑡0 𝑥 𝑡 𝑑𝑡 = [∫−11 𝑑𝑡 + ∫1 −1 𝑑𝑡] = 𝑡 −1 − 1 𝑡 1 = (1 − (−1 ) − (3 − 1)]
𝑇 4 4 4
1
= [2 − 2] = 𝟎
4
Evaluation of 𝒂𝒏
2 𝑡0+𝑇 2 1 3
( )
𝒂𝒏 =
∫ 𝑥 𝑡 cos 𝑛𝛺0𝑡 𝑑𝑡 =
[∫ cos 𝑛𝛺0𝑡 𝑑𝑡 + ∫1 (−1)cos 𝑛𝛺0𝑡 𝑑𝑡]
𝑇 𝑡0 4 −1 1 3
1 sin 𝑛𝛺 𝑡 1 sin 𝑛𝛺 𝑡 3 1 sin 𝑛 𝜋 𝑡 sin 𝑛 𝜋 𝑡
2
= [[ 0 ] −[ 0 ] ] = [[ 𝜋2 ] −[ 𝜋 ] ]
2 𝑛𝛺0 −1 𝑛𝛺0 1 2 𝑛2 𝑛
−1 2 1
1 2 𝜋 𝜋 𝜋 𝜋
= ( ) [sin 𝑛 − (sin 𝑛 (−1)) − (sin 𝑛 (3) − sin 𝑛 )]
2 𝑛𝜋 2 1 2 𝜋 2 2 𝑛𝜋 𝑛𝜋
= [ ][ 𝜋 𝜋 𝜋 1 [3sin − sin(2n𝜋 − )]
]=
1 𝑛𝜋2 www.EnggTree.com
𝑛𝜋 sin 𝑛 + sin 𝑛 − sin 3𝑛 + sin 𝑛
2𝜋 𝟒2 𝝅2
𝑛𝜋 2 2
= [3sin − (−sin 𝑛 )] = [𝐬𝐢𝐧 𝒏 ]
𝑛𝜋 2 2 𝒏𝝅 𝟐
Evaluation of 𝒃𝒏
2 𝑡0+𝑇 2 1 3
( )
𝒃𝒏 = ∫𝑡0 𝑥 𝑡 sin 𝑛𝛺0𝑡 𝑑𝑡 = [∫−1 sin 𝑛𝛺0𝑡 𝑑𝑡 + ∫1 − sin 𝑛𝛺0𝑡 𝑑𝑡]
𝑇 4 1 3
1 −cos 𝑛𝛺 𝑡 1 −cos 𝑛𝛺 𝑡 3 1 −cos 𝑛 𝜋 𝑡 cos 𝑛 𝜋 𝑡
2
= [[ 0 ] −[ 0 ] ] = [[ 𝜋 2 ] +[ 𝜋 ] ]
2 𝑛𝛺0 −1 𝑛𝛺 0 1 2 𝑛 𝑛
𝜋 2 𝜋 2 −1 𝜋 2 1
1 −2 𝜋 (−1)) + (cos 𝑛 (3) − cos 𝑛 )]
= [ (
1 𝑛𝜋 cos
2 2 𝑛 − cos 𝑛 𝑛𝜋 𝑛𝜋𝜋 21 𝜋 2 𝜋
= [0 + (cos 2 (2𝑛𝜋 −2 ) − cos 𝑛 )] = [ ( )] = 𝟎
2 𝑛𝜋 2 2 𝑛𝜋 cos 𝑛 − cos 𝑛
2 2
Trigonometric Fourier
∞ ∞
series
𝑥(𝑡) = 𝑎0 + ∑ 𝑎𝑛 cos 𝑛𝛺0𝑡 + ∑ 𝑏𝑛 sin 𝑛𝛺0𝑡
𝑛=1
∞ 4 𝑛𝜋𝑛=1 ∞ 4 𝑛𝜋 𝜋
=∑ sin ( ) cos 𝑛𝛺 𝑡 = ∑ sin ( )
0 cos 𝑛 𝑡
𝑛𝜋 2 𝑛𝜋 2 2
𝑛=1 𝑛=1
Example 2 Obtain Fourier series of the following full wave rectified sine wave shown in figure
2
EnggTree.com
x(t)
-2 -1 0 1 2 t
Solution:
𝑥(𝑡) = 𝑥(−𝑡); ∴ 𝐺𝑖𝑣𝑒𝑛 𝑠𝑖𝑔𝑛𝑎𝑙 𝑖𝑠 𝑒𝑣𝑒𝑛 𝑠𝑖𝑔𝑛𝑎𝑙, 𝑠𝑜 𝒃𝒏 = 𝟎
2𝜋
𝑇 = 1 𝑎𝑛𝑑 Ω0 = = 2𝜋1
The given signal is sinusoidal signal, ∴ 𝑥(𝑡) = 𝐴 sin Ω𝑡
2π 2π
Here Ω = = = π and A = 1
T 2
∴ 𝒙(𝒕) = 𝐬𝐢𝐧 𝝅𝒕
Evaluation of 𝒂𝟎
𝑇 1 1
2 𝖥 2 2 1 cos 𝜋𝑡 12 2 𝜋 𝟐
𝒂 = 2 ∫ 𝑥(𝑡) 𝑑𝑡 = 2 ∫ 𝑥(𝑡) 𝑑𝑡 = I2 ∫ sin 𝜋𝑡 𝑑𝑡I = 2 [− ] = − [cos − cos 0] =
𝟎 𝑇 1 I I 𝜋 0 𝜋 2 𝝅
0 0
L 0
𝑇 1 1
4 2
4 2 2
𝒂 = ∫ 𝑥(𝑡) cos 𝑛Ω 𝑡 𝑑𝑡 = ∫ sin 𝜋𝑡 cos 𝑛2𝜋𝑡 𝑑𝑡 = 2 ∫[sin((1 + 2𝑛)𝜋𝑡) + sin((1 − 2𝑛)𝜋𝑡)] 𝑑𝑡
𝒏 0
𝑇 1
0 0 0
1
cos((1 + 2𝑛)𝜋𝑡) − cos((1 − 2𝑛)𝜋𝑡) 2
= 2 [− (1 + 2𝑛)𝜋 www.EnggTree.com
(1 − 2𝑛)𝜋
]
0
2 cos ((1 + 2𝑛) 𝜋) cos ((1 − 2𝑛) 𝜋) 1 1
= [− 2 − 2 + + ]
𝜋 1 + 2𝑛 1 − 2𝑛 1 + 2𝑛 1 − 2𝑛
2 1 1 2 1 − 2𝑛 + 1 + 2𝑛 𝟒
= [ + ]= [ 2
] =
𝜋 1 + 2𝑛 1 − 2𝑛 𝜋 1 − 4𝑛 𝝅(𝟏 − 𝟒𝒏𝟐)
Trigonometric Fourier series

∞ ∞
𝑥(𝑡) = 𝑎0 + ∑ 𝑎𝑛 cos 𝑛𝛺0𝑡 + ∑ 𝑏𝑛 sin 𝑛𝛺0𝑡

𝑛=1 𝑛=1
∞
2 4 cos 𝑛2𝜋𝑡
∴ 𝑥(𝑡 ) = +∑
𝜋 𝜋(1 − 4𝑛2)
𝑛=1
Exponential Fourier series

The exponential form of Fourier series of a periodic signal 𝑥(𝑡) with period 𝑇 is defined as,
∞
𝑥(𝑡) = ∑ 𝑐𝑛 𝑒𝑗𝑛 Ω0 𝑡
𝑛=−∞
The Fourier coefficient 𝑐𝑛 can be evaluated using the following formulae
3
EnggTree.com
𝑇
2
1
−𝑗𝑛 Ω0 𝑡 𝑑𝑡
𝑐𝑛 = 𝑇 ∫ 𝑥(𝑡)𝑒
𝑇
−2
Example 3 Find exponential series for the signal shown in figure

x(t)
0 1 2 3
Fig 2.26
Solution:
𝑇 = 1, Ω0 = 2π = 2π = 2π
T 1
Consider the equation of a straight line
𝑦 − 𝑦1 𝑥 − 𝑥1 … … … (9)
=
𝑦 2 − 𝑦 1 𝑥2 − 𝑥 1
Consider one period of the given signal Fig 2.26 as shown in Fig 2.27Consider x(t)
points P,Q as shown in fig 2.27 1 Q[1,1]

Coordinates of point 𝑃 = [0,0]
Coordinates of point 𝑄 = [1,1]
0 1 t
On substituting the coordinates of points 𝑃 and 𝑄 in eq (9) P[0,0]
𝑥(𝑡) − 0 𝑡 − 0 Fig 2.27
= ⇒ 𝑥(𝑡) = 𝑡
1−0 1−0
[∵ 𝑥 = 𝑡, 𝑦 = 𝑥(𝑡)]
Evaluation of 𝒄𝟎 www.EnggTree.com
1 𝑇 1 1 𝑡2 1 1
𝑐0 = ∫ 𝑥(𝑡)𝑑𝑡 = ∫ (𝑡)𝑑𝑡 = [ ] =
𝑇 0 1 0 2 0 2
Evaluation of 𝒄𝒏
1
1 𝑇 1 1 𝑒−𝑗𝑛 2𝜋𝑡 1 𝑒−𝑗𝑛 2𝜋𝑡
𝑑𝑡
𝑐𝑛 = ∫ 𝑥(𝑡)𝑒−𝑗𝑛 Ω0 𝑡 𝑑𝑡 = ∫ 𝑡𝑒−𝑗𝑛 2𝜋𝑡 𝑑𝑡 = [𝑡 ] − ∫ −𝑗𝑛2𝜋
𝑇 1 −𝑗𝑛2𝜋
0 0 0 0
1
𝑒−𝑗𝑛 2𝜋 𝑒−𝑗𝑛 2𝜋𝑡 1
𝑒−𝑗𝑛 2𝜋 𝑒−𝑗𝑛 2𝜋
= +0+[ 2 ] = 𝑗 + −
−𝑗𝑛2𝜋 −𝑗 (𝑛2𝜋)2 0 𝑛2𝜋 𝑛24𝜋2 𝑛24𝜋2
𝑗 1 1 𝑗
=
𝑛2𝜋 + 𝑛24𝜋2 − 𝑛24𝜋2 = 𝑛2𝜋
𝑗
𝑐𝑛 =
𝑗 𝑛2𝜋
𝑐 = , 𝑐 𝑗 , 𝑐 𝑗 , 𝑐 = 𝑗 , 𝑐 𝑗 , 𝑐 𝑗
= = = =
1 2𝜋 2 4𝜋 3 6𝜋 −1 −2𝜋 −2 −4𝜋 −3 −6𝜋
Exponential Fourier
∞
series
𝑥(𝑡) = ∑ 𝑐𝑛 𝑒𝑗𝑛 𝛺0 𝑡
𝑛=−∞
4
EnggTree.com
𝑗 𝑗 𝑗 1 𝑗 𝑗 𝑗
∴ 𝑥(𝑡) = + ⋯ − 𝑒−𝑗 6𝜋𝑡 − 𝑒−𝑗 4𝜋𝑡 − 𝑒−𝑗 2𝜋𝑡 𝑒𝑗
+ 𝑒𝑗 4𝜋𝑡 +2𝜋𝑡
𝑒𝑗 6𝜋𝑡 + ⋯
+ +
6𝜋 4𝜋 2𝜋 2 2𝜋 4𝜋 6𝜋
1 𝑗 𝑗 2𝜋𝑡 𝑗 𝑗 4𝜋𝑡 𝑗
= + [𝑒 −𝑒 −𝑗 2𝜋𝑡 ]+ [𝑒 −𝑒 −𝑗 4𝜋𝑡 ]+ [𝑒𝑗 6𝜋𝑡 − 𝑒−𝑗 6𝜋𝑡 ] + ⋯
2 2𝜋 4𝜋 6𝜋
1 1 𝑒𝑗 2𝜋𝑡 − 𝑒−𝑗 2𝜋𝑡 1 𝑒𝑗 4𝜋𝑡 − 𝑒−𝑗 4𝜋𝑡 1 𝑒𝑗 6𝜋𝑡 − 𝑒−𝑗 6𝜋𝑡
= + [ ] [ ] [ ]
2 𝜋 (−1)2𝑗 + (−1)2𝑗 + (−1)2𝑗
1 −1 1 2𝜋 1 3𝜋
= + ( ) 𝑠𝑖𝑛 2𝜋𝑡 − 𝑠𝑖𝑛 4𝜋𝑡 − sin 6𝜋𝑡
2 𝜋 2𝜋 3𝜋
1 1 1 1
[ ]
= − 𝑠𝑖𝑛 2𝜋𝑡 + 𝑠𝑖𝑛 4𝜋𝑡 + 𝑠𝑖𝑛 6𝜋𝑡 + ⋯
2 𝜋 2 3
Cosine Fourier series
Cosine representation of 𝑥(𝑡) is

∞
𝑥(𝑡) = 𝐴0 + ∑ 𝐴𝑛 𝑐𝑜𝑠 (𝑛𝛺0𝑡 + 𝜃𝑛 )

