Signals & Systems

MT - 252
Lecture # 11: Continuous-Time LTI System
Impulse Response & Convolution Integral

Dr. Hafiz Zia Ur Rehman

[email protected]

Department of Mechatronics Engineering

Air University, Islamabad
 Representation CT signals in terms of Impulses

◼ Discrete time vs Continuous time Convolution

◼ In analogy with the results derived and discussed in the preceding lecture,
the goal of this lecture is to obtain a complete characterization of a
continuous-time LTI system in terms of its unit impulse response.
◼ In discrete time, the key to developing the convolution sum was the sifting
property of DT unit impulse i.e.,
◼ The mathematical representation of a signal as the superposition of scaled and
shifted unit impulse functions.
◼ The discrete-time system as responding to a sequence of individual impulses.
◼ In continuous time
◼ A unit impulse is the idealization of a pulse which is so short that its duration is
inconsequential for any real, physical system.
◼ We can develop a representation for arbitrary continuous-time signals in terms of
these idealized pulses with vanishingly small duration, or equivalently,
impulses.
2
 Representation CT signals in Terms of Impulses

◼ Representation of Continuous-Time Signals in terms of Impulses

◼ To develop the continuous-time counterpart of the discrete-time sifting
property, we begin by considering a pulse or "staircase" approximation, 𝒙(𝒕),
to a continuous-time signal 𝒙(𝒕).

3
 Digital Image Processing, 3rd ed.
Gonzalez & Woods
www.ImageProcessingPlace.c
om
Chapter 10
Image Segmentation


xˆ ( t ) =  x ( k  )  (t − k  ) 
k =−


© 1992–2008 R. C. Gonzalez & R. E. Woods

 Representation CT signals in Terms of Impulses

◼ Representation of Continuous-Time Signals in terms of Impulses

◼ In a manner similar to that employed in the discrete-time case,

this approximation can be expressed as a linear combination of

delayed pulses.

xˆ ( t ) =  x ( k  )  (t − k  ) 
k =−



x ( t ) = lim
→ 0
 x ( k  )  (t − k  ) 
k =−


◼ Representing the signal 𝒙(𝒕) as a "sum" (more precisely, an

integral) of weighted, shifted impulses.

x (t ) =  x ( )  ( t −  ) d
− 5
 Digital Image Processing, 3rd ed.
Gonzalez & Woods
www.ImageProcessingPlace.c
◼ Example-01: Unit Step Signal in terms
om of Impulses
Chapter 10
Image Segmentation

u (t ) =  u ( )  ( t −  ) d
−

u ( t ) =  (1)  ( t −  ) d
0

 u ( ) = 0 for   0 and
u ( t ) =   ( t −  ) d
0
u ( ) = 1 for   0

◼ Sifting property:


 x ( t )  ( t ) dt = x ( 0 )
−


 x ( t )  ( t − t ) dt = x ( t )
−
0 0

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Digital Image Processing, 3rd ed.
Gonzalez & Woods
Properties ofomUnit Impulse
www.ImageProcessingPlace.c

➢ Time Scaling Property: Chapter 10

Image Segmentation
1
 ( t ) =  (t )   n  =  n 

   n    =  n   
  ( t   ) =
1
 (t   )

➢ Product property:

x (t ) (t ) = x ( 0) (t ) Point to Ponder!

x ( t )  ( t − t0 ) = x ( t0 )  ( t − t0 ) 100

 x ( t )  ( t − 1) dt = 0
2
➢ Sifting property:


 x ( t )  ( t ) dt = x ( 0 )
−


 x ( t )  ( t − t ) dt = x ( t )
−
0 0

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Digital Image Processing, 3rd ed.
Gonzalez & Woods
Sifting property:
 www.ImageProcessingPlace.c
y (t ) =  x ( )  ( − t ) d , om

−
Chapter 10 

y (t ) =


 x ( t )  ( − t ) d ,
Image Segmentation
 x ( t )  ( t − t ) dt = x ( t )
−
0 0

−
 
y ( t ) = x ( t )   ( − t ) d ,   ( − t ) d = 1
− −

y (t ) = x (t ) i.e., a constant value

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Unit Impulse Response
◼ Response of CT-LTI system
◼ The impulse response, as it is named, is the response of the system to a unit impulse
input.

x (t ) =  (t ) System y (t ) = h (t )

Unit Impulse Response to a Unit Impulse

◼ In another way, the “impulse response” of a system, i.e., 𝒉 𝒕 , is the output that it
produces in response to an impulse input.

