0% found this document useful (0 votes)
33 views20 pages

Frequency-Domain and Frequency Response: 2.6 Representation of Discrete-Time Signals and Systems

1. The document discusses frequency-domain representation of discrete-time signals and systems. Complex exponentials are the eigenfunctions for linear time-invariant (LTI) discrete-time systems. 2. The frequency response of a system is defined as its output when the input is a complex exponential. It represents how the system transforms inputs of different frequencies. 3. For LTI systems, the frequency response is the transfer function evaluated on the unit circle. It is a periodic function with period 2π.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
33 views20 pages

Frequency-Domain and Frequency Response: 2.6 Representation of Discrete-Time Signals and Systems

1. The document discusses frequency-domain representation of discrete-time signals and systems. Complex exponentials are the eigenfunctions for linear time-invariant (LTI) discrete-time systems. 2. The frequency response of a system is defined as its output when the input is a complex exponential. It represents how the system transforms inputs of different frequencies. 3. For LTI systems, the frequency response is the transfer function evaluated on the unit circle. It is a periodic function with period 2π.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 20

SY20212022 12/7/2021

Frequency-Domain and
Frequency Response

1 ECEA108

2.6 Frequency-Domain
Representation of Discrete-
Time Signals and systems
2.6.1 Eigenfunction and Eigenvalue for LTI

If y n  T x n    x n 

 xn is called as the eigenfunction of the


system T   , and the associated eigenvalue
is  .

2 ECEA108

Compiled by LDV 1
SY20212022 12/7/2021

Eigenfunction and Eigenvalue


Complex exponentials is the eigenfunction
for LTI discrete-time systems hn  :
jwn
x  n  e ,   n  

y  n  h  n  x  n   h k  x n  k 
k 

jw n  k 
  h k e  
   h k e
 jwk  jwn
k  e
 k  
 jw  jwn
 H e e eigenfunction
 
eigenvalue frequency response
3 ECEA108

Frequency response
   jwk  jwn  jw  jwn
y  n    h  k  e  e  H e e
 k    
 
 H e jw is called as frequency response of
the system.
i.e. DTFT, (Discrete-Time) Fourier Transform
Real part, imaginary part
H  e jw   H R  e jw   jH I  e jw 
Magnitude, phase
H e
 
jw
  H e jw
e jH e jw

4 ECEA108

Compiled by LDV 2
SY20212022 12/7/2021

Example 2.14 Find Frequency


response of the ideal Delay
solution 1: x  n   e jwn yn  xn  nd 
yn  e jwn nd   e  jwnd e jwn
H  e jw   e
 jwnd
solution 2:
From defination of frequency response:

H e    h  n e
jw  jwn
, h  n     n  nd 
 n 

   n  n  e e
 jwn  jwnd
d
n 
5 ECEA108

Example 2.14 Frequency


response of the ideal Delay
H  e jw   e  jwnd
 cos  wnd   j sin  wnd 
 H R  e jw   jH I  e jw 
 1e  jwnd
H  e jw   1, H  e jw    wnd

6 ECEA108

Compiled by LDV 3
SY20212022 12/7/2021

x[n]: Linear combination of complex exponential

If x  n   e jwk n
k , output of LTI system?
k

From the principle of superposition, and


 jw  jwn
y  n  H  e  e x  n  e jwn
    n  

the corresponding output of an LTI system is

y  n     k H e jwk e jwk n  
k

7 ECEA108

Example 2.15 Sinusoidal


response of LTI systems
A j jw0n A  j  jw0n
x  n  A cos  w0 n     e e  e e
2 2
y  n  H  
e jw0 A j jw0n
e e
2
 H   e e
e  jw0 A  j  jw0n
2
if h n is real, then H  e jw   H *  e jw 0 0

 H e  e
jw0 H e jw0   , H  e jw 0  

   h k e
 jw0 k
  k 

 
y  n   A H e jw0 cos  w0 n      ,   H e jw0  
8 ECEA108

Compiled by LDV 4
SY20212022 12/7/2021

Sinusoidal response of the ideal Delay y n  x n  nd 

Given: x  n   A cos  w0 n    , find y[n].