𝑛=1
Where 𝐴0 is dc component, 𝐴𝑛 is harmonic amplitude or spectral amplitude and 𝜃𝑛 is phase
coefficient or phase angle 𝑜𝑟 𝑠𝑝𝑒𝑐𝑡𝑟𝑎𝑙 𝑎𝑛𝑔𝑙𝑒
Example 4 www.EnggTree.com
Determine the cosine Fourier series of the signal shown in Figure
x(t)
-2π -π 0 π 2π 3π t
Solution:
The signal shown in is periodic with period 𝑇 = 2𝜋 𝑎𝑛𝑑 Ω = 2𝜋 = 1
0 2𝜋
The given signal is sinusoidal signal, ∴ 𝑥(𝑡) = 𝐴 sin Ω𝑡
2π 2π
Here Ω = = = 1, A = 1
T 2π
∴ 𝒙(𝒕) = 𝒔𝒊𝒏 𝒕
Evaluation of 𝒂𝟎
1 𝑇 1 𝜋 1 1 1 1
[− 𝑐𝑜𝑠 𝑡 ]𝜋
𝑎0 = ∫0 𝑥(𝑡)𝑑𝑡 = ∫ 𝑠𝑖𝑛 𝑡 𝑑𝑡 = 0 = [− 𝑐𝑜𝑠 𝜋 + 𝑐𝑜𝑠 0] = [2] =
𝑇 2𝜋 0 2𝜋 2𝜋 2𝜋 𝜋
5
EnggTree.com
𝑇
2
( ) 2 𝜋 1 𝜋
[ ( ) ( ) ]
𝑎𝑛 = ∫0 𝑥 𝑡 𝑐𝑜𝑠 𝑛𝛺0𝑡 𝑑𝑡 = ∫ 𝑠𝑖𝑛 𝑡 𝑐𝑜𝑠 𝑛𝑡 𝑑𝑡 = ∫0 𝑠𝑖𝑛 1 + 𝑛 𝑡 + 𝑠𝑖𝑛 1 − 𝑛 𝑡 𝑑𝑡
𝑇 2𝜋 0 2𝜋
1 𝑐𝑜𝑠(1 + 𝑛)𝑡 𝑐𝑜𝑠(1 − 𝑛)𝑡 𝜋
= [− − ]
2𝜋 (1 + 𝑛) (1 − 𝑛) 0
1 𝑐𝑜𝑠(1 + 𝑛)𝜋 𝑐𝑜𝑠(1 − 𝑛)𝜋 1 + 1 ]
= [− − +
2𝜋 (1 + 𝑛) (1 − 𝑛) 1+𝑛 1−𝑛
1 [− 1 1 1 1
𝑓𝑜𝑟 𝑛 = 𝑜𝑑𝑑 ∶ 𝑎𝑛 = − + + ]=0
2𝜋 1 + 𝑛 1 − 𝑛 1 + 𝑛 1 − 𝑛
1 1 1 1 1 ]= 1 2 + 2 )
𝑓𝑜𝑟 𝑛 = 𝑒𝑣𝑒𝑛 ∶ 𝑎𝑛 = [ + + + (
2𝜋 1 + 𝑛 1 − 𝑛 1 + 𝑛 1 − 𝑛 2𝜋 1 + 𝑛 1 − 𝑛
1 1−𝑛+1+𝑛 2
= [ ]=
𝜋 1 − 𝑛2 𝜋(1 − 𝑛2)
𝟎 𝒇𝒐𝒓 𝒏 = 𝒐𝒅𝒅
∴ 𝒂𝒏 = { 𝟐
𝒇𝒐𝒓 𝒏 = 𝒆𝒗𝒆𝒏
𝝅(𝟏 − 𝒏𝟐)
Evaluation of 𝒃𝒏
2 𝑇 2 𝜋 1 𝜋( ( ) ( ) )
𝒃𝒏 = ∫0 𝑥(𝑡) 𝑠𝑖𝑛 𝑛𝛺0𝑡 𝑑𝑡 = ∫ 𝑠𝑖𝑛 𝑡 𝑠𝑖𝑛 𝑛𝑡 𝑑𝑡 = ∫ 𝑐𝑜𝑠 1 − 𝑛 𝑡 − 𝑐𝑜𝑠 1 + 𝑛 𝑡 𝑑𝑡
𝑇 2𝜋 0 2𝜋 0
1 𝑠𝑖𝑛(1 − 𝑛)𝑡 𝑠𝑖𝑛(1 + 𝑛)𝑡 𝜋 1 𝑠𝑖𝑛(1 − 𝑛)𝜋 𝑠𝑖𝑛(1 + 𝑛)𝜋
= [ − ] = [ − − 0] = 𝟎
2𝜋 (1 − 𝑛) (1 + 𝑛) 0 2𝜋 (1 − 𝑛) (1 + 𝑛)
Evaluation of Fourier coefficients of Cosine Fourier series from Trigonometric Fourier series:
1
www.EnggTree.com
𝐴0 = 𝑎0 =
𝜋
2 2
𝐴𝑛 = √𝑎𝑛 2 + 𝑏𝑛 = , 𝑓𝑜𝑟 𝑛 = 𝑒𝑣𝑒𝑛
𝜋(1 − 𝑛2)
𝑏𝑛
𝜃𝑛 = −𝑡𝑎𝑛 =0
−1
𝑎𝑛
Cosine Fourier
series ∞
x(t) = A0 + ∑ An cos(nΩ0t + θn )
n=1
∞
1 2 1 2 2
∑
∴ 𝑥(𝑡) = + 𝑛=1 𝑐𝑜𝑠 𝑛𝑡 = + 𝑐𝑜𝑠 2𝑡 + 𝑐𝑜𝑠 4𝑡 + ⋯
𝜋 (𝑛=𝑒𝑣𝑒𝑛 ) 𝜋(1 − 𝑛2) 𝜋 𝜋(1 − 4) 𝜋(1 − 16)
1 2 2 1 2 1 1
= − 𝑐𝑜𝑠 2𝑡 − 𝑐𝑜𝑠 4𝑡 + ⋯ = − [ 𝑐𝑜𝑠 2𝑡 + 𝑐𝑜𝑠 4𝑡 + ⋯ ]
𝜋 3𝜋 15𝜋 𝜋 𝜋 3 15
Fourier transform
The Fourier representation of periodic signals has been extended to non-periodic signals by
letting the fundamental period T tend to infinity and this Fourier method of representing non-
periodic signals as a function of frequency is called Fourier transform.
Definition of Continuous time Fourier Transform
6
EnggTree.com
The Fourier transform (FT) of Continuous time signals is called Continuous Time Fourier
Transform
𝐿𝑒𝑡 𝑥(𝑡) = 𝐶𝑜𝑛𝑡𝑖𝑛𝑢𝑜𝑢𝑠 𝑡𝑖𝑚𝑒 𝑠𝑖𝑔𝑛𝑎𝑙
𝑋(𝑗Ω) = 𝐹{𝑥(𝑡)}
The Fourier transform of continuous time signal, x(t) is defined as,
∞
𝑋(𝑗Ω) = 𝐹{𝑥(𝑡)} = ∫ 𝑥(𝑡)𝑒−𝑗Ω𝑡 𝑑𝑡

−∞
Conditions for existence of Fourier transform
The Fourier transform 𝑥(𝑡) exist if it satisfies the following Dirichlet condition
1. 𝑥(𝑡) should be absolutely integrable
∞
𝑖𝑒 , ∫ 𝑥(𝑡)𝑑𝑡 < ∞
−∞
2. 𝑥(𝑡) should have a finite number of maxima and minima with in any finite interval.
3. 𝑥(𝑡) should have a finite number of discontinuities with in any interval.
Definition of Inverse Fourier Transform

The inverse Fourier Transform of 𝑋(𝑗Ω) is defined as,
1 ∞
𝑥(𝑡) = 𝐹−1{𝑋(𝑗Ω)} =
2𝜋 ∫ 𝑋(𝑗Ω)𝑒𝑗Ω𝑡 𝑑Ω
www.EnggTree.com
−∞
Example 5 Find Fourier transform of impulse signal

Solution:
By definition of Fourier transform
∞
𝐹{𝑥(𝑡)} = 𝑋(𝑗Ω) = ∫ 𝑥(𝑡)𝑒−𝑗Ω𝑡 𝑑𝑡

−∞
∞
∴ 𝐹[𝛿(𝑡)] = ∫ 𝛿(𝑡)𝑒−𝑗 𝛺𝑡 𝑑𝑡
−∞
1 𝑓𝑜𝑟 𝑡 = 0
𝐹[𝛿(𝑡)] = 𝛿(0)𝑒−𝑗 𝛺(0) =1 [∵ 𝐼𝑚𝑝𝑢𝑙𝑠𝑒 𝑠𝑖𝑔𝑛𝑎𝑙 𝛿(𝑡) = { ]
0 𝑓𝑜𝑟 𝑡 ≠ 0
Example 6 Find Fourier transform of double sided exponential signal

Solution:
Double sided exponential signal is given by
𝑒−𝑎𝑡 ∶𝑡≥0
𝐹[𝑒−𝑎|𝑡|] = {
𝑒 𝑎𝑡 ∶𝑡≤0
0 ∞ 0 ∞
𝐹[𝑒−𝑎|𝑡|] = ∫ 𝑒𝑎𝑡 𝑒−𝑗 𝛺𝑡 𝑑𝑡 + ∫ 𝑒−𝑎𝑡 . 𝑒−𝑗 𝛺𝑡 𝑑𝑡 = ∫ 𝑒(𝑎−𝑗 𝛺)𝑡 𝑑𝑡 + ∫ 𝑒−(𝑎+𝑗 𝛺)𝑡 𝑑𝑡
−∞ 0 −∞ 0
0 ∞ 1 𝑎+𝑗𝛺+𝑎−𝑗𝛺
=[ 𝑒(𝑎−𝑗 𝛺)𝑡 ] +[ 𝑒−(𝑎+𝑗 𝛺)𝑡 ] 1
𝑎−𝑗𝛺 −(𝑎 + 𝑗 𝛺 = + =
) 𝑎−𝑗𝛺 𝑎+𝑗𝛺 𝑎2 + 𝛺2
−∞ 0
2𝑎
=
𝑎2 + 𝛺2
7
EnggTree.com
Example 7 Find Fourier transform of rectangular pulse function shown in figure

x(t)
-T/2 0 T/2 t
Solution:
−𝑇 𝑇
𝑥(𝑡) = 𝜋(𝑡) = 𝐴 ; ≤𝑡≤
2 2
𝑇 𝑇 𝑇
𝑇
2 𝑒−𝑗𝛺𝑡 2 𝐴 𝑇
−𝑗𝛺 2 𝑗𝛺
𝑇 2𝐴 𝑒 𝑗𝛺 2 − 𝑒 −𝑗𝛺 2 2𝐴 𝑠𝑖𝑛 𝛺 𝑇
𝐹[𝜋(𝑡)] = ∫ 𝐴𝑒 −𝑗𝛺𝑡 𝑑𝑡 = 𝐴 [ ] = [𝑒 − 𝑒 2 ]= [ ]=
−𝑇2
−𝑗𝛺 𝑇 −𝑗𝛺 𝑗𝛺 2 𝛺 2
− 2
2𝐴 𝑇 𝑠𝑖𝑛 𝛺2𝑇 𝑇
= 𝑇 𝑠𝑖𝑛 𝛺 = 𝐴𝑇 𝑇 𝐴𝑇 𝑠𝑖𝑛 𝑐 𝛺
𝛺𝑇 2 𝛺2 2
Example 8 Find inverse Fourier transform 𝑋(𝑗 𝛺) = 𝛿(𝛺)

Solution:
∴ 𝐹−1[𝑋(𝑗 𝛺)] =
1 𝐹−1
∞ [𝛿(𝛺)]
1
( ) ( ) 𝑗 𝛺𝑡 1 ∞ ) 𝑗 𝛺𝑡 [ ] 1 𝑓𝑜𝑟 𝛺 = 0
( )
𝑥 𝑡 = ∫ 𝑋 𝑗𝛺 𝑒 𝑑𝛺 = ∫ 𝛿 ( 𝛺 𝑒 𝑑𝛺 = 1 ∵ 𝛿 𝛺 ={
2𝜋 −∞ 2𝜋 −∞ 2𝜋 0 𝑓𝑜𝑟 𝛺 ≠ 0
www.EnggTree.com
𝐹−1[𝛿(𝛺)] =
1
2𝜋
Laplace transform
It is used to transform a time domain to complex frequency domain signal(s-domain)
Two Sided Laplace transform (or) Bilateral Laplace transform

Let 𝑥(𝑡) be a continuous time signal defined for all values of 𝑡. Let 𝑋(𝑆) be Laplace transform of
𝑥(𝑡).
∞
𝐿{𝑥(𝑡)} = 𝑋(𝑆) = ∫ 𝑥(𝑡)𝑒−𝑆𝑡 𝑑𝑡

−∞
One sided Laplace transform (or) Unilateral Laplace transform

Let 𝑥(𝑡) be a continuous time signal defined for 𝑡 ≥ 0 (ie If 𝑥(𝑡) is causal) then,
∞
𝐿{𝑥(𝑡)} = 𝑋(𝑆) = ∫ 𝑥(𝑡)𝑒−𝑆𝑡 𝑑𝑡

0
Inverse Laplace transform

The S-domain signal 𝑋(𝑆) can be transformed to time domain signal x(t) by using inverse Laplace
transform.
8
EnggTree.com
The inverse Laplace transform of X(S) is defined as,

1 𝑠=𝜎+𝑗Ω
𝐿−1{𝑋(𝑠)} = 𝑥(𝑡) =
2𝜋𝑗 ∫ 𝑋(𝑆)𝑒𝑠𝑡 𝑑𝑠
𝑠=𝜎−𝑗Ω
Existence of Laplace transform

The necessary and sufficient conditions for the existence of Laplace transform are
• 𝑥(𝑡) should be continuous in the given closed interval
• 𝑥(𝑡)𝑒−𝜎𝑡 must be absolutely intergrable
∞
𝑖. 𝑒. , 𝑋(𝑆) 𝑒𝑥𝑖𝑠𝑡𝑠 𝑜𝑛𝑙𝑦 𝑖𝑓 ∫−∞|𝑥(𝑡)𝑒−𝜎𝑡 |𝑑𝑡 < ∞
Example 9 Find unilateral Laplace transform for the following signals
𝒊) 𝒙(𝒕) = 𝜹(𝒕)
( ) ∞ ( ) −st ∞ −st −s(0) () 1 for t = 0
()
X S = ∫ x t e dt = ∫ δ t e dt = e =1 ∵δ t ={
0 0 0 for t ≠ 0
𝒊𝒊) 𝒙(𝒕) = 𝒖(𝒕)

∞ 1 for t ≥ 0
∞ ∞ 𝑒−𝑠𝑡
∞ 1
−st −st
www.EnggTree.com
𝑋(𝑆) = ∫0 x(t)e dt = ∫0 u(t)e dt = ∫0 1𝑒 𝑑𝑡 = [
−𝑠𝑡
−𝑠 0
] =
𝑠
∵ u(t) = {
0 for t < 0
Example 10 Find Laplace transform of 𝑥(𝑡) = 𝑒𝑎𝑡 𝑢(𝑡)