Definition: if and only if 𝒙 𝒕 = 𝜹 𝒕 then 𝒚 𝒕 = 𝒉 𝒕

9
 Digital Image Processing, 3rd ed.
Gonzalez & Woods
www.ImageProcessingPlace.c
Just remove this to make
om it more simple
Chapter 10
Image Segmentation

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Digital Image Processing, 3rd ed.
Gonzalez & Woods
▪ Response of CT-LTI system www.ImageProcessingPlace.c
om
Chapter 10
Image Segmentation
x ( 0 ) hˆ0 ( t )  → x ( 0 ) hˆ ( t − 0 ) 

x ( k  ) hˆk  ( t )  → x ( k  ) hˆ ( t − k  ) 

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Digital Image Processing, 3rd ed.
Gonzalez & Woods
www.ImageProcessingPlace.c
Linear Time Invariance (LTI) om
Chapter 10
Image
LTISegmentation
𝛿 𝑡 ℎ 𝑡

– Time invariance property:

𝛿 𝑡 − 𝑡0 LTI ℎ 𝑡 − 𝑡0

– Scaling Property:
𝑎𝛿 𝑡−1 LTI 𝑎ℎ 𝑡−1

– Time invariance and Scaling Property:

𝑥 1 𝛿 𝑡−1 LTI 𝑥 1 ℎ𝑥 𝑡 − 1

– Time invariance, Scaling, and Additive Property or LTI property:

𝑥 1 𝛿 𝑡−1 +𝑥 2 𝛿 𝑡−2 LTI 𝑥 1 ℎ𝑥 𝑡 − 1 + 𝑥 2 ℎ𝑥 𝑡 − 2

+∞ +∞
𝑦 𝑡
𝑥 𝑡
න 𝑥 𝜏 𝛿 𝑡 − 𝜏 𝑑𝜏 LTI න 𝑥 𝜏 ℎ 𝑡 − 𝜏 𝑑𝜏
−∞ −∞

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Digital Image Processing, 3rd ed.
Gonzalez & Woods
▪ Response of CT-LTI system www.ImageProcessingPlace.c
om
Chapter 10
Image Segmentation

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Digital Image Processing, 3rd ed.
Gonzalez & Woods
◼ Convolution-Integral www.ImageProcessingPlace.c
om
Chapter 10
෡ 𝒌∆ (𝒕) denote the response of an CT-LTI system to the input 𝛿∆ (𝑡 − 𝑘∆).
◼ Let 𝒉
Image Segmentation
Then, from the superposition property for continuous-time linear systems:

 
yˆ ( t ) =  x ( k  ) hˆ ( t − k  )  yˆ ( t ) =  x ( k  ) hˆ ( t − k  ) 
k =−
k =−
 
y ( t ) = lim
→
 x ( k  ) hˆk  ( t )  yˆ ( t ) = lim
→
 x ( k  ) hˆ ( t − k  ) 
k =−
k =−

 
y (t ) =  x ( ) h ( t ) d y (t ) =  x ( ) h ( t −  ) d
− −

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Convolution-Integral Representation (CT-LTI)

◼ Convolution-Integral (summary)
◼ The expression (given below) referred to as the convolution integral or the
superposition integral, is the CT counterpart of the convolution sum of DT
and corresponds to the representation of a CT LTI system in terms of its
response to a unit impulse.