solution 1: y  n  x  n  nd   A cos w0  n  nd    
solution 2: 
H e jw0  e   jw0nd
e jw0 n
 
H e jw0  1,    w0 nd

 
y  n   A H e jw0 cos  w0 n      ,   H e jw0  
 A cos  w0 n    w0 nd 
 A cos  w0  n  nd    
9 ECEA108

Periodic Frequency Response


The frequency response of discrete-time
LTI systems is always a periodic function of
the frequency variable w with period 2

 

H e j  w 2 
  h  n  e  j  w 2  n
=H  e jw 
n 

 h  n e  jwn
=H  e jw 
n 

e j  w 2 n  e jwn e j 2 n  e jwn
H e j  w 2 r 
  H e  , jw
for r an integer
10 ECEA108

Compiled by LDV 5
SY20212022 12/7/2021

Periodic Frequency Response


 
We need only specify H e jw over 0  w  2
or    w  
The “low frequencies” are frequencies close
to zero or an even multiple of π.

The “high frequencies” are frequencies close


to ±π or an odd multiple of π.

More generally, modify the frequency with


2 r , r is integer.
11 ECEA108

Ideal Frequency-Selective Filters


periodicity

 
Frequency Response of Ideal Low-pass Filter

repeats periodically
one period with period 2πr outside
the plotted interval

 
12 ECEA108

Compiled by LDV 6
SY20212022 12/7/2021

Frequency Response of Ideal


High-pass Filter

 

 
Ideal Band-pass Filter
13 ECEA108

Frequency Response of Ideal


Band-stop Filter

 

periodicity

 
Ideal Low-pass Filter
14 ECEA108

Compiled by LDV 7
SY20212022 12/7/2021

Example 2.16 Frequency Response of


the Moving-Average System
solution:
 1
 ,  M1  n  M 2
hn   M 1  M 2  1
 0, otherwise
M2
H e  jw 1
 e jwn
M 1  M 2  1 n M
1
jwM1  jw M 2 1

1 e e
M1  M 2  1  jw
1 e
15 ECEA108

 jw M 2 1
H e  jw e jwM
1 e 1

M1 M 2 1
1  e jw
jw M1M21  jw M1M21
 jw M2M11

1 e 2 e 2
e 2
M1M21
1e  jw

jw M1M2 1  jw M1M21


 jw M2 M1

1 e 2 e 2
e 2
M1M21
e jw 2
e  jw 2

1 sin  w  M1  M 2  1 2  jw M2 M1


 e 2

16
M1M 2 1 sin  
w 2 ECEA108

Compiled by LDV 8
SY20212022 12/7/2021

Frequency Response of the Moving-


Average System
M1 5w
sin1w sin M22  e1 j22w  jw M2M1
H  e     M M 1
jw1
H jw
e 
M1M2 1 1 2 sin sinww 22
e 2

M1 = 0 and M2 = 4

4
5

17
 4
5
 ECEA108

2.6.2 Suddenly applied Complex


Exponential Inputs
In practice, we may not apply the complex
exponential inputs e jwn to a system, but the
more practical-appearing inputs of the form
x[n] = ejwn  u[n]
i.e., x[n] suddenly applied at an arbitrary time,
which for convenience we choose n=0.
For causal LTI system:
y  n   0, n  0; for x  n   0, n  0

18 ECEA108

Compiled by LDV 9
SY20212022 12/7/2021

2.6.2 Suddenly applied Complex


Exponential Inputs
xn  e jwnun For causal LTI system
 0, n0
y  n   x  n   h  n    n  jwk  jwn
  h  k  e e , n  0

 k 0   nk 0
  h  k  e  jw n  k u[ n  k ]   h  k  e  jw n  k u[ n  k ]
k  k 0 For n≥0
   jwk  jwn    jwk  jwn
   h k  e  e   h  k  e e
 k 0   k  n 1 
  