Solution:
∞ ∞ ∞
−(𝑠−𝑎)𝑡
𝑋(𝑆) = 𝐿[𝑒𝑎𝑡 𝑢(𝑡)] = ∫ 𝑒𝑎𝑡 𝑒−𝑠𝑡 𝑑𝑡 = ∫ 𝑒−(𝑠−𝑎)𝑡 𝑑𝑡 = [ 𝑒 ] = 1
(
− 𝑠 − 𝑎) 0 𝑠−𝑎
0 0
Example 11 Determine initial value and final value of the following signal 𝑋(𝑆) = 1
𝑠( 𝑠+2)
Solution:
Initial value
1 1
𝑥(0) = Lt 𝑆𝑋(𝑆) = Lt 𝑠 = =0
𝑠→∞ 𝑠→∞ 𝑠(𝑠 + 2) ∞
Final value
1 1
𝑥(∞) = Lt 𝑆𝑋(𝑆) = Lt 𝑠 =
𝑠→0 𝑠→0 𝑠(𝑠 + 2) 2
2+9𝑆+1
Example 12 Find inverse Laplace Transform of 𝑋(𝑆) = 𝑆
𝑆[𝑆2 +6𝑆+8]. Find ROC for 𝑖) 𝑅𝑒(𝑠) > 0
𝑖𝑖) 𝑅𝑒(𝑠) < −4 𝑖𝑖𝑖) − 2 > 𝑅𝑒(𝑠) > −4
Solution:
9
EnggTree.com
𝑆2 + 9𝑆 + 1 𝑆2 + 9𝑆 + 1 𝐴 𝐵 𝐶
( )
X 𝑆 = = = + +
𝑆[𝑆2 + 6𝑆 + 8] 𝑆(𝑆 + 4)(𝑆 + 2) 𝑆 (𝑆 + 4) (𝑆 + 2)
𝑆2 + 9𝑆 + 1 = 𝐴(𝑆 + 4)(𝑆 + 2) + B𝑆(𝑆 + 2) + C𝑆(𝑆 + 4)
at 𝑆 = 0 𝑆 = −4 𝑆 = −2
1 19 13
𝐴= 𝐵=− 𝐶=
8 8 1 19 13 4
∴ 𝑋(𝑆) = 8 + − 8 4
( +
𝑆 𝑆 + 4 ) (𝑆 + 2)
Applying inverse Laplace transform 1
( ) ( ) 19 −4𝑡 13 −2𝑡
𝑥 𝑡 = 𝑢 𝑡 − 𝑒 𝑢(𝑡) + 𝑒 𝑢(𝑡)
8 8 4
ROC
𝒊) 𝑹𝒆(𝒔) > 𝟎
jΩ
ROC lies right
1 side of all poles
( ) ( ) 19 −4𝑡 13 −2𝑡
∴𝑥 𝑡 = 𝑢 𝑡 − 𝑒 𝑢(𝑡) + 𝑒 𝑢(𝑡)
-4 -2 0 σ 8 8 4
𝒊𝒊) 𝑹𝒆(𝒔) < −𝟒
jΩ
ROC lies left side of all poles
www.EnggTree.com 1
∴ 𝑥(𝑡) = − 𝑢(−𝑡) + 𝑒 −4𝑡
19 13
-4 -2 0 σ
𝑢(−𝑡) − 𝑢(−𝑡)
8 8
𝑒−2𝑡
4
𝒊𝒊𝒊) − 𝟐 > 𝑅𝑒(𝒔) > −𝟒

ROC lies left side of poles s=-2, s=0 and right side of
jΩ pole s=-4 1
( ) ( ) 19 −4𝑡 ( ) 13 −2𝑡
∴ 𝑥 𝑡 = − 𝑢 −𝑡
8
-4 -2 0 σ
10
EnggTree.com
Unit 3: Linear Time Invariant-Continuous

Time Systems
LTI-CT (Linear Time Invariant-Continuous Time) Systems
When continuous time system satisfies the properties of linearity and time invariant
then it is called an LTI-CT (Linear Time Invariant-Continuous Time) System.
Impulse Response
When the input to a continuous time system is an unit impulse signal δ(t) then the
output is called an impulse response of the system and it is denoted by h(t)
Impulse response, h(t) = H{𝛿(t)}
Continuous time system

Impulse input
𝛿(𝑡) H h(t)
Fig 3.1
Convolution Integral
∞
𝑦(𝑡) = ∫ 𝑥(𝑟)𝑕(𝑡 − 𝑟)𝑑𝑟
www.EnggTree.com
−∞
This is called convolution integral or simply convolution. The convolution of two signal x(t) and
h(t) can be represented as
𝑦(𝑡) = 𝑥(𝑡) ∗ 𝑕(𝑡)
Systems connected in series/parallel(Block diagram representation)

System Realization
There are four types of system realization in continuous time linear time invariant
systems.
They are
• Direct form I realization
• Direct form II realization
• Cascade form realization
• Parallel form realization
Direct form I realization

It is the direct implementation of differential equation or transfer function describing the
system. It uses separate integrators for input and output variables. It provides direct relation
between time domain and s-domain equations. In general, this form requires 2N delay elements
(for both input and output signals) for a filter of order N. This form is practical for small filters.
Advantages:
• Simplicity
• Most straight forward realization
202
EnggTree.com
Disadvantages:
• More number of integrators are used
• Inefficient and impractical (numerically unstable) for complex design
Direct form II realization
It is the direct implementation of differential equation or transfer function describing the

system. Instead of using separate integrators for integrating input and output variables
separately, an intermediate variable is integrated. It provides direct relation between time
domain and s-domain equations.
Advantages:
• It uses minimum number of integrators
• Straight forward realization
Disadvantages:
• It increases the possibility of arithmetic overflow for filters of high Q or resonance
Cascade form
In cascade form realization the given transfer function is expressed as a product of several
transfer function and each of these transfer function is realized in direct form II and then all those
realized structures are cascaded i.e., is connected in series.
Parallel form realization
The given transfer function is expressed into its partial fractions and each factor is
www.EnggTree.com
realized in direct form II and all those realized structures are connected in parallel.
Solved Problems
Example 3.1: Find the convolution by graphical method
1 𝑓𝑜𝑟 0 ≤ 𝑡 ≤ 2 1 𝑓𝑜𝑟 0 ≤ 𝑡 ≤ 3
𝑥(𝑡) = { ; 𝑕(𝑡) = {
0 𝑜𝑡𝑕𝑒𝑟𝑤𝑖𝑠𝑒 0 𝑜𝑡𝑕𝑒𝑟𝑤𝑖𝑠𝑒
Solution:
∞
𝐼𝑛 𝑔𝑒𝑛𝑒𝑟𝑎𝑙 𝑥1(𝑡) ∗ 𝑥2(𝑡) = ∫ 𝑥1(𝑟) 𝑥2(𝑡 − 𝑟) 𝑑𝑟

−∞
∞
𝑆𝑖𝑚𝑖𝑙𝑎𝑟𝑙𝑦 𝑕(𝑡) ∗ 𝑥(𝑡) = ∫ 𝑕(𝑟) 𝑥(𝑡 − 𝑟) 𝑑𝑟

−∞
Replacing 𝑡 by 𝑟 in 𝑥(𝑡)and 𝑕(𝑡)
𝑥(𝑟) = {1 𝑓𝑜𝑟 0 ≤ 𝑟 ≤ 2 ; 𝑕(𝑟) = {1 𝑓𝑜𝑟 0 ≤ 𝑟 ≤ 3
0 𝑜𝑡𝑕𝑒𝑟𝑤𝑖𝑠𝑒 0 𝑜𝑡𝑕𝑒𝑟𝑤𝑖𝑠𝑒
203
2
EnggTree.com
h(τ) x(τ) x(-τ)
1 1 1
0 3 τ 0 2 τ -2 0 τ
Fig 3.21 Fig 3.22 Fig 3.23
Case (i) 𝒕 < 0
h(τ)x(t-τ) Since overlap is absent between 𝑕(𝑟) and 𝑥(−𝑟 + 𝑡)

∴ 𝑦(𝑡) = 𝑕(𝑡) ∗ 𝑥(𝑡) = 0
1
t-2 t 0 3 τ
Fig 3.24
Case (ii) 𝟎 ≤ 𝒕 < 2
h(τ)x(t-τ) Since overlap is present

∞ 𝑡
1 ∴ 𝑦(𝑡) = 𝑕(𝑡) ∗ 𝑥(𝑡) = ∫ 𝑕(𝑟) 𝑥(𝑡 − 𝑟) 𝑑𝑟 = ∫(1)(1) 𝑑𝑟 = [𝑟]𝑡0 = 𝑡
t-2 0 t 3 τ −∞ 0
Fig 3.25
www.EnggTree.com
Case (iii) 𝟐 ≤ 𝒕 < 3

∞ 𝑡
1 ∴ 𝑦(𝑡) = 𝑕(𝑡) ∗ 𝑥(𝑡) = ∫ 𝑕(𝑟) 𝑥(𝑡 − 𝑟) 𝑑𝑟 = ∫(1)(1) 𝑑𝑟 = [𝑟]𝑡−2
𝑡 =2
0 t-2 t 3 τ −∞ 𝑡−2
Fig 3.26
Case (iv) 𝟑 ≤ 𝒕 < 5

∞ 3
1 ∴ 𝑦(𝑡) = 𝑕(𝑡) ∗ 𝑥(𝑡) = ∫ 𝑕(𝑟) 𝑥(𝑡 − 𝑟) 𝑑𝑟 = ∫(1)(1) 𝑑𝑟 = [𝑟]3𝑡−2

0 t-2 3 t τ −∞ 𝑡−2
=5−𝑡
Fig 3.27
Case (v) 𝒕 > 5
h(τ)x(t-τ) Since overlap is absent

∴ 𝑦(𝑡) = 𝑕(𝑡) ∗ 𝑥(𝑡) = 0
1
0 3 t-2 t τ
Fig 3.28
204

EnggTree.com
Linear Time Invariant-Continuous Time systems
0 𝑓𝑜𝑟 𝑡<0
𝑡 𝑓𝑜𝑟 0 ≤ 𝑡 < 2
∴ 𝑦(𝑡) = 𝑕(𝑡) ∗ 𝑥(𝑡) = 2 𝑓𝑜𝑟 2 ≤ 𝑡 < 3
5−𝑡 𝑓𝑜𝑟 3 ≤ 𝑡 < 5
𝗅0 𝑓𝑜𝑟 𝑡≥5
Example 3.2: Find impulse response of the following equation

𝑑2𝑦(𝑡) 𝑑𝑦(𝑡)
+ 5 + 6𝑦(𝑡) = 𝑥(𝑡)
𝑑𝑡2 𝑑𝑡
Solution:
+5 + 6𝑦(𝑡) = 𝑥(𝑡)
𝑑𝑡2 𝑑𝑡
Assume all the initial conditions are zero

Applying Laplace transform of the given equation
𝑆2𝑌(𝑆) + 5𝑆𝑌(𝑆) + 6𝑌(𝑆) = 𝑋(𝑆)
𝑌(𝑆)(𝑆2 + 5𝑆 + 6) = 𝑋(𝑆)
𝑌(𝑆) 1
Transfer function 𝐻(𝑆) = =
𝑋(𝑆) (𝑆2 +5𝑆+6)
1
𝐻(𝑆) = 𝑌(𝑆) = (∵ 𝐹𝑜𝑟 𝑖𝑚𝑝𝑢𝑙𝑠𝑒 𝑖𝑛𝑝𝑢𝑡 𝑥(𝑡) = 𝛿(𝑡) => 𝑋(𝑆) = 1)
𝑆2 + 5𝑆 + 6
1 𝐴 𝐵
www.EnggTree.com
𝐻(𝑆) = = +
(𝑆 + 3)(𝑆 + 2) 𝑆 + 3 𝑆 + 2
1 = 𝐴(𝑆 + 2) + 𝐵(𝑆 + 3)
at 𝑆 = −3 𝑆 = −2
𝐴 = −1 𝐵=1
1 1
∴ 𝐻(𝑆) = − +
𝑆+3 𝑆+2
Applying Inverse Laplace transform
𝒉(𝒕) = −𝒆−𝟑𝒕𝒖(𝒕) + 𝒆−𝟐𝒕𝒖(𝒕)
Example 3.3: Using Laplace transform solve differential equation

𝑑2𝑦(𝑡)
+ 𝑦(𝑡) = 𝑥(𝑡)
𝑑𝑡2
Where y′ (0) = 2; 𝑦(0) = 1; 𝑖𝑛𝑝𝑢𝑡 𝑥(𝑡) = 𝐶𝑜𝑠2𝑡
Solution:
𝑑2𝑦(𝑡)
+ 𝑦(𝑡) = 𝑥(𝑡)
𝑑𝑡2
Applying Laplace transform
𝑆2𝑌(𝑆) − 𝑆𝑦(0) − 𝑦′ (0) + 𝑌(𝑆) = 𝑋(𝑆)
𝑆
𝑆2𝑌(𝑆) − 𝑆 − 2 + 𝑌(𝑆) =
𝑆2 + 4
𝑆
∴ 𝑌(𝑆)(𝑆2 + 1) = +𝑆+2
𝑆2 + 4
4 205

