Chapter 2

Basic Concepts
§ Role of Linear Time Invariant Systems
– Many physical systems possesses these properties and can be
modelled with LTI Systems
– LTI Systems can be analysed in considerable detail, providing
both insight into their properties and a set of powerful tools that
form the core of signal and system analysis

– Reason for good for analysis:

• Superposition property
• Any input signal can be broken into basic signals and
corresponding ‘response’ (output signal) is observed
• The response to the sum of input signals will be just the sum
of individual responses

Dr. Hassan Tariq Chattha 2

 Basic Concepts
§ Discrete-Time Linear Time Invariant Systems
– Any DT- Signal can be represented as a combinations of shifted
and weighted impulses

§ Convolution Sum
– For an LTI System, if the response to a Unit Impulse signal (input
signal) is known, Impulse Response of the system, then the
response to any signal is sum of the individual Impulse
Responses with corresponding Time-shifts and weights

• Linear => weighted sum of responses

• Time Invariant => for shifted impulse, same shift in output

Dr. Hassan Tariq Chattha 3

 Convolution Sum
§ Any Discrete Time Signal can be represented as a sequence of
impulses

§ An arbitrary sequence is a linear combination of shifted unit

impulses, δ[n – k], with corresponding weight “x[k]”

Dr. Hassan Tariq Chattha 4

 Convolution Sum
§ Convolution Sum
– For an LTI System, if the response to a Unit Impulse signal (input
signal) is known, Impulse Response of the system, then the
response to any signal is sum of the individual Impulse
Responses with corresponding Time-shifts and weights
• Linear => weighted sum of responses
• Time Invariant => for shifted impulse, same shift in output
x[n ] = x[0]V [n ] + x[1]V [n - 1] + x[2]V [n - 2]
– Unit Impulse response x[0 ]V [n - 0 ] ® x[0 ]h[n - 0 ]
V [n ] ® h[n ] x[1]V [n - 1] ® x[1]h[n - 1]
x[2 ]V [n - 2 ] ® x[2 ]h[n - 2 ]
– Response to ‘x[n]’
y[n ] = x[0]h[n ] + x[1]h[n - 1] + x[2]h[n - 2]
Dr. Hassan Tariq Chattha 5
 Convolution Sum
§ Convolution/Superposition Sum: LTI System

x[ n ] = ... + x[0 ]V [n ] + x[1]V [n - 1] + x[2 ]V [n - 2 ] + ...

V[n ] ®h[n ]
y[n ] = ... + x[0 ]h[n ] + x[1]h[n - 1] + x[2 ]h[n - 2 ] + ...

k = +¥

y[n ] = å x[k ]h[n - k ]

k = -¥

– n→ sample no. for y[n] (output); k→ sample no. for x[k] (input)
– An LTI System is completely characterised by its unit impulse
response
y[n ] = x[n ] * h[n ]
Dr. Hassan Tariq Chattha 6
Properties of convolution

7
 Example 2.1

Dr. Hassan Tariq Chattha 8

 Example 2.1

Dr. Hassan Tariq Chattha 9

 Example 2.2

Dr. Hassan Tariq Chattha 1

0
 Example 2.2

Dr. Hassan Tariq Chattha 11

 Example 2.2

Dr. Hassan Tariq Chattha 12

 Example 2.2

Dr. Hassan Tariq Chattha 13

 Example 2.3

Dr. Hassan Tariq Chattha 14

 Example 2.3

Dr. Hassan Tariq Chattha 15

 Example 2.3

Dr. Hassan Tariq Chattha 16

 Math Review

N
N +1
=a -a
M
åan a ¹1
For Reference- See
Geometric Series,
n=M 1-a Finite and Infinite
+¥ Sum Formulas and
1
å =
a n
a <1
Problem 1.54

n=0 1-a
+¥
a
å na n =
1-a 2
a <1
n=0

Dr. Hassan Tariq Chattha 17

 Convolution Sum
§ Example:

Dr. Hassan Tariq Chattha 17

1
8
 Convolution Sum

11 Adopted from Oppe

Dr. Hassan Tariq Chattha 19 nheim & Willsky

 Convolution Sum

Dr. Hassan Tariq Chattha 20

 Example 2.4

Dr. Hassan Tariq Chattha 21

 Convolution Sum

Dr. Hassan Tariq Chattha 22

 Example 2.4

Dr. Hassan Tariq Chattha 23

 Example 2.4

§ Study
– Example 2.5

Dr. Hassan Tariq Chattha 24

 Convolution Integral
k = +¥

y[ n ] = å x[ k ]h[ n - k ]
k = -¥

§ Continuous Time counter-part of Convolution Sum

– Convolution Integral/Superposition Integral

+¥

y (t ) = ò x (t )h (t - t )d t
-¥

y (t ) = x (t ) * h (t )
– t→ y(t) (output) time; Ƭ→ x(t) (input) time
– An LTI System is completely characterised by its unit impulse
response
Dr. Hassan Tariq Chattha 25
 Convolution Integral

