Digital Signal Processing

DTFT
 Digital Signal

Continuous

Continuous
amplitude

amplitude
Analog Discrete-
Signal time Signal

Continuous Discrete
time time

amplitude
Discrete
Digital
Signal

Discrete
time

2
Discrete-Time Fourier Transform
– Introduction to DTFT
– DTFT
– Convolution Using the DTFT
Next lecture is on Z-Transform
– Introduction to z-Transform
– ROC of a Rational z-Transform
– Inverse z-Transform
– z-Transform Properties

3
•Three useful representations of discrete time
sequences in the transform domain:
1. Discrete-time Fourier Transform (DTFT)
2. Discrete Fourier Transform (DFT)
3. z-Transform

4
 Frequency Response
For a discrete-time LTI system, the frequency response
is defined as
j
j Y (e )
H (e ) 
X ( e j )
 Frequency response explained
 For a discrete-time LTI system, a complex
exponential input, the output is a complex
exponential sequence, with the same complex
frequency but with a different amplitude = H (e  j ).
The change in the amplitude is referred to as the
frequency response of the system.
 The expression for the frequency response in terms of
the unit sample response is :

The FR is the sum of h[n] (the unit sample response)

x H ( e  j )
7
•

8
• Definition: The discrete-time Fourier
transform (DTFT) X (e j ) of a sequence
x[n] is given by

X (e )   x[n].e
j  jn

n 
• In general, X (e j ) is a complex function of
the real variable ω and can be written as
j j j
X (e )  X re (e )  jX im (e ) 9
• X re (e j ) and X im (e j ) are, respectively,
j
the real and imaginary parts of X (e ) ,and
are real functions of w
• X (e j ) can alternately be expressed as
j j j (  )
X (e )  X (e ) .e
where
j
 ( ) arg{ X (e )}
10
• X (e j ) is called the magnitude function
•  ( ) is called the phase function
• Both quantities are again real functions of ω
• In many applications, the DTFT is called
the Fourier spectrum
j
• Likewise, X (e ) and  ( ) are called the
magnitude and phase spectra

11
 j
• For a real sequence x[n], X (e ) and X re (e j )
are even functions of w, whereas,  ( )
and are odd functions of w
• Note: X (e j )  X (e j ) .e j ( 2k )

 X (e j ) .e j ( )
for any integer k
The phase function  ( ) cannot be
uniquely specified for any DTFT
12
• Unless otherwise stated, we shall assume
that the phase function  ( ) is restricted to the
following range of values:
   ( )  
called the principal value

13
• The DTFTs of some sequences exhibit
discontinuities of 2p in their phase responses
• An alternate type of phase function that is a
continuous function of w is often used
• It is derived from the original phase function
by removing the discontinuities of 2p

14
• Example: The DTFT of the unit sample
sequence δ[n] is 1, given by
j   j n
 (e )    [ n].e ??????
n  
• Example: Consider the causal sequence

x[ n]  .[ n],   1
n

15
Its DTFT is given by
j   j n  j n
X (e )    .[n].en
   n .e
n   n 0
  j n 1
  ( .e )   j
n 0 1   .e
as e  j
  1

16
 j
• The DTFT X (e ) of a sequence x[n] is a
continuous function of w
• It is also a periodic function