EnggTree.com
Faculty: Mrs.M.Shakunthala & Dr.M.N.Vimal Kumar
𝑆 𝑆 2
𝑌(𝑆) = + +
(𝑆2 + 4)(𝑆2 + 1) (𝑆2 + 1) (𝑆2 + 1)
𝑆 𝐴𝑆 + 𝐵 𝐶𝑆 + 𝐷
𝐿𝑒𝑡 2 = +
(𝑆 + 4)(𝑆2 + 1) (𝑆2 + 4) (𝑆2 + 1)
𝑆 = (𝐴𝑆 + 𝐵)(𝑆2 + 1) + (𝐶𝑆 + 𝐷)(𝑆2 + 4)
𝑆 = 𝐴𝑆3 + 𝐵𝑆2 + 𝐴𝑆 + 𝐵 + 𝐶𝑆3 + 𝐷𝑆2 + 4𝐶𝑆 + 4𝐷
Comparing constant term
0 = 𝐵 + 4𝐷
𝐵 = −4𝐷 … (7)
Comparing coeff of 𝑆3
0=𝐴+𝐶
𝐴 = −𝐶 … (8)
Comparing coeff of 𝑆 2
0= 𝐵+𝐷 … (9)
Comparing coeff of S
1 = 𝐴 + 4𝐶 … (10)
Substitute 𝑒𝑞 (8) in 𝑒𝑞 (10) and 𝑒𝑞 (7) in 𝑒𝑞 (9)
𝟏
𝑪= ,𝑫 = 𝟎
𝟑
Substitute value of C and D in 𝑒𝑞 (8) and 𝑒𝑞 (7)
𝟏
𝑨 = − ,𝑩 = 𝟎
𝟑
1 1
𝑆 −3𝑆 𝑆
www.EnggTree.com
∴ 2 3
(𝑆 + 4)(𝑆2 + 1) = (𝑆2 + 4) + (𝑆2 + 1)
−13𝑆 + 3𝑆
1 𝑆 2
𝑌(𝑆) = + 2 + 2
(𝑆2 + 4) (𝑆2 + 1) (𝑆 + 1) (𝑆 + 1)
−1 𝑆 4
𝑆 2
3
𝑌(𝑆) = + 23 + 2
(𝑆2 + 4) (𝑆 + 1) (𝑆 + 1)
Taking Inverse Laplace transform
𝟏 𝟒
𝒚(𝒕) = − 𝒄𝒐𝒔𝟐𝒕 𝒖(𝒕) + 𝒄𝒐𝒔𝒕 𝒖(𝒕) + 𝟐𝒔𝒊𝒏𝒕 𝒖(𝒕)
𝟑 𝟑
Example 3.4: Find step response of the circuit shown in Fig 3.30
R
x(t) L y(t)
i(t)
Fig 3.30
Solution:
Applying KVL to the circuit shown in Fig 3.30
𝑑𝑖(𝑡) 𝑑𝑖(𝑡)
𝑥(𝑡) = 𝑅𝑖(𝑡) + 𝐿 𝑦(𝑡) = 𝐿
𝑑𝑡 𝑑𝑡
Applying Laplace transform
𝑋(𝑆) = 𝑅𝐼(𝑆) + 𝐿𝑆𝐼(𝑆) 𝑌(𝑆) = 𝐿𝑆𝐼(𝑆)
𝑋(𝑆) = [𝑅 + 𝐿𝑆]𝐼(𝑆)
206

EnggTree.com
𝑋(𝑆)
𝐼(𝑆) =
[𝑅 + 𝐿𝑆]
𝑋(𝑆)
𝑌(𝑆) = 𝐿𝑆
[𝑅 + 𝐿𝑆]
1
For Step response 𝑥(𝑡) = 𝑢(𝑡) => 𝑋(𝑆) =
𝑆
1 𝐿 1
𝑌(𝑆) = 𝐿𝑆 𝑆 = =
𝑅 + 𝐿𝑆 𝐿𝑆 + 𝑅 𝑅
𝑆+
𝐿
Applying Inverse Laplace transform
𝑹
𝒚(𝒕) = 𝒆−𝑳𝒕𝒖(𝒕)
Example 3.5: Solve the differential equation using Fourier transform

+ 6 + 8𝑦(𝑡) = 2𝑥(𝑡)
𝑑𝑡2 𝑑𝑡
(i) Find the impulse response of the system
(ii) What is the response of the system if 𝑥(𝑡) = 𝑡𝑒−2𝑡𝑢(𝑡)
Solution:
+6 + 8𝑦(𝑡) = 2𝑥(𝑡)
𝑑𝑡2 𝑑𝑡
Applying Fourier transform
(𝑗𝛺)2Y(𝑗Ω) + 6𝑗Ω𝑌(𝑗Ω) + 8𝑌(𝑗Ω) = 2X(𝑗Ω)
Y(𝑗Ω)[(𝑗𝛺)2 + 6𝑗Ω + 8] = 2X(𝑗Ω)
www.EnggTree.com
H(𝑗Ω) =
Y(𝑗Ω)
=
2
X(𝑗Ω) [(𝑗𝛺)2 + 6𝑗Ω + 8]
(i) Impulse response 𝑥(𝑡) = 𝛿(𝑡) => 𝑋(𝑗Ω) = 1

2 𝐴 𝐵
∴ H(𝑗Ω) = Y(𝑗Ω) = = +
[(𝑗𝛺)2 + 6𝑗Ω + 8] 𝑗Ω + 4 𝑗Ω + 2
2 = 𝐴(𝑗Ω + 2) + 𝐵(𝑗Ω + 4)
at 𝑗Ω = −4 𝑗𝛺 = −2
𝐴 = −1 𝐵=1
−1 1
H(𝑗Ω) = +
𝑗Ω + 4 𝑗Ω + 2
Applying Inverse Fourier Transform
𝒉(𝒕) = −𝒆−𝟒𝒕𝒖(𝒕) + 𝒆−𝟐𝒕𝒖(𝒕)
(ii) 𝑥(𝑡) = 𝑡𝑒−2𝑡𝑢(𝑡)

1
𝑋(𝑗Ω) =
(𝑗Ω + 2)2
Y(𝑗Ω) 2
=
X(𝑗Ω) [(𝑗𝛺)2 + 6𝑗Ω + 8]
2 1 2
∴ Y(𝑗Ω) = . =
[(𝑗𝛺)2 + 6𝑗Ω + 8] (𝑗Ω + 2)2 (𝑗Ω + 4)(𝑗Ω + 2)3
𝐴 𝐵 𝐶 𝐷
= + + +
𝑗Ω + 4 𝑗Ω + 2 (𝑗Ω + 2) 2 (𝑗Ω + 2)3
6 207

EnggTree.com
Faculty: Mrs.M.Shakunthala & Dr.M.N.Vimal Kumar
at 𝑗Ω = −4 𝑗𝛺 = −2
2 2
𝐴= (𝑗Ω + 4) 𝑑2 (𝑗Ω + 2)3
1 (𝑗Ω + 4)(𝑗Ω + 2)3
(𝑗Ω + 4)(𝑗Ω + 2)3 𝑗Ω=−4 𝐵=
1 2! 𝑑(𝑗Ω)2
𝐴=− 𝑗𝛺 =−2
4 1
𝐵=
4
𝑗𝛺 = −2 𝑗𝛺 = −2
2 𝐷= 2
𝑑 (𝑗Ω + 2)3 (𝑗Ω + 4)(𝑗Ω + 2)3 (𝑗Ω + 2)3
(𝑗Ω + 4)(𝑗Ω + 2)3
𝐶= 𝑗𝛺 =−2
𝑑(𝑗Ω)
𝐷=1
𝑗𝛺 =−2
1
𝐶=−
2
1 1 1
−4 4 −2 1
∴ Y(𝑗Ω) = + + +
𝑗Ω + 4 𝑗Ω + 2 (𝑗Ω + 2)2 (𝑗Ω + 2)3
Applying Inverse Fourier Transform
𝟏 𝟏 𝟏 𝟏
𝒚(𝒕) = − 𝒆−𝟒𝒕 𝒖(𝒕) + 𝒆−𝟐𝒕 𝒖(𝒕) − 𝒆−𝟐𝒕 𝒕𝒖(𝒕) + 𝒆 𝒕 𝒖(𝒕)
−𝟐𝒕 𝟐
𝟒 𝟒 𝟐 𝟐
Example 3.6: Find the direct form II structure of

5𝑆3 − 4𝑆2 + 11𝑆 − 2
𝐻(𝑆) = 1 1
(𝑆 − 4) (𝑆2 − 𝑆 + 2)
Solution: www.EnggTree.com
5𝑆3 − 4𝑆2 + 11𝑆 − 2 = 5𝑆3 − 4𝑆2 + 11𝑆 − 2
𝐻(𝑆) = 1 𝑆2 𝑆 𝑆 1
1
(𝑆 − 4) (𝑆2 − 𝑆 + 2) 𝑆3 − 4 − 𝑆2 + 4 + 2 − 8
4 11 2
5𝑆3 − 4𝑆2 + 11𝑆 − 2 5−𝑆+ 2 − 3
= 𝑆 𝑆
𝐻(𝑆) = 5𝑆2 3𝑆 1 5 3 1
𝑆3 − 4 + 4 − 8 1 − 4𝑆 + 2 − 3
4𝑆 8𝑆
Direct form II structure

X(S) W(S) 5 Y(S)
1/S
5/4 -4
1/S
-3/4 11
1/8 -2
Fig 3.39
Example 3.7: Realize the system with following differential equation in direct form I
𝑑 3 𝑦(𝑡) 𝑑 2 𝑦(𝑡) 𝑑𝑦 (𝑡) 𝑑 2 𝑥(𝑡) 𝑑𝑥 (𝑡)
𝑑𝑡 3
+3 𝑑𝑡 2
+5 𝑑𝑡
+ 7𝑦(𝑡) = 2 𝑑𝑡 2
+ 0.4 𝑑𝑡
+ 0.5𝑥(𝑡)
208

EnggTree.com
Solution:
𝑑 3 𝑦(𝑡) 𝑑 2 𝑦(𝑡) 𝑑𝑦 (𝑡) 𝑑 2 𝑥(𝑡) 𝑑𝑥 (𝑡)
+3 +5 + 7𝑦(𝑡) = 2 + 0.4 + 0.5𝑥(𝑡)
𝑑𝑡 3 𝑑𝑡 2 𝑑𝑡 𝑑𝑡 2 𝑑𝑡
Taking Laplace transform
𝑆3𝑌(𝑆) + 3𝑆2𝑌(𝑆) + 5𝑆𝑌(𝑆) + 7𝑌(𝑆) = 2𝑆2𝑋(𝑆) + 0.4𝑆𝑋(𝑆) + 0.5𝑋(𝑆)
Dividing both the side by 𝑆3
3 5 7 2 0.4 0.5
𝑌(𝑆) + 𝑌(𝑆) + 2 𝑌(𝑆) + 3 𝑌(𝑆) = 𝑋(𝑆) + 2 𝑋(𝑆) + 3 𝑋(𝑆)
𝑆 𝑆 𝑆 𝑆 𝑆 𝑆
2 0.4 0.5 3 5 7
𝑌(𝑆) = 𝑋(𝑆) + 2 𝑋(𝑆) + 3 𝑋(𝑆) − 𝑌(𝑆) − 2 𝑌(𝑆) − 3 𝑌(𝑆)
𝑆 𝑆 𝑆 𝑆 𝑆 𝑆
Direct form I structure
X(S) W(S) Y(S)
1/S 1/S
2 -3
1/S 1/S
-5
0.4
1/S 1/S
0.5 -7
Fig 3.42
Example 3.8:
www.EnggTree.com
Realize the system with transfer function in cascade form
4(𝑆2 + 4𝑆 + 3)
𝐻(𝑆) =
𝑆3 + 6.5𝑆2 + 11𝑆 + 4
Solution:
4(𝑆2 + 4𝑆 + 3) 4(𝑆 + 1)(𝑆 + 3)
𝑆+1 𝑆+3 4
𝐻(𝑆) = = . = .
𝑆3 + 6.5𝑆2 + 11𝑆 + 4 (𝑆 + 0.5)(𝑆 + 2)(𝑆 + 4) 𝑆 + 0.5 𝑆 + 2 𝑆 + 4
4 𝑆+1 𝑆+3
𝐻1(𝑆)𝐻2(𝑆)𝐻3(𝑆) = . .
𝑆 + 0.5 𝑆 + 2 𝑆 + 4
4 4/ 𝑆 + 1 1 + 1/𝑆 𝑆 + 3 1 + 3/𝑆
𝑆
𝐻1(𝑆) = = 𝐻2(𝑆) = = 𝐻3(𝑆) = =
𝑆 + 0.5 1 + 0.5/𝑆 𝑆 + 2 1 + 2/ 𝑆 + 4 1 + 4/𝑆
𝑌1(𝑆) 4 𝑌2(𝑆) 𝑆 𝑌3(𝑆)
= = 1 + 1/𝑆 = 1 + 3/𝑆
𝑊1(𝑆) 𝑆 𝑊 2(𝑆) 𝑊 3(𝑆)
𝑊1(𝑆) 1 𝑊2(𝑆) 1 𝑊3(𝑆) 1
= = =
𝑋1(𝑆) 1 + 0.5 𝑋2(𝑆) 1 + 2/ 𝑋3(𝑆) 1 + 4/
𝑆 𝑆 𝑆
X2(S) W2(S) Y2(S) Y3(S)
W1(S) X3(S) W3(S)
X1(S)
1/S 1/S 1/S

-0.5 4 Y1(S) -2 1 -4 3
Fig 3.54 Fig 3.55 Fig 3.56
209
8
EnggTree.com
Cascade form:
X(S)
1/S
Y(S)
-0.5 4
1/S
1/S
1 -4 3
-2
Fig 3.57
Example 3.9: Realize the following system in parallel form

𝑆(𝑆 + 2)
𝐻(𝑆) =
(𝑆 + 1)(𝑆 + 3)(𝑆 + 4)
Solution:
𝑆(𝑆 + 2) 𝐴 𝐵 𝐶
𝐻(𝑆) = = + +
(𝑆 + 1)(𝑆 + 3)(𝑆 + 4) 𝑆 + 1 𝑆 + 3 𝑆 + 4
𝑆(𝑆 + 2) = 𝐴(𝑆 + 3)(𝑆 + 4) + 𝐵(𝑆 + 1)(𝑆 + 4) + 𝐶(𝑆 + 1)(𝑆 + 3)
Let S= −1 Let S= −3 Let S= −4
−1(1) = 𝐴(2)(3) −3(−1) = 𝐵(−2)(1) −4(−2) = 𝐶(−3)(−1)
𝟏 𝟑 𝟖
𝑨=− 𝑩=− 𝑪=
𝟔 𝟐 𝟑
1 3 8
−6 −2
∴ 𝐻(𝑆) = + + 3
𝑆+1 𝑆+3 𝑆+4
Parallel form structure
X (S)
www.EnggTree.com
1/S
-1 -1/6
1/S
-3/2
-3
1/S
8/3 Y (S)
-4
210

EnggTree.com
Unit 4: Analysis of Discrete Time Signals

Sampling of CT signals
Sampling theorem (or) uniform sampling theorem (or) Low pass sampling theorem
It is one of useful theorem that applies to digital communication systems.
Sampling theorem states that “A band limited signal 𝒙(𝒕) with 𝑿(𝑚) = 𝟎 for |𝑚| ≥ 𝑚𝒎 can
be represented into and uniquely determined from its samples 𝒙(𝒏𝑻) if the sampling frequency
𝒇𝒔 ≥ 𝟐𝒇𝒎, where 𝒇𝒎 is the frequency component present in it”.
(i.e) for signal recovery, the sampling frequency must be at least twice the highest
frequency present in the signal.
Proof:
Sampling Operation:
𝝳T(t) is impulse train
𝝳T(t)
xs(t)=x(nT)
x(t)
-5T -4T -3T -2T -T 0 T 2T 3T 4T 5T