 H e jw
e jwn
   h  k  e  jwk  e jwn  yss  n   yt  n 
19
 k  n 1  ECEA108

2.6.2 Suddenly applied Complex


Exponential Inputs
Steady-state Response

yss  n   H  e jw
e jwn
  h  k  e  jwk e jwn
k 0

Transient response

yt n    hk e  jwk e jwn
k  n 1

20 ECEA108

Compiled by LDV 10
SY20212022 12/7/2021

2.6.2 Suddenly Applied Complex


Exponential Inputs (continue)
For infinite-duration impulse response (IIR)
  
yt  n    h k  e  jwk
e jwn
  h k    h k 
k  n 1 k  n 1 k 0

For stable system, transient response must
become increasingly smaller as n  ,
 

 h  k ' ne  k '  k   h  n  k  e
k ' n 1
jw ( n  k ')

k  1
jwk
h n  k 

n0
Illustration of a real part of suddenly applied
21 complex exponential Input with
ECEA108 IIR

2.6.2 Suddenly Applied Complex


Exponential Inputs (continue)
If h[n] = 0 except for 0 n  M (FIR), then the
transient response yt[n] = 0 for n+1 > M.
For n  M, only the steady-state response
exists

Illustration of a real part of suddenly applied


22 complex exponential Input with FIR
ECEA108

Compiled by LDV 11
SY20212022 12/7/2021

2.6.2 Suddenly Applied Complex


Exponential Inputs (continue)
The condition for stability is also a sufficient
condition for the existence of the frequency
response function, as
 jw  
 jwk
H  e    h k e
  k 

 jwk
  h k e
k 

  h k   
k 
23 ECEA108

2.7 Representation of Sequences


by Fourier Transforms
(Discrete-Time) Fourier Transform, DTFT,

analyzing  jwn
X  e    x  n  e
jw
  n

If xn is absolutely summable, i.e. x  n   


 
then X e jw exists. (Stability) n 

Inverse Fourier Transform, synthesis

  
1
xn 

X e jw e jwn dw
2 

24 ECEA108

Compiled by LDV 12
SY20212022 12/7/2021

Fourier Transform
       
X e jw  X R e jw  jX I e jw  X e jw e jX e 
jw

rectangular form polar form


 
X e jw : Fourier transform,
Fourier spectrum, spectrum
X  e jw  : magnitude, magnitude spectrum,
amplitude spectrum
 
X e jw : phase , phase spectrum
25 ECEA108

Principal Value
 
 X e is not unique because any 2 r
jw

 
may be added to X e jw without affecting
the result of the complex exponentiation.
 e 
j  X e jw
e

j  X ( e jw )  2  r

Principle value: X e  is restricted to the


jw

range of values between   and  . It is


 
denoted as ARG  X e jw 

 arg  X  e  : phase function is referred as a


jw

continuous function of w for 0  w  


26 ECEA108

Compiled by LDV 13
SY20212022 12/7/2021

Impulse response and


Frequency response
The frequency response of a LTI
system is the Fourier transform of the
impulse response of the LTI .

H e jw
  h  n  e  jwn
n 

  
1
hn 

H e jw e jwn dw
2 

27 ECEA108

Example 2.17: Absolute Summability


Let xn  a nun , find condition of existence of
the fourier transform or Absolute Summability.
Soltution:  
X e jw
  a e n  jwn
   ae  jw n

n 0 n 0

1 (ae jw)n+1 1
 lim  if ae jw  1
n  1  ae jw 1  ae  jw

or if a  1
1
a 
Absolute n
Summability  , if a  1
n0 1 a
28 ECEA108

Compiled by LDV 14
SY20212022 12/7/2021

Discussion of convergence

Absolute summability is a sufficient


condition for the existence of a Fourier
transform representation, and it also
guarantees uniform convergence.
Some sequences are not absolutely
summable, but are square summable,
i.e., 

   
2
x n
n 

29 ECEA108

Discussion of convergence

 x  n  
2

n 
Sequences which are square summable,
can be represented by a Fourier transform,
if we are willing to relax the condition of
uniform convergence of the infinite sum
 

defining . X  e jw    x  n e jwn
jw
Xe
  n 

Then we have Mean-square Convergence.