Dr. Hassan Tariq Chattha 26

 Convolution Integral

Dr. Hassan Tariq Chattha 27

 Convolution Integral

+¥

y (t ) = ò x (t ) h (t - t ) d t
-¥

x (t ) = 1 0 <t <T h (t ) = t 0 < t < 2T

h (t - t ) = t - t 0 < t - t < 2T

h (t - t ) = t - t t - 2T < t < t
Dr. Hassan Tariq Chattha
28
 Convolution Integral

Dr. Hassan Tariq Chattha 29

Dr. Hassan Tariq Chattha 30
Dr. Hassan Tariq Chattha 31 Adopted from Oppenheim & Willsky
 Convolution Integral

§ Study
– Example 2.8

Dr. Hassan Tariq Chattha 32

 LTI Systems Properties
+¥

y[n ] = å x[k ]h[n - k ] = x[n ] * h[n ]

k = -¥
+¥

y (t ) = ò x (t ) h (t - t ) d t = x (t )*h (t )
-¥

§ The Commutative Property

+¥

x[ n ] * h[ n ] = h[ n ] * x[ n ] = å h[ k ] x[n - k]
k = -¥

"
x(t) * h(t) =h(t) * x(t) = ∫!" ℎ(𝜏) 𝑥(𝑡 − 𝜏)d𝜏

Dr. Hassan Tariq Chattha 33

 LTI Systems Properties
§ The Distributive Property

x[n ] * {h1 [n ] + h 2 [n ]} = x[n ] * h1 [n ] + x[n ] * h 2 [ n ]

x (t) * {h1 (t)+ h 2 (t)} = x (t) * h1 (t) + x (t) * h 2 (t)

Dr. Hassan Tariq Chattha 34

 LTI Systems Properties
1
x[n ] = ( ) n u[n ] + (2) n u[- n ] h[n ] = u[n ]
2

y[n ] = x[n ] * h[n ]

y[n ] = { x1 [n ] + x 2 [n ]} * h[n ]

y[n ] = x1 [ n ] * h[ n ] + x 2 [n ] * h[ n ]

y[ n ] = y 1 [n ] + y 2 [ n ]

§ Using Results from

– Examples: 2.3, 2.5

Dr. Hassan Tariq Chattha 35

 LTI Systems Properties
§ The Associative Property

x[n ] * {h1 [n ] * h 2 [n ]} = { x[n ] * h1 [n ]} * h 2 [n ]

= { x[n ] * h 2 [n ]} * h1 [n ]

Dr. Hassan Tariq Chattha 36

 LTI Systems Properties
§ With or Without Memory
– A system is memoryless if its output at any time depends only on
the value of the input at that same time.
– That is true only if

– Causal/Not Causal ?

Dr. Hassan Tariq Chattha 37

 LTI Systems Properties
§ Invertibility

Dr. Hassan Tariq Chattha 38

 LTI Systems Properties
§ Example: 2.11

§ Impulse Response?
– Take input as unit impulse function

– Thus

– If we take

– Then

Dr. Hassan Tariq Chattha 30

3
9
 LTI Systems Properties
§ Example: 2.12

§ Summer/accumulator/running sum

§ Inverse of summer

§ Impulse Response

§ Thus

Dr. Hassan Tariq Chattha 40

 LTI Systems Properties
§ Causality

Ø From Example 2.12

Ø And its inverse

Ø Both Causal

Dr. Hassan Tariq Chattha 41

 LTI Systems Properties
§ Stability
– Impulse Response must be absolutely summable/integrable.

§ Example 2.13
– Time shifting systems are stable

– Summer is Unstable.

Dr. Hassan Tariq Chattha 42

 LTI Systems Properties
§ Unit Step Response
– Output of the system when input is a Unit Step Function

– From Example 2.12, it is clear that h[n]

can be recovered from s[n] using the
relation

• For Continuous time, the unit step

response is

• And unit impulse response is the first

derivative of unit step response
Dr. Hassan Tariq Chattha 43
 Causal LTI Systems
§ CT-System described by Linear Constant-Coefficient Differential
Equation
𝑑𝑦(𝑡)
+ 2𝑦 𝑡 = 𝑥(𝑡)
𝑑𝑡
– Implicit relationship between input and output
– Explicit relationship: Solution of differential equation

– For a particular input

x(t)= 𝐾𝑒 #$ u(t)
– Complete Solution: Particular solution (forced response) +
Homogeneous solution (natural response)

𝑦 𝑡 = 𝑦% (𝑡) + 𝑦& (𝑡)

Dr. Hassan Tariq Chattha 44
 Causal LTI Systems
§ Since x(t)= Ke3t for t >0, the forced Response solution for t> 0, is of the
same form as the input
yp(t)= 𝑌𝑒 #$