0 nT
t
t
www.EnggTree.com
Fig 4.2
x(t) xs(t)=x(nT)
Fig 4.1 Fig 4.4

𝝳T(t)
Fig 4.3
Analog signal 𝑥(𝑡) is input signal as shown in Fig 4.1, 𝛿𝑇(𝑡) is the train of impulse shown in Fig 4.2
Sampled signal 𝑥𝑠(𝑡) is the product of signal 𝑥(𝑡) and impulse train 𝛿𝑇 (𝑡) as shown in Fig 4.2
∴ 𝑥𝑠(𝑡) = 𝑥(𝑡). 𝛿𝑇(𝑡)
∞ ∞
1
𝑤𝑒 𝑘𝑛𝑜𝑤 𝛿𝑇 (𝑡) = ∑ 𝛿(𝑡 − 𝑛𝑇) = ∑ 𝑒𝑗𝑛 𝜔𝑠𝑡
𝑇
𝑛=−∞ 𝑛=−∞
∞
∴ 𝑥 (𝑡) = 𝑥(𝑡). 1 ∑ 𝑒𝑗𝑛 𝜔𝑠𝑡
𝑠 𝑇
𝑛=−∞
Applying Fourier transform on both sides
∞
𝑋 (𝜔) = 1 ∑ 𝐹[𝑥(𝑡)𝑒𝑗𝑛 𝜔𝑠 𝑡 ]
𝑠
𝑇
𝑛=−∞
∞
1
𝑋𝑠(𝜔) = ∑ 𝑋(𝜔 − 𝑛𝜔𝑠)
𝑇
𝑛=−∞
2𝜋
𝑤𝑕𝑒𝑟𝑒 𝜔𝑠 = 2𝜋𝑓𝑠 =
𝑇

EnggTree.com
∞
1 2𝜋𝑛
∴ 𝑋𝑠(𝜔) = ∑ 𝑋(𝜔 − )
𝑇 𝑇
𝑛=−∞
(𝑜𝑟)
∞
1
𝑋𝑠(𝑓) = 𝑓𝑠 ∑ 𝑋(𝑓 − 𝑛𝑓𝑠 ) 𝑤𝑕𝑒𝑟𝑒 𝑓𝑠 =
𝑛=−∞
𝑇
Where 𝑋(𝜔) 𝑜𝑟 𝑋(𝑓) 𝑖𝑠 𝑆𝑝𝑒𝑐𝑡𝑟𝑢𝑚 𝑜𝑓 𝑖𝑛𝑝𝑢𝑡 𝑠𝑖𝑔𝑛𝑎𝑙.
Where 𝑋𝑠(𝜔) 𝑜𝑟 𝑋𝑠(𝑓) 𝑖𝑠 𝑆𝑝𝑒𝑐𝑡𝑢𝑟𝑚 𝑜𝑓 𝑠𝑎𝑚𝑝𝑙𝑒𝑑 𝑠𝑖𝑔𝑛𝑎𝑙.
Spectrum of continuous time signal x(t) with maximum frequency 𝜔𝑚 is shown in Fig 4.5.
X(Ѡ)
-Ѡm 0 Ѡm Ѡ
Fig 4.5 Spectrum of x(t)
Xs(Ѡ)
1/T
-Ѡs-Ѡm -Ѡs -Ѡs+Ѡm -Ѡm 0 Ѡm Ѡs-Ѡm Ѡs Ѡs+Ѡm Ѡ
Fig 4.6 Spectrum of 𝑥𝑠(𝑡) when 𝑚𝒔 − 𝑚𝒎 > 𝑚𝒎

www.EnggTree.com
X (Ѡ) s
1/T
-Ѡs-Ѡm -Ѡs -Ѡm 0 Ѡm Ѡs Ѡs+Ѡm Ѡ

-Ѡs+Ѡm Ѡs-Ѡm
Fig 4.7 Spectrum of 𝑥𝑠(𝑡) when 𝑚𝒔 − 𝑚𝒎 = 𝑚𝒎

Xs(Ѡ)
1/T
-Ѡs-Ѡm -Ѡs -Ѡm 0 Ѡm Ѡs Ѡs+Ѡm Ѡ

-Ѡs+Ѡm Ѡs-Ѡm
Fig 4.8 Spectrum of 𝑥𝑠(𝑡) when 𝑚𝒔 − 𝑚𝒎 < 𝑚𝒎
From the plot of 𝑋𝑠(𝜔) (Fig 4.6, Fig 4.7, Fig 4.8),
For 𝑚𝒔 > 2𝑚𝒎

The spectral replicates have a larger separation between them known as guard band which
makes process of filtering much easier and effective. Even a non-ideal filter which does not have a
sharp cut off can also be used.
281

EnggTree.com
For 𝑚𝒔 = 𝟐𝑚𝒎
There is no separation between the spectral replicates so no guard band exists and 𝑋(𝜔)
can be obtained from 𝑋𝑠(𝜔) by using only an ideal low pass filter (LPF) with sharp cutoff.
For 𝑚𝒔 < 2𝑚𝒎
The low frequency component in 𝑋𝑠(𝜔) overlap on high frequency components of 𝑋(𝜔) so
that there is presence of distortion and 𝑋(𝜔) cannot be recovered from 𝑋𝑠(𝜔) by using any
filter. This distortion is called aliasing.
So we can conclude that the frequency spectrum of 𝑋𝑠(𝜔) is not overlapped for
𝑚𝒔 − 𝑚𝒎 ≥ 𝑚𝒎, therefore the Original signal can be recovered from the sampled signal.
For 𝑚𝒔 − 𝑚𝒎 < 𝑚𝒎, the frequency spectrum will overlap and hence the original signal cannot be
recovered from the sampled signal.
∴ For signal recovery,

𝜔𝑠 − 𝜔𝑚 ≥ 𝜔𝑚 (𝑖. 𝑒) 𝑚𝒔 ≥ 𝟐𝑚𝒎
(𝑜𝑟)
𝒇𝒔 ≥ 𝟐𝒇𝒎
i.e., Aliasing can be avoided if 𝑓𝑠 ≥ 2𝑓𝑚
Aliasing effect (or) fold over effect
It is defined as the phenomenon in which a high frequency component in the frequency

spectrum of signal takes identity of a lower frequency component in the spectrum of the sampled
signal.
www.EnggTree.com
When 𝑓𝑠 <1 2𝑓𝑚 , (i.e) when signal is under sampled, the individual terms in equation
𝑋 (ω) = ∑∞ 𝑥(𝜔 − 𝑛𝜔 ) get overlap. This process of spectral overlap is called frequency
𝑠 T n=−∞ 𝑠
folding effect.
Occurrence of aliasing
Aliasing Occurs if
i) The signal is not band-Limited to a finite range.
ii) The sampling rate is too low.
To Avoid Aliasing
i) 𝑥(𝑡) should be strictly band limited.
It can be ensured by using anti-aliasing filter before the sampler.
ii) 𝑓𝑠 should be greater than 2𝑓𝑚 .
Nyquist rate
It is the theoretical minimum sampling rate at which a signal can be sampled and still be
reconstructed from its samples without any distortion
𝑁𝑦𝑞𝑢𝑖𝑠𝑡 𝑟𝑎𝑡𝑒 𝑓𝑁 = 2𝑓𝑚 . 𝐻𝑧

EnggTree.com
Data Reconstruction or Interpolation
The process of obtaining analog signal 𝑥(𝑡) from the sampled signal 𝑥𝑠(𝑡) is called data
reconstruction or interpolation.
∞
𝑤𝑒 𝑘𝑛𝑜𝑤 𝑥𝑠(𝑡) = 𝑥(𝑡). 𝛿𝑇 (𝑡) = 𝑥(𝑡) ∑ 𝛿(𝑡 − 𝑛𝑇)

𝑛=−∞
𝛿(𝑡 − 𝑛𝑇) 𝑒𝑥𝑖𝑠𝑡 𝑜𝑛𝑙𝑦 𝑎𝑡 𝑡 = 𝑛𝑇
∞
∴ 𝑥𝑠(𝑡) = 𝑥(𝑛𝑡) ∑ 𝛿(𝑡 − 𝑛𝑇)

𝑛=−∞
The reconstruction filter, which is assumed to be linear and time invariant, has unit impulse
response 𝑕(𝑡).
The reconstruction filter, output 𝑦(𝑡) is given by convolution of 𝑥𝑠(𝑡) and 𝑕(𝑡).
∞ ∞
∴ 𝑦(𝑡) = 𝑥𝑠(𝑡) ∗ 𝑕(𝑡) = ∫ 𝑥(𝑛𝑇) ∑ 𝛿(𝑟 − 𝑛𝑇). 𝑕(𝑡 − 𝑟)𝑑𝑟
−∞ 𝑛=−∞
∞ ∞
= ∑ 𝑥(𝑛𝑇) ∫ 𝛿(𝑟 − 𝑛𝑇) 𝑕(𝑡 − 𝑟)𝑑𝑟

𝑛=−∞ −∞
𝛿(𝑟 − 𝑛𝑇)𝑒𝑥𝑖𝑠𝑡 𝑜𝑛𝑙𝑦 𝑎𝑡 𝑟 = 𝑛𝑇
𝛿(𝑟 − 𝑛𝑇) = 1 𝑎𝑡 𝑟 = 𝑛𝑇
∞
∴ 𝑦(𝑡) = ∑ 𝑥(𝑛𝑇) 𝑕(𝑡 − 𝑛𝑇)

www.EnggTree.com
𝑛=−∞
Ideal Reconstruction filter
The sampled signal 𝑥𝑠(𝑡) is passed through an ideal LPF (Fig 4.9) with bandwidth greater
than 𝑓𝑚 and a pass band amplitude response of T, then the filter output is 𝑥(𝑡).
Transfer function of ideal reconstruction filter is
𝐻(𝑓) = 𝑇 ; |𝑓| < 0.5𝑓𝑠
Ideal Reconstruction filter

H(f)
Xs(t) T x(t)
-fs/2 fs/2 f
Fig 4.9
The impulse response of ideal reconstruction filter is
𝑓𝑠 𝑓𝑠
𝑓𝑠
2 2 𝑇 𝑗 2𝜋𝑓𝑠 𝑡 −𝑗 2𝜋𝑓𝑠𝑡
𝑒𝑗 2𝜋𝑓𝑡 2
𝑕(𝑡) = ∫ 𝑇 𝑒𝑗𝜔𝑡 𝑑𝑓 = ∫ 𝑇 𝑒𝑗2𝜋𝑓𝑡 𝑑𝑓 = 𝑇 [ 𝑗2𝜋𝑡 ] 𝑓 = 𝑗2𝜋𝑡 [𝑒 2 −𝑒 2 ]
𝑠
𝑓 𝑓 −2
− 2𝑠 − 2𝑠
𝑗 2𝜋𝑓𝑠 𝑡 −𝑗 2𝜋 𝑡 𝑓𝑠
1 𝑒 2 −𝑒 2 1
= [ ]= sin 𝜋𝑓𝑠𝑡 = sin 𝑐 𝜋𝑓𝑠𝑡
𝑓𝑠𝜋𝑡 2𝑗 𝜋𝑓𝑠𝑡
∴ 𝑕(𝑡 − 𝑛𝑇) = sin 𝑐 𝜋𝑓𝑠 (𝑡 − 𝑛𝑇) … … … (1)
∞
𝑦(𝑡) = ∑ 𝑥(𝑛𝑇)𝑕(𝑡 − 𝑛𝑇)

𝑛=−∞
283

EnggTree.com
Substitute equation (1) in above equation

∞ ∞
𝑡
∴ 𝑦(𝑡) = ∑ 𝑥(𝑛𝑇) sin 𝑐 𝜋𝑓𝑠 (𝑡 − 𝑛𝑇) = ∑ 𝑥(𝑛𝑇) sin 𝑐 𝜋 ( − 𝑛)
𝑇
𝑛=−∞ 𝑛=−∞
1
[∵ 𝑓𝑠 = ]
𝑇
Example 4.1: Determine Nyquist rate and Nyquist interval corresponding to each of the following
signals
𝑥(𝑡) = cos 1000𝜋𝑡 + cos 3000𝜋𝑡 + sin 4000𝜋𝑡
𝜔𝑚 = 4000𝜋 ⇒ 2𝜋𝑓𝑚 = 4000𝜋 ⇒ 𝑓𝑚 = 2000𝐻𝑧
𝑁𝑦𝑞𝑢𝑖𝑠𝑡 𝑟𝑎𝑡𝑒 = 2𝑓𝑚 = 4000 𝐻𝑧
1 1
𝑁𝑦𝑞𝑢𝑖𝑠𝑡 𝑖𝑛𝑡𝑒𝑟𝑣𝑎𝑙 = = = 0.25 𝑚𝑆
2𝑓𝑚 4000
Discrete time Fourier Transform

∞
𝐹[𝑥(𝑛)] = 𝑋(𝑒𝑗𝜔 ) = ∑ 𝑥(𝑛)𝑒−𝑗𝜔𝑛

𝑛=−∞
4.2.2 Inverse Discrete Time Fourier Transform
www.EnggTree.com
1 𝜋
𝑥(𝑛) = 𝐹−1[𝑋(𝑒𝑗𝜔 )] = ∫ 𝑋(𝑒𝑗𝜔 ) 𝑒𝑗𝜔𝑛 𝑑𝜔 , 𝑓𝑜𝑟 𝑛 = −∞ 𝑡𝑜 ∞
2𝜋
−𝜋
Example 4.9: Find Fourier transform of the following
i) 𝑥(𝑛) = 𝛿(𝑛)
ii) 𝑥(𝑛) = 𝑢(𝑛)
iii) 𝑥(𝑛) = 𝑎𝑛 𝑢(𝑛)
Solution:
i) 𝑥(𝑛) = 𝛿(𝑛)
∞ 0, 𝑛≠0
𝛿(𝑛) = {
𝑋(𝑒𝑗𝜔 ) = 𝐹{𝛿(𝑛)} = ∑ 𝛿(𝑛) 𝑒−𝑗𝜔𝑛 = 𝑒0 = 1 1, 𝑛=0
𝑛=−∞