y (t ) =  x ( ) h ( t −  ) d
−

◼ The convolution of two signals 𝒙(𝒕) and 𝒉(𝒕) will be represented

symbolically as

y (t ) = x (t )  h (t )
15
 Digital Image Processing, 3rd ed.
Gonzalez & Woods
www.ImageProcessingPlace.c
Example-02: Let 𝒙(𝒕) be the inputomto an LTI system with unit impulse
Chapter 10
response 𝒉(𝒕), where
Image Segmentation
x ( t ) = e − at u ( t )
x ( t ) = e − at u ( t ) , a0

h (t ) = u (t )
x ( ) = e − a u ( )
Solution:

y (t ) = x (t )  h (t ) h (t ) = u (t )
 
y (t ) =  x ( ) h ( t −  ) d h (t − ) = u (t − )
−

y (t ) =  u ( ) u ( t −  ) d
e − a

−

y (t ) =  u ( ) u ( − + t ) d
e − a

−

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Digital Image Processing, 3rd ed.
Gonzalez & Woods
www.ImageProcessingPlace.c
Example-02: Let 𝒙(𝒕) be the inputomto an LTI system with unit impulse
Chapter 10
response 𝒉(𝒕), where
Image Segmentation

x ( t ) = e − at u ( t ) , a0
h (t ) = u (t )
Solution:

y (t ) = x (t )  h (t )


y (t ) =  x ( ) h ( t −  ) d
−

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Digital Image Processing, 3rd ed.
Gonzalez & Woods
Solution: www.ImageProcessingPlace.c
om
Chapter 10
Image Segmentation


y (t ) =  x ( ) h ( t −  ) d
−

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Digital Image Processing, 3rd ed.
Gonzalez & Woods
Solution: www.ImageProcessingPlace.c
om
Chapter 10
Image Segmentation

Zero

0, otherwise
x ( ) h ( t −  )   − a
 e , 0   t

x ( ) h ( t −  )  e − a u ( )u ( t −  )

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Digital Image Processing, 3rd ed.
Gonzalez & Woods
Solution: www.ImageProcessingPlace.c
om
Chapter 10

y (t ) =  x ( ) h ( t −  Image
) d Segmentation
−

y (t ) =  u ( ) u ( t −  ) d
e − a

−

t
y ( t ) =  e − a (1) (1) d
0

t
1 − a
y (t ) = − e
a 0

y (t ) =
1
a
(1 − e − at ) u ( t )

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Digital Image Processing, 3rd ed.
Gonzalez & Woods
◼ Convolution Comparison www.ImageProcessingPlace.c
om
Chapter 10
Image Segmentation

Convolution sum Convolution integral

(for DT-LTI systems) (for CT-LTI systems)


+
x  n =  x k   n − k  x (t ) =  x ( )  ( t −  ) d
k =− −



y  n =  x k  h n − k  y (t ) =  x ( ) h ( t −  ) d
k =− −
y  n = x  n  h  n y (t ) = x (t )  h (t )

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Convolution-sum Representation (DT-LTI)

◼ Example 2.8: Let 𝒙(𝒕) be the input to an LTI system with unit impulse
response 𝒉(𝒕), where

x ( t ) = e 2t u ( −t ) ,
h ( t ) = u ( t − 3) .

Solution: ~Try yourself~ 22

 Digital Image Processing, 3rd ed.
Gonzalez & Woods
www.ImageProcessingPlace.c
Solution: om
Chapter 10
Image Segmentation

t −3 2  0 2
y ( t ) =   e d , −    t − 3 y ( t ) =   e d , −    0
 −  −

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Digital Image Processing, 3rd ed.
Gonzalez & Woods
Solution: www.ImageProcessingPlace.c
om
Chapter 10
Image Segmentation
t −3 2
  e d , −    t − 3
 −
y (t ) =  0 ,
 e 2 d ,

−    0
 −

 1 2( t − 3)
 2 e , −    t − 3
y (t ) = 
 1, −    0
 2

© 1992–2008 R. C. Gonzalez & R. E. Woods

 Acknowledgement
◼ The slides are prepared based on the following textbook:
◼ Chapter # 2, Discrete-Time Signal Processing
by Alan V. Oppenheim, Ronald W. Schafer & John R. Buck. 2nd Edition,
Pearson Education - Prentice Hall, 1999
◼ Reading Assignment :2.1.1,2.1.2,
◼ Special thanks to
◼ Prof. Dr. Muhammad Shehzad Hanif