30 ECEA108

Compiled by LDV 15
SY20212022 12/7/2021

Discussion of convergence
Mean-square convergence

    xne     xne
 M
jw  jwn jw  jwn
Xe XM e
n   n M

lim
M  

   X e 
X e jw
M
jw 2
dw  0

   
The error X e jw  X M e jw may not approach
zero at each value of w as M   , but
total “energy” in the error does.
31 ECEA108

2.6.2 Suddenly applied Complex


Review Exponential Inputs
xn  e jwnun For causal LTI system
 0, n0

y  n   x  n   h  n    n  jwk  jwn
  h  k  e e , n  0
 k  0 
   jwk  jwn    jwk  jwn
   h k  e  e    h  k  e e For n≥0
 k 0   k  n 1 
yss  n  yt  n 

Illustration of a real part of suddenly applied


32 complex exponential Input with
ECEA108IIR

Compiled by LDV 16
SY20212022 12/7/2021

Impulse response and


Frequency response
The frequency response of a LTI
system is the Fourier transform of the
impulse response of the LTI .

H e jw
  h  n  e  jwn
n 

  
1
hn 

H e jw e jwn dw
2 

33 ECEA108

2.7 Representation of Sequences


Review by Fourier Transforms
(Discrete-Time) Fourier Transform, DTFT,

 jwn
analyzing X  e    x  n  e
jw
  n

 uniform convergence(Stability):  x n  


n  

Inverse Fourier Transform, synthesis


xn 
1 

2 
 
X e jw e jwn dw

Mean-square convergence:  x  n   
2

n 
34 ECEA108

Compiled by LDV 17
SY20212022 12/7/2021

Example 2.18 : Square-summability


for the ideal Lowpass Filter
 1, w  wc
 
H lp e   jw

Soltution:
0, wc  w  
1 wc jwn 1  jwn  wc
hlp  n    e dw  
e w
2  wc 2 jn c


1
2 jn
 e jwc n
e
 jwc n
  sinwn n ,
c
  n  

Since hlp n  is nonzero for n  0 , the ideal


lowpass filter is noncausal.
35 ECEA108

Example 2.18 Square-summability


for the ideal Lowpass Filter
sin wc n
hlp  n   approaches zero as n  ,
n
but only as 1n.
 hlp n is not absolutely summable.

sin wc n  jwn does not converge

n   n
e
uniformly for all w.

  
M
Define H M e jw sin wc n  jwn
e
n M n
36 ECEA108

Compiled by LDV 18
SY20212022 12/7/2021

Example 2.18 Square-summability


for the ideal Lowpass Filter
since 1 wc j n sin wc n
hlp  n   e
2  wc
d   ,   n  
n

   sin wn n e
M
HM e jw c  jwn

n  M
M
 1 wc 
  wc e d  e  j w n
j n

n  M  2  1 wc  M j  wn 
 e
2  wc  n  M d

1 wc sin[( 2M+1) ( w   ) / 2]

2  wc sin[( w   ) / 2] d
37 ECEA108

Gibbs Phenomenon
M
sin( wwc n ) / 2]
H M  e H
jw
 M  e   
1 wjwc sin[( 2M+1)
e jw nd
2  wc nsin[  M( w 
 n) / 2]

M=1 M=3

M=7 M=19

38 ECEA108

Compiled by LDV 19
SY20212022 12/7/2021

Example 2.18 continued


As M increases, oscillatory behavior at
w wc is more rapid, but the size of
the ripple does not decrease. (Gibbs
Phenomenon)

As M  , the maximum amplitude of


the oscillation does not approach zero,
but the oscillations converge in location
toward the point w w .
c
39 ECEA108

Example 2.18 continued



sin wc n  jwn
 e does not converge uniformly
n    n
to the discontinuous function H lp e
jw
.  
However, hlp n is square summable,
and H M e jw  converges in the mean-
square sense to H lp e jw 

   H e  dw  0
 2


jw jw
lim H lp e M
M  

40 ECEA108

Compiled by LDV 20

You might also like