∞ ∞
𝑋(𝑒𝑗𝜔 ) = 𝐹{𝑢(𝑛)} = ∑ 𝑢(𝑛) 𝑒−𝑗𝜔𝑛 = ∑ 𝑒−𝑗𝜔𝑛 𝑢(𝑛) = { 0,

𝑛<0
𝑛=−∞ 𝑛=0 1,
𝑛≥0
1 1
= 1 + 𝑒−𝑗𝜔 + 𝑒−2𝑗𝜔 + ⋯ ∞ = 1+𝑥+𝑥 +⋯ =
2
1 − 𝑒−𝑗𝜔 1−𝑥
iii) 𝑥(𝑛) = 𝑎𝑛 𝑢(𝑛)

It is Right handed exponential signal

EnggTree.com
∞ ∞
𝑋(𝑒𝑗𝜔 )= 𝐹{𝑎𝑛 𝑢(𝑛)} = ∑ 𝑎𝑛 𝑢(𝑛)𝑒−𝑗𝜔𝑛 = ∑ 𝑎𝑛 𝑒−𝑗𝜔𝑛

𝑛=−∞ 𝑛=0
= 1 + 𝑎𝑒−𝑗𝜔 + (𝑎𝑒−𝑗𝜔 2 + ⋯+∞ = 1
)
1 − 𝑎𝑒−𝑗𝜔
Example 4.11: Obtain DTFT of rectangular pulse
𝐴, 0≤𝑛 ≤𝐿−1
𝑥(𝑛) = {
0, 𝑜𝑡𝑕𝑒𝑟𝑤𝑖𝑠𝑒
Solution:
𝐿−1 𝑁−1
1 − 𝑒−𝑗𝜔𝐿 1 − 𝑥𝑁
𝑋(𝑒𝑗𝜔 ) = ∑ 𝐴𝑒−𝑗𝜔𝑛 = 𝐴[ ] ∑ 𝑥𝑛 =
1 − 𝑒−𝑗𝜔 1−𝑥
𝑛=0 𝑛=0
𝑗𝜔𝐿 −𝑗𝜔𝐿 −𝑗𝜔𝐿 𝜔𝐿 𝜔𝐿
(𝑒 2 −𝑒 2 )𝑒 2 2𝑗 sin 𝑗𝜔 (𝐿−1) 𝑗𝜔 (𝐿−1) sin
𝑋(𝑒𝑗𝜔 ) = 𝐴[ 𝑗𝜔
] = 𝐴[ 2
𝜔 ] 𝑒− 2 = 𝐴𝑒− 2 [ 2
𝜔 ]
− 𝑗𝜔 −𝑗𝜔
2𝑗 sin sin
(𝑒 2 − 𝑒 2 )𝑒 2 2 2
Example 4.16: Find DTFT of 𝑥(𝑛) = sin(n θ)u(n)

Solution:
𝑥(𝑛) = sin(n θ)u(n)
∞
∞ ∞ ∞
ej𝜃n − e−j𝜃n −j𝜔𝑛 1
𝑋(𝑒𝑗𝜔 ) = ∑ sin(n θ)𝑒−j𝜔𝑛 = ∑( )𝑒 = (∑ ej(𝜃−𝜔)n − ∑ e−j(𝜃+𝜔)n )
2j 2𝑗
𝑛=0 www.EnggTree.com
𝑛=0
𝑛=0 𝑛=0
1 1 1 1 1 − e−j(𝜃+𝜔) − 1 + ej(𝜃−𝜔)
= ( − ) = ( )
2𝑗 1 − ej(𝜃−𝜔) 1 − e−j(𝜃+𝜔) 2𝑗 1 − 2e−j𝜔 cos θ + 𝑒−2𝑗𝜔
1 2je−j𝜔 sin θ e−j𝜔 sin θ
= ( ) =
2𝑗 1 − 2e−j𝜔 cos θ + 𝑒−2𝑗𝜔 1 − 2e−j𝜔 cos θ + 𝑒−2𝑗𝜔
4.4 Z-Transform
The Z-transform of discrete time signal 𝑥(𝑛) is defined as

∞
𝑍[𝑥(𝑛)] = 𝑋(𝑍) = ∑ 𝑥(𝑛)𝑍−𝑛

𝑛=−∞
4.4.2 Inverse Z-transform
The inverse Z-transform of 𝑋(𝑍) is defined as

1
𝑥(𝑛) = 𝑍−1 𝑋(𝑍) = 𝑋(𝑍) 𝑍𝑛−1𝑑𝑍
2𝜋𝑗
𝐶
Example 4.21: Find Z-transform of the following

i) 𝑥(𝑛) = 𝛿(𝑛)
iii) 𝑥(𝑛) = −𝑎𝑛 𝑢(−𝑛 − 1) 𝑎𝑛𝑑 𝑓𝑖𝑛𝑑 𝑅𝑂𝐶
285

EnggTree.com
i) 𝑥(𝑛) = 𝛿(𝑛)
1 𝑓𝑜𝑟 𝑛 = 0
𝛿(𝑛) = {
0 𝑓𝑜𝑟 𝑛 ≠ 0
∞
𝑍[𝛿(𝑛)] = ∑ δ(n)𝑧−n = Z−0 = 1

𝑛=−∞
∴ 𝑍[𝛿(𝑛)] = 1

∞ ∞
𝑍[𝑢(𝑛)] = ∑ u(n)𝑧−𝑛 = ∑ 𝑧−𝑛 = 1 + 𝑧−1 + 𝑧−2 + 𝑧−3 1
1 + 1 + ⋯⋯
+ ⋯ ⋯ = 1 + + 2 𝑧3
𝑛=−∞ 𝑛=0 𝑧 𝑧
1 1 𝑍
= 1
= =
1− 1− 𝑧−1 𝑍−1
𝑧
The above series convergence if |𝑍−1| < 1 i.e ROC is |𝑍| > 1
iii) 𝑥(𝑛) = −𝑎𝑛 𝑢(−𝑛 − 1)

∞ ∞ −1
𝑋(𝑍) = ∑ 𝑥(𝑛)𝑧−𝑛 = ∑ −𝑎𝑛 𝑢(−𝑛 − 1)𝑧−𝑛 = − ∑ 𝑎𝑛 𝑧−𝑛

𝑛=∞ 𝑛=−∞ 𝑛=−∞
∞
𝑋(𝑍) = − ∑ 𝑎−𝑛 𝑧𝑛 = −[𝑎−1𝑍 + (𝑎−1𝑍)2 + (𝑎−1𝑍)3 + ⋯ ] = −𝑎−1𝑍[1 + 𝑎−1𝑍 + (𝑎−1𝑍)2 + ⋯ ]

𝑛=1
𝑎−1𝑧 −𝑎−1 𝑧 1 𝑧
−1] = −1 =
= −[ −1 ] =
1−𝑎 𝑧
−1 [
www.EnggTree.com
−𝑎 𝑧 1 − 𝑎𝑧 1 − 𝑎𝑧 𝑧−𝑎
ROC: |𝑎−1𝑍| < 1 ⇒ |𝑍| < |𝑎|

Im(z)
Re(z)
xample 4.25: Obtain Inverse Z-Transform of

1
𝑋(𝑧) = 𝑓𝑜𝑟 𝑖) |𝑍| > 0.4 𝑖𝑖) |𝑍| < 0.2 𝑖𝑖𝑖)0.2 < |𝑍| < 0.4
1 − 0.6𝑍−1 + 0.08𝑍−2
Solution:
1 𝑍2
𝑋(𝑧) = =
1 − 0.6𝑍−1 + 0.08𝑍−2 𝑍2 − 0.6𝑍 + 0.08
𝑋(𝑍) 𝑍 𝐴 𝐵
= = +
𝑍 (𝑍 − 0.2)(𝑍 − 0.4) 𝑍 − 0.2 𝑍 − 0.4
𝑍 𝑍
𝐴= (𝑍 − 0.2)| = −1 (𝑍 − 0.4)|
𝐵= =2
(𝑍 − 0.2)(𝑍 − 0.4) 𝑍=0.2
(𝑍 − 0.2)(𝑍 − 0.4) 𝑍=0.4
𝑋(𝑍) −1 2 −𝑍 2𝑍
∴ = + ⇒ 𝑋(𝑍) = +
𝑍 𝑍 − 0.2 𝑍 − 0.4 𝑍 − 0.2 𝑍 − 0.4

EnggTree.com
Applying inverse Z-transform

𝑥(𝑛) = −(0.2)𝑛 𝑢(𝑛) + 2(0.4)𝑛 𝑢(𝑛)
Im (z)
ROC: |𝑍| > 0.4
ROC lies outside of all poles. So both the terms
are causal 0.2 0.4 Re(z)
∴ 𝑥(𝑛) = −(0.2)𝑛 𝑢(𝑛) + 2(0.4)𝑛 𝑢(𝑛)
Im(z)
ROC : |𝑍| < 0.2

ROC lies inside of all poles. So both the terms are
non-causal 0.2 0.4Re(z)
∴ 𝑥(𝑛) = (0.2)𝑛 𝑢(−𝑛 − 1) − 2(0.4)𝑛 𝑢(−𝑛 − 1)
ROC :
0.5 < |𝑧| < 1 Im(z)
ROC lies inside of pole Z=0.4 and lies outside of
pole Z=0.2. So the term with pole Z=0.4 is non- Re(z)
causal and the term with pole Z=0.2 is causal 0.2 0.4
∴ 𝑥(𝑛) = −(0.2)𝑛 𝑢(𝑛) − 2(0.4)𝑛 𝑢(−𝑛 − 1)
www.EnggTree.com
𝑍3
Example 4.29: Find the inverse Z transform of 𝑋(𝑍) = using Cauchy residue method.
(𝑍+1)(𝑍−1)2
Solution:
𝑍3
𝑋(𝑍) =
(𝑍 + 1)(𝑍 − 1)2
𝑥(𝑛) = 𝑅𝑒𝑠𝑖𝑑𝑢𝑒 𝑜𝑓 𝑋(𝑍) 𝑍𝑛−1𝑎𝑡 𝑝𝑜𝑙𝑒 (𝑍 = −1)
+ 𝑅𝑒𝑠𝑖𝑑𝑢𝑒 𝑜𝑓𝑋(𝑍) 𝑍𝑛−1𝑎𝑡 𝑝𝑜𝑙𝑒 (𝑍 = 1) 𝑤𝑖𝑡𝑕 𝑚𝑢𝑙𝑡𝑖𝑝𝑙𝑖𝑐𝑖𝑡𝑦 2
𝑍3(𝑍 + 1) 𝑑 𝑍3(𝑍 − 1)2
𝑥(𝑛) = 𝑍𝑛−1| + [ ] 𝑍𝑛−1|
(𝑍 + 1)(𝑍 − 1)2 𝑍=−1
𝑑𝑍 (𝑍 + 1)(𝑍 − 1)2
𝑍=1
1 2(𝑛 + 2) − 1 1 2𝑛 + 3 𝑢(𝑛)
= (− ) (−1)𝑛−1 + = (− ) (−1)𝑛−1 𝑢(𝑛) +
4 4 4 4
𝑥(0) = 1 𝑥(1) = 1 𝑥(2) = 2 𝑥(3) = 2

∴ 𝑥(𝑛) = {1,1,2,2, … }
287

EnggTree.com
UNIT 5: Linear Time Invariant –Discrete

Time Systems
Linear Time Invariant Discrete Time System (LTI-DT System)
When a discrete time system satisfies the properties of linearity and time invariance,
then it is called an LTI System.
Discrete time system
A discrete time system is a device that operates on a discrete time signal to produce
another discrete time signal called the output or response of the system.
Discrete time system
input output
𝑥(𝑛) 𝑦(𝑛)
H
The input signal x(n) is transformed to output signal y(n) through the above system
Impulse Response
When the input to a discrete time system is a unit impulse 𝛿(𝑛) then the output is called
an impulse response of the system and is denoted by ℎ(𝑛)
∴Impulse response ℎ(𝑛) = Η{𝛿(𝑛)}
𝛿(𝑛) → 𝐻 → ℎ(𝑛)
Impulse response of interconnected systems

www.EnggTree.com
Parallel connections of discrete time systems (Distributive property)
Consider two LTI systems with impulse response h1(n) and h2(n) connected in parallel as
shown in Fig 5.1
h1(n) y1(n)
y(n) x(n) y(n)
x(n) System 1
h(n)=h1(n)+h2(n)
h2(n)
y2(n)
System 2
Fig 5.1 Parallel connections of discrete time systems
Cascade connection of discrete time systems (Associative property)

𝑥(𝑛) 𝑦1(𝑛) 𝑦(𝑛) 𝑥(𝑛)
ℎ1(𝑛) ℎ2(𝑛) ℎ(𝑛) = ℎ1(𝑛) ∗ ℎ2(𝑛) 𝑦(𝑛)
System1 System2
Fig 5.2
Let us consider two systems with impulse ℎ1(𝑛) and ℎ2(𝑛) connected in cascade as shown in
Fig 5.2
348
EnggTree.com
Block diagram representation (System connected in series/parallel)
System Realization
There are four types of system realization in discrete time linear time invariant systems.
They are
• Direct form I realization
• Direct form II realization
• Cascade form realization
• Parallel form realization
Direct form I realization
It is the direct implementation of transfer function describing the system. It uses
separate unit delay element for input and output variables. It provides direct relation between
time domain and Z-domain equations. This form is practical for small filters.
Advantages:
• Simplicity
• Most straight forward realization
Disadvantages:
• More number of unit delay elements are used
• Inefficient and impractical for complex design
Consider a system with system function

𝑌(𝑍) = 𝑏0 + 𝑏1𝑍−1 + 𝑏2𝑍−2
𝐻(𝑍) = 𝑋(𝑍) 1 + 𝑎 𝑍−1 + 𝑎 𝑍−2
1 2
www.EnggTree.com
𝑌(𝑍) + 𝑎1𝑍−1𝑌(𝑍) + 𝑎2𝑍−2𝑌(𝑍) = 𝑏0𝑋(𝑍) + 𝑏1𝑍−1𝑋(𝑍) + 𝑏2𝑍−2𝑋(𝑍)
Direct form – I realization of 𝐻(𝑍)
𝑋(𝑍) 𝑏0 𝑊(𝑍) 𝑌(𝑍)
𝑍 −1 𝑍− 1
𝑏1 −𝑎1
𝑍 −1 𝑍− 1
𝑏2 −𝑎2
Fig 5.3 Direct form – I realization
Direct form II realization