§ Where Y is the number we need to determine, as

§ So the forced response for t >0, we get

3𝑌𝑒 #$ + 2𝑌𝑒 #$ = Ke3t

Dr. Hassan Tariq Chattha 45

 Causal LTI Systems
§ Cancelling the factor e3t from both sides, we obtain,

3Y + 2Y = K

'
Y= (
So that
' #$
yp(t)= (
𝑒 , 𝑡>0
§ Homogeneous Differential Equation
𝑑𝑦(𝑡)
+ 2𝑦 𝑡 = 0
𝑑𝑡
§ Solutionof Homogeneousdifferential equation

Dr. Hassan Tariq Chattha 46

 Causal LTI Systems
§ Solution of Homogeneous differential equation

§ From this equation, we see that s = -2 and Ae-2t is the solution for
any choice of A.
§ So the total solution of the differential equation for t>0 is

'
y(t)= 𝐴𝑒 !)$ + 𝑒 #$ , 𝑡>0
(

Dr. Hassan Tariq Chattha 47

 Causal LTI Systems
§ In ordertohaveexactrelationship between input and output,
auxiliaryconditionsshouldbe used.
§ For a causal LTI system, if x(t) = 0 for t <0, then y(t) = 0 for t <0,
i.e. y(0) = y0

§ This implied that '

0= 𝐴 + (
'
or A= −
(
!
§ Thus for t >0, y(t)= "
𝑒 #$ − 𝑒 !)$
§ While for t <0, y(t) = 0. Combining these two cases, we obtain the
full solution
'
y(t)= 𝑒 #$ − 𝑒 !)$ u(t)
(

Dr. Hassan Tariq Chattha 48

 Causal LTI Systems
§ So far, we have used the first-order differential equation.
§ A general Nth-order linear constant–coefficient differential equation is
given by

§ The order refers to the highest derivative of the output y(t) appearing
in the equation. In case when N=0, we have

§ The solution y(t) will also consist on two parts – a particular solution
and a solution of the homogenous differential equation

§ Rest of the procedure is same as described earlier.

Dr. Hassan Tariq Chattha 49

 Causal LTI Systems
§ DT-System described by Linear Constant-Coefficient Difference
Equation
– N-th order

– System is linear if auxiliary conditions are zero

– Complete Solution: Particular solution (forced response) +
Homogeneous solution (natural response)

– Homogeneous

§ Alternative method to solve, Nth-order difference equation

– Re-arrange

Dr. Hassan Tariq Chattha 50

 Causal LTI Systems
§ Alternative method to solve, N-th order difference equation
– Re-arrange the difference equation

– If previous output values are known then next value can be calculated
‘recursively’. And this equation is called recursive equation as it
specifies a recursive procedure for determining the output in terms of
input and previous outputs and Auxiliary conditions required.
§ In the special case when N = 0, then

– Explicit relationship between input and output and this equation is often
called a nonrecursive equation as it does not recursively use the
previous computed values of output to find the computed value of output
• No auxiliary conditions required
• Convolution sum equation
• FIR system
Dr. Hassan Tariq Chattha 51
 Causal LTI Systems
§ Example:

Dr. Hassan Tariq Chattha 52

 Causal LTI Systems
§ Example:

Dr. Hassan Tariq Chattha 53

 Causal LTI Systems
§ Example:

§ To ensure Linearity and causality, System must be in initial rest.

– Then impulse response of this system

– IIR system
• A system described by recursive difference equation with
infinite duration impulse response.

Dr. Hassan Tariq Chattha 54

 Block Diagram of LTI System
§ LTI systems represented by Difference Equation
– Structure and realisation of LTI system (Block Diagram)
– Difference equation requires
• (i) Adder (ii) Multiplier (iii) Delay

Dr. Hassan Tariq Chattha 55

 Block Diagram of LTI System
§ Difference Equation
– Initial Rest

§ Recursive Solution

Dr. Hassan Tariq Chattha 56

 Block Diagram of LTI System
§ Difference Equation

§ Non-Recursive Solution

Dr. Hassan Tariq Chattha 57

 Block Diagram of LTI System
§ Difference Equation
– Initial Rest

§ Recursive Solution

Dr. Hassan Tariq Chattha

58
 Block Diagram of LTI System
§ LTI systems represented by Differential Equation
– Structure and realisation of LTI system (Block Diagram)
§ Consider next the causal continuous-time system described by first-order
differential equation:

§ Let us rewrite as

§ The right hand side of this equation shows that differential equation also
requires these three basic functions
• (a) Adder (b) Multiplier (c) Delay

Dr. Hassan Tariq Chattha 58

 Block Diagram of LTI System

§ Therefor, the above can be represented by block diagram as

§ Alternatively the above equations can be rewritten as

Dr. Hassan Tariq Chattha 59

 Block Diagram of LTI System

§ And the integrating from -∞ to t. If we assume that the system is initially at

rest, then the integral of dy(t) /dt from -∞ to t is precisely y(t). Consequently
we obtain

§ The block diagram of the system is

shown as

Dr. Hassan Tariq Chattha 60

 END PROBLEMS CHAPTER 2

§ 2.1 – 2.39

Dr. Hassan Tariq Chattha 62

Dr. Hassan Tariq Chattha 63