It is the direct implementation of transfer function describing the system. Instead of
using separate unit delay elements for input and output variables separately, an intermediate
variable is unit delay element. It provides direct relation between time domain and z-domain
equations.
Advantages:
It uses minimum number of unit delay element
Straight forward realization
Consider a system with system function

𝑌(𝑍) = 𝑏0 + 𝑏1𝑍−1 + 𝑏2𝑍−2
𝐻(𝑍) = 𝑋(𝑍) 1 + 𝑎 𝑍−1 + 𝑎 −2 … … … 〈4〉
1 2𝑍
349
EnggTree.com
𝑌(𝑍)
𝑌(𝑍) 𝑊(𝑍)
𝐿𝑒𝑡 . =
𝑋(𝑍) 𝑊(𝑍) 𝑋(𝑍)
𝑊(𝑍) 1
𝑊ℎ𝑒𝑟𝑒 𝑋(𝑍) = 1 + 𝑎 𝑍−1 + 𝑎 −2 … … … 〈5〉
1 2𝑍
𝑎𝑛𝑑 𝑌(𝑍) = 𝑏 + 𝑏 𝑍−1 + 𝑏 𝑍−2 … … … 〈6〉
𝑊(𝑍) 0 1 2
From eq 〈5〉 we have

𝑊(𝑍) = 𝑋(𝑍) − 𝑎1𝑍−1𝑊(𝑍) − 𝑎2𝑍−2𝑊(𝑍) … … … 〈7〉
From eq 〈6〉 we have
𝑌(𝑍) = 𝑏0𝑊(𝑍) + 𝑏1𝑍−1𝑊(𝑍) + 𝑏2𝑍−2 𝑊(𝑍) … … … 〈8〉
Realization of eq 〈7〉 Realization of eq 〈8〉
X(Z) W(Z) b0
W(Z) Y(Z)
𝑍 −1 𝑍 −1
-a1 b1
𝑍− 1 𝑍 −1
-a2 b2
Fig 5.4 Fig 5.5
Combined form of Fig 5.4 and Fig 5.5 gives direct form II realization as shown in Fig 5.6
X(Z) b0 Y(Z)
−1
www.EnggTree.com
-a
𝑍
b 1 1
𝑍−1
-a2 b2
Fig 5.6 Direct form II realization
Cascade form(Series form)

In cascade form realization the given transfer function is expressed as a product of
several transfer function and each of these transfer function is realized in direct form II and
then all those realized structures are cascaded i.e., connected in series.
Consider a system with the following system function
(𝑏𝑘0 + 𝑏𝑘1𝑍−1 + 𝑏𝑘2𝑍−2)(𝑏𝑚0 + 𝑏𝑚1𝑍−1 + 𝑏𝑚2𝑍−2)
𝐻(𝑍) = (𝑍)𝐻2(𝑍)
(1 + 𝑎𝑘1 𝑍−1 + 𝑎𝑘2 𝑍−2)(1 + 𝑎𝑚1 𝑍−1 + 𝑎𝑚2 𝑍−2) = 𝐻1
Where
(𝑏𝑘0 + 𝑏𝑘1𝑍−1 + 𝑏𝑘2𝑍−2)
𝐻1(𝑍) = (1
+ 𝑎𝑘1 𝑍−1 + 𝑎𝑘2 𝑍−2)
(𝑏𝑚0 + 𝑏𝑚1𝑍−1 + 𝑏𝑚2𝑍−2)
𝐻2(𝑍) =
(1 + 𝑎𝑚1 𝑍−1 + 𝑎𝑚2 𝑍−2)
350
EnggTree.com
X(Z) bk0 bm0
Y(Z)
𝑍−1 𝑍−1
-ak1 bk1 -am1 bm1
𝑍−1 𝑍−1
-ak2 bk2 -am2 bm2
Fig 5.7 Cascade form realization
Realizing 𝐻1(𝑍) and 𝐻2(𝑍) in direct form II and the cascade form of the system function 𝐻(𝑍) is
shown in Fig 5.7
Parallel form realization

The given transfer function is expressed into its partial fractions and each factor is
realized in direct form II and all those realized structures are connected in parallel as shown in
Fig 5.8.
Consider a system with the following system function

𝑁
𝐶𝑘
𝐻(𝑍) = 𝑐 + ∑
1 − 𝑃𝑘 𝑍−1
𝑘=1
Where {𝑃𝑘} are poles of the system function
𝐶1 𝐶2 𝐶𝑁
𝐻(𝑍) = 𝑐 + + + ………+
1 − 𝑃1𝑍−1 1 − 𝑃2𝑍−1 1 − 𝑃𝑁 𝑍−1
X(Z) C
www.EnggTree.com
C 1
𝑍−1
P1
C2
P2 𝑍 −1
CN Y(Z)
PN 𝑍 −1
Fig 5.8 Parallel form realization
Solved Problems
Example 5.1: Determine the frequency response and impulse response
1 1
𝑦(𝑛) − 𝑦(𝑛 − 1) − 𝑦(𝑛 − 2) = 𝑥(𝑛)
6 6
Solution:
1 1
𝑦(𝑛) − 𝑦(𝑛 − 1) − 𝑦(𝑛 − 2) = 𝑥(𝑛)
6 6
351
EnggTree.com
Applying DTFT
1 1
𝑌(𝑒𝑗𝜔 ) − 𝑒 −𝑗𝜔 𝑌(𝑒 𝑗𝜔 ) − 𝑒 −2𝑗𝜔 𝑌(𝑒 𝑗𝜔 ) = 𝑋(𝑒 𝑗𝜔)
6 6
𝑌(𝑒𝑗𝜔) = 1 𝑒2𝑗𝜔
Frequency response 𝐻(𝑒𝑗𝜔) = 𝑋(𝑒𝑗𝜔) 1 −𝑗𝜔 1 −2𝑗𝜔 = 2𝑗𝜔
𝑒 − 1𝑒𝑗𝜔 − 1
1− 𝑒 − 𝑒 6 6
6 6
𝐻(𝑒𝑗𝜔) 𝑒𝑗𝜔 𝐴 𝐵
= 2𝑗𝜔 1 𝑗𝜔 1 = 𝑗𝜔 1 + 𝑗𝜔 1
𝑒𝑗𝜔 𝑒 − 6𝑒 − 6 𝑒 − 2 𝑒 + 3
1 1
𝑒𝑗𝜔 = 𝐴 (𝑒𝑗𝜔 + ) + 𝐵 (𝑒𝑗𝜔 − )
3 2
At 𝑒𝑗𝜔 = −1 At 𝑒𝑗𝜔 = 1
3 2
−1 = 𝐵(−1 − 1) , ∴ 𝐵 = 2 1 = 𝐴(1 + 1) , ∴ 𝐴 = 3
3 3 2 5 2 2 3 5
3 𝑗𝜔 2 𝑗𝜔
𝑒 𝑒
5 5
𝐻(𝑒𝑗𝜔) = +
𝑒𝑗𝜔 − 12 𝑒𝑗𝜔 + 31
Applying inverse DTFT
3 1 𝑛 2 1 𝑛
ℎ(𝑛) = ( ) 𝑢(𝑛) + (− ) 𝑢(𝑛)
5 2 5 3
Example 5.2: Find response of system using DTFT

1 𝑛 3 𝑛
ℎ(𝑛) = ( ) 𝑢(𝑛) ; 𝑥(𝑛) = ( ) 𝑢(𝑛)
2 4
Solution:
www.EnggTree.com
1 𝑛 3 𝑛
ℎ(𝑛) = ( ) 𝑢(𝑛) ; 𝑥(𝑛) = ( ) 𝑢(𝑛)
2 4
Applying DTFT
1 1
𝐻(𝑒𝑗𝜔) = ; 𝑋(𝑒𝑗𝜔) =
1 − 1𝑒 −𝑗𝜔 −𝑗𝜔
2 1 − 34 𝑒
𝑌(𝑒𝑗𝜔) = 𝐻(𝑒𝑗𝜔)𝑋(𝑒𝑗𝜔)
1 1 𝑒𝑗𝜔 𝑒𝑗𝜔
𝑌(𝑒𝑗𝜔) = 1 . 3 −𝑗𝜔 = 1 . 𝑗𝜔 3
1 − 2𝑒 1 − 4𝑒 𝑒𝑗𝜔 − 2 𝑒 − 4
𝑌(𝑒𝑗𝜔) −𝑗𝜔
𝑒𝑗𝜔 𝐴 𝐵
𝑗𝜔
= 1 3
= 𝑗𝜔 1 + 𝑗𝜔 3
𝑒 (𝑒 − )(𝑒 − ) 𝑒 −
𝑗𝜔 𝑗𝜔 𝑒 −
2 4 2 4
3 1
𝑒𝑗𝜔 = 𝐴 (𝑒𝑗𝜔 − ) + 𝐵 (𝑒𝑗𝜔 − )
4 2
At 𝑒𝑗𝜔 = 1 At 𝑒𝑗𝜔 = 3
2 4
1 = 𝐴(1 − 3) , ∴ 𝐴 = −2 3 = 𝐵(3 − 1) , ∴ 𝐵 = 3
2 2 4 4 4 2
𝑌(𝑒𝑗𝜔) −2 3 −2𝑒𝑗𝜔 3𝑒𝑗𝜔

= 𝑗𝜔 1 + 𝑗𝜔 3 => 𝑌(𝑒 ) = 𝑗𝜔 1 + 𝑒𝑗𝜔 − 3
𝑗𝜔
𝑒𝑗𝜔 𝑒 −2 𝑒 −4 𝑒 −2 4
Applying IDTFT
1 𝑛 3 𝑛
𝑦(𝑛) = −2 ( ) 𝑢(𝑛) + 3 ( ) 𝑢(𝑛)
2 4
352
EnggTree.com
Example 5.3: Find output response using Z-transform

3 1 3
𝑦(𝑛) − 𝑦(𝑛 − 1) + 𝑦(𝑛 − 2) = 2𝑥(𝑛) + 𝑥(𝑛 − 1)
2 2 2
Where 𝑦(−1) = 0 𝑦(−2) = 1 ( ) 1 𝑛
𝑥 𝑛 = (4) 𝑢(𝑛)
Solution:
3 1 3
𝑦(𝑛) − 𝑦(𝑛 − 1) + 𝑦(𝑛 − 2) = 2𝑥(𝑛) + 𝑥(𝑛 − 1)
2 2 2
Taking Z-transform
3 1
𝑌(𝑍) − [𝑍−1𝑌(𝑍) + 𝑦(−1)] +
[𝑍−2𝑌(𝑍) + 𝑍−1𝑦(−1) + 𝑦(−2)]
2 2
3
= 2𝑋 (𝑍) + [𝑍−1𝑋(𝑍) + 𝑥(−1)]
2
3 1 3
𝑌(𝑍) − [𝑍 𝑌(𝑍)] + [𝑍 𝑌(𝑍) + 1] = 𝑋(𝑍) [2 + 𝑍−1] ∵ 𝑦(−1) = 0, 𝑦(−2) = 1
−1 −2
2 2 2
𝑆𝑖𝑛𝑐𝑒 𝑥(𝑛)𝑖𝑠 𝑐𝑎𝑢𝑠𝑎𝑙 𝑠𝑖𝑔𝑛𝑎𝑙 𝑥(−1) = 0
1 𝑛 1 𝑍
𝑥(𝑛) = ( ) 𝑢(𝑛) ⟹ 𝑋(𝑍) = 1 = 1
4 1 − 𝑍−1 𝑍 −
3 −1 1 4 4
( ) −2 𝑍 3 −1 1
∴ 𝑌 𝑍 [1 − 𝑍 + 𝑍
2 ]= 1 [2 + 2 𝑍
]−
2 𝑍 −4 2
3 −1 1
𝑍 (2 + 𝑍 )
𝑌(𝑍) = 1
2
− 3
2
𝑍 − (1 − 3 𝑍−1 + 1 𝑍−2) 1 − 𝑍−1 + 1 𝑍−2
www.EnggTree.com
4 2
2
2
3
2
1 2
2
𝑍 2𝑍 + 𝑍 𝑍
𝑌(𝑍) = 𝑍 − 1 ( 2 3 2 1) − 2 2 3 1
𝑍 − 𝑍+ 𝑍 − 𝑍+
4 2 2 2 2
𝑌(𝑍) = (2𝑍 2 +3 𝑍) 1 𝑍 (2𝑍2 + 3 𝑍) − 1 𝑍 (𝑍 − 1 )

2 2 = 2 2 4
1 − 1 1
𝑍 (𝑍 − 1) (𝑍 − 1) (𝑍 − ) (𝑍 − 1) (𝑍 − ) (𝑍 − 1) ( 𝑍 − 1) (𝑍 − )
4 2 2 4 2
3 1
2𝑍2 + 𝑍 − 1 𝑍2 + 𝑍 3 2
𝑍 +
13
𝑍
2 2 8 2 8
= 1 = 1
1) (𝑍
(𝑍 − − 1) (𝑍 − ) (𝑍 − 1) (𝑍 − 1) (𝑍 − )
4 2 4 2
𝐴 𝐵 𝐶
= + +
(𝑍 − )
1 (𝑍 − 1) (𝑍 − 1)
4 2
3 13 1 1 1 1
𝑍2 + 𝑍 = 𝐴(𝑍 − 1) (𝑍 − ) + 𝐵 (𝑍 − ) (𝑍 − ) + 𝐶 (𝑍 − ) (𝑍 − 1)
1
2 8 8 2 4 2 4
At 𝑍 = :𝐴=
4 3
25
At 𝑍 = 1 :𝐵 =
3
1 −19
At 𝑍 = :𝐶 =
2 2
8 25 19
𝑌(𝑍) 3 3 2
∴ = 1 + − 1
𝑍 𝑍− 𝑍−1 𝑍−
4 2
8 𝑍 19 𝑍 25 𝑍
𝑌(𝑍) = + −
3𝑍 − 1 3𝑍−1 2𝑍−1
24
353
EnggTree.com
Applying inverse Z transform
8 1 𝑛 25 19 1 𝑛
𝑦(𝑛) = ( ) 𝑢(𝑛) + 𝑢(𝑛) − ( ) 𝑢(𝑛)
3 4 3 2 2
Example 5.4: Convolve the following sequences using Tabulation method

𝑛
𝑥(𝑛) = 𝑓𝑜𝑟 0 ≤ 𝑛 ≤ 6
3
ℎ(𝑛) = 1 𝑓𝑜𝑟 − 2 ≤ 𝑛 ≤ 2
0 1 2 3 4 5 6 1
𝑥(𝑛) = { , , , , , , } ℎ(𝑛) = {1, 1, , 1, 1}
𝗍 3 3 3 3 3 3 𝗍
Tabulation method
∞ ∞
𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛) = ∑ 𝑥(𝑘)ℎ(𝑛 − 𝑘) 𝑜𝑟 𝑦(𝑛) = ℎ(𝑛) ∗ 𝑥(𝑛) = ∑ ℎ(𝑘)𝑥(𝑛 − 𝑘)

𝑘=−∞ 𝑘=−∞
0 1 2 3 4 5 6 1
𝑥(𝑛) = { , , , , , , } ℎ(𝑛) = {1, 1, , 1, 1}
𝗍 3 3 3 3 3 3 𝗍
𝒌 −𝟒 −𝟑 −𝟐 −𝟏 𝟎 𝟏 𝟐 𝟑 𝟒 𝟓 𝟔 𝟕 𝟖 9
𝒙(𝒌) 0 1 2 3 4 5 6
3 3 3 3 3 3
𝒉(𝒌) 1 1 1 1 1
𝒉(−𝒌) 1 www.EnggTree.com
1 1 1 1
𝒉(−𝒌 − 𝟐) 1 1 1 1 1
𝒉(−𝒌 − 𝟏) 1 1 1 1 1
𝒉(−𝒌) 1 1 1 1
𝒉(−𝒌 + 𝟏) 1 1 1 1 1
𝒉(−𝒌 + 𝟐) 1 1 1 1 1
𝒉(−𝒌 + 𝟑) 1 1 1 1 1
𝒉(−𝒌 + 𝟒) 1 1 1 1 1
𝒉(−𝒌 + 𝟓) 1 1 1 1
𝒉(−𝒌 + 𝟔) 1 1 1 1 1
𝒉(−𝒌 + 𝟕) 1 1 1 1
𝒉(−𝒌 + 𝟖) 1 1 1 1
𝑦(−2) = (0)(1) = 0
1 1
𝑦(−1) = (0)(1) + ( ) (1) =
3 3
1 2
𝑦(0) = (0)(1) + ( ) (1) + ( ) (1) = 1
3 3
1 2 3
𝑦(1) = (0)(1) + ( ) (1) + ( ) (1) + ( ) (1) = 2
3 3 3
1 2 3 4 10
𝑦(2) = (0)(1) + ( ) (1) + ( ) (1) + ( ) (1) + ( ) (1) =
3 3 3 3 3
354
EnggTree.com
1 2 3 4 5
𝑦(3) = ( ) (1) + ( ) (1) + ( ) (1) + ( ) (1) + ( ) (1) = 5
3 3 3 3 3
2 3 4 5 6 20
𝑦(4) = ( ) (1) + ( ) (1) + ( ) (1) + ( ) (1) + ( ) (1) =
3 3 3 3 3 3
3 4 5 6
𝑦(5) = ( ) (1) + ( ) (1) + ( ) (1) + ( ) (1) = 6
3 3 3 3
4 5 6
𝑦(6) = ( ) (1) + ( ) (1) + ( ) (1) = 5
3 3 3
5 6 11
𝑦(7) = ( ) (1) + ( ) (1) =
3 3 3
6
𝑦(8) = ( ) (1) = 2
3 1 1 10 20 11
∴ 𝑦(𝑛) = {0, , , 2, , 5, , 6, 5, , 2}
3 𝗍 3 3 3
Example 5.5: Obtain Cascade form realization

1 1
𝑦(𝑛) − 𝑦(𝑛 − 1) − 𝑦(𝑛 − 2) = 𝑥(𝑛) + 3𝑥(𝑛 − 1) + 2𝑥(𝑛 − 2)
4 8
Solution:
1 1
𝑦(𝑛) − 𝑦(𝑛 − 1) − 𝑦(𝑛 − 2) = 𝑥(𝑛) + 3𝑥(𝑛 − 1) + 2𝑥(𝑛 − 2)
4 8
Taking Z-transform
1 1
𝑌(𝑍) − 𝑍−1 𝑌(𝑍) − 𝑍−2 𝑌(𝑍) = 𝑋(𝑍) + 3𝑍−1 𝑋(𝑍) + 2𝑍−1 𝑋(𝑍)
4 8
𝑌(𝑍) 1 + 3𝑍−1 + 2𝑍−1
= =
www.EnggTree.com
(1 + 𝑍−1)(1 + 2𝑍−1)
1 = 𝐻1(𝑍). 𝐻2(𝑍)
𝑋(𝑍) 1 − 𝑍−1 − 1 𝑍−2 (1 − 1 𝑍−1 )(1 + 1 𝑍−1)
4 8 2 4
(1 + 𝑍−1) (1 + 2𝑍−1)
𝐻1(𝑍) = 1 𝐻2(𝑍) = 1
(1 − 𝑍−1) (1 + 𝑍−1)
2 4
X 1(Z) 1 Y1(Z) X2(Z) 1 1 Y2(Z)
𝑍−1 𝑍−1
1/2 1 -1/4 2
Fig 5.19 Direct form II structure of 𝐻1(𝑍) Fig 5.20 Direct form II structure of 𝐻2(𝑍)
X(Z) 1 1 1 Y(Z)
𝑍−1 𝑍−1
1/2 1 -1/4 2
Fig 5.21 Cascade form
Example 5.6: Obtain Parallel form realization

1 1
𝑦(𝑛) − 𝑦(𝑛 − 1) − 𝑦(𝑛 − 2) = 𝑥(𝑛) + 3𝑥(𝑛 − 1) + 2𝑥(𝑛 − 2)
4 8
Solution:
355
EnggTree.com
1 1
𝑦(𝑛) − 𝑦(𝑛 − 1) − 𝑦(𝑛 − 2) = 𝑥(𝑛) + 3𝑥(𝑛 − 1) + 2𝑥(𝑛 − 2)
4 8
𝑌(𝑍) 1 + 3𝑍−1 + 2𝑍−2
= 1
𝑋(𝑍) 1 − 𝑍−1 − 1 𝑍−2
4 8
−16
1 1
− 𝑍−2 − 𝑍−1 + 1 2𝑍−2 + 3𝑍−1 + 1
8 4
2𝑍−2 + 4𝑍−1 − 16
(−) (−) (+)
−𝑍 + 17
−1
𝑌(𝑍) −𝑍−1 + 17 17−𝑍−1

= −16 + 1 1 = −16 +
𝑋(𝑍) 1 − 𝑍−1 − 𝑍−2 1 1
(1 − 𝑍−1 )(1 + 𝑍−1)
4 8 2 4
17−𝑍−1 = 𝐴 + 𝐵
𝐿𝑒𝑡 1 −1 1 −1 −1 1 −1
1
(1 − 𝑍 )(1 + 𝑍 ) 1− 𝑍 1+ 𝑍
2 4 2 4
1 1
17−𝑍−1 = 𝐴 (1 + 𝑍−1 ) + 𝐵(1 − 𝑍−1)
4 2
𝑎𝑡 𝑍−1 = −4 𝑎𝑡 𝑍 = 2
−1
1 www.EnggTree.com 1
17 + 4 = 𝐵(1 − (−4)) 17 − 2 = 𝐴(1 + (2))
2 4
∴𝐵=7 ∴ 𝐴 = 10
10 7
∴ 𝐻(𝑍) = −16 + 1 −1 + 1 −1
1− 𝑍 1+ 𝑍
2 4
𝑌1(𝑍) 10 𝑌2(𝑍) 7
𝐻1(𝑍) = = 1 𝐻2(𝑍) = = 1
𝑋1 (𝑍) 1 − 𝑍 −1 𝑋 2(𝑍) 1 + 𝑍 −1
2 4
𝑌1(𝑍) 𝑌2(𝑍)
= 10 ⇒ 𝑌1(𝑍) = 10𝑊1(𝑍) = 7 ⇒ 𝑌2(𝑍) = 7𝑊2(𝑍)
𝑊1(𝑍) 𝑊2(𝑍)
𝑊1(𝑍) = 1 𝑊2(𝑍) = 1
𝑋 (𝑍) 1 1 −1
−1 𝑋 (𝑍)
1 1− 𝑍 2 1+ 𝑍
2 4
1 1
𝑊1(𝑍) = 𝑋1(𝑍) + 𝑍−1𝑊1(𝑍) 𝑊2(𝑍) = 𝑋2(𝑍) − 𝑍−1𝑊2(𝑍)
2 4
X1(Z) 10 Y (Z) X2(Z) 7
1 Y2(Z)
𝑍−1 -1/4 𝑍−1

1/2
Fig 5.22 Direct form II structure of 𝐻1(𝑍) Fig 5.23 Direct form II structure of 𝐻2(𝑍)
Combining figures Fig 5.22 and Fig 5.23 we can form parallel form realization as shown in
356
EnggTree.com
Fig 5.24
X(Z) -16
10
𝑍−1
1/2
Y(Z)
7
-1/4
𝑍−1
Fig 5.24 Parallel form
Example 5.7: Convolve the following discrete time signals using graphical convolution
𝑥(𝑛) = ℎ(𝑛) = 𝑢(𝑛)
Solution:
𝑥(𝑛) = 𝑢(𝑛) = 1; 𝑛 ≥ 0
ℎ(𝑛) = 𝑢(𝑛) = 1; 𝑛 ≥ 0
x(n) h(n) h(-k)
1 1 1
www.EnggTree.com
0 1 2 3 4 5 n 0 1 2 3 4 5 n
-5 -4 -3 -2 -1 0 n
∞
𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛) = ∑ 𝑥(𝑘)ℎ(𝑛 − 𝑘)
𝑘=−∞
𝑤ℎ𝑒𝑛 𝑛 = 0
𝑦(0) = (1)(1) = 1
𝑦(1) = (1)(1) + (1)(1) = 2
h(-k+1)
-4 -3 -2 -1 0 1 n
𝑦(2) = (1)(1) + (1)(1) + (1)(1) = 3
h(-k+2)
-3 -2 -1 0 1 2 n
∴ 𝑦(𝑛) = {1,2,3,4,5, … . }
357
EnggTree.com
Example 5.8: Compute linear convolution.

𝑥(𝑛) = {2,2,0,1,1} ℎ(𝑛) = {1,2,3,4}
Solution:
2 2 0 1 1
1 2 2 0 1 1
2 4 4 0 2 2
3 6 6 0 3 3
4 8 8 0 4 4
𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛) = {2,6,10,15,11,5,7,4}
www.EnggTree.com
358

EC3354 Signals and Systems Lecture Notes 2

Uploaded by

Copyright:

Available Formats

EC3354 Signals and Systems Lecture Notes 2

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

EC3354 Signals and Systems Lecture Notes 2

Uploaded by

Copyright:

Available Formats

`

Unit 1: Classification of signals and systems

Amplitude Sine Wave

Fig 1.2 Digital Signal

Continuous time and discrete time signal

Continuous time signal:

Discrete time signal:

Fig 1.3 Discrete time signal

Even (symmetric) and Odd (Anti-symmetric) signal

Periodic and Aperiodic signal

Energy and Power signal

For Discrete time signals,

Random signal (Non-Deterministic signal):

Causal and Non-causal signal

Basic(Elementary or Standard) continuous time signals

Unit ramp signal is defined as

Unit Parabolic signal

• Unit ramp signal is obtained by integrating unit step signal

• Unit step signal is obtained by differentiating unit ramp signal

Unit Pulse signal is defined as

Unit Pulse signal

Unit Impulse signal

𝑥(𝑡)𝛿 𝑡 𝑑𝑡 = 𝑥 0 𝛿 0 = 𝑥 0 [∵ 𝛿 𝑡 𝑒𝑥𝑖𝑠𝑡𝑠 𝑜𝑛𝑙𝑦 𝑎𝑡 𝑡 = 0 𝑎𝑛𝑑 𝛿 0 = 1]

where Ω = 2πf = 2π and Ω is angular frequency in rad/sec

Cosinusoidal signal Sinusoidal signal

x(t) a=0 x(t) a>0 x(t) a<0

DC signal Exponentially Exponentially

Exponentially growing sinusoidal signal

𝑤𝑕𝑒𝑛 𝜎 = −𝑣𝑒, 𝑡𝑕𝑒𝑛 𝑥 𝑡 = 𝐴𝑒 −𝜎𝑡 (𝑐𝑜𝑠Ω𝑡 + 𝑗𝑠𝑖𝑛Ω𝑡),

Exponentially decaying Cosinusoidal signal Exponentially decaying sinusoidal signal

Basic(Elementary or Standard) Discrete time signals

www.EnggTree.comUnit step signal

Unit Ramp signal

Pulse signal (Rectangular pulse function)

Unit Impulse signal

Unit Impulse signal

where 𝜔 = 2πf = 2π m and 𝜔 is frequency in radians/sample

Decreasing exponential signal Increasing exponential signal

Complex Exponential signal is defined as 𝑥 𝑛 = 𝑎𝑛 𝑒𝑗 (𝜔 0𝑛) = 𝑎𝑛 [𝑐𝑜𝑠𝜔0𝑛 + 𝑗𝑠𝑖𝑛𝜔0𝑛]

Exponentially decreasing Cosinusoidal signal

Exponentially growing Cosinusoidal signal

Exponentially growing sinusoidal signal

Continuous time and Discrete time system

Fig 1.84 Representation of continuous time system

Linear system and Non Linear system

Discrete time domain:

where y1 n and y2 n are the responses of x1 n and x2 n respectively

Static (Memoryless) and Dynamic (Memory) system

Discrete time domain:

Time variant system:

Discrete time domain:

Time invariant system:

Causal and Non-Causal system

Discrete time domain:

𝑦(𝑛) = 𝑥(−𝑛) + 𝑥(𝑛 + 4)

Stable and Unstable system

𝑖. 𝑒. , 0 < 𝑕(𝜏) 𝑑𝜏 < ∞

𝑖. 𝑒. , 0 < 𝑕(𝑘) < ∞

1. Draw 𝑟(𝑡 + 3), where r(t) is ramp signal

2. Sketch 𝑥(𝑡) = 3𝑟(𝑡 − 1) + 𝑟(−𝑡 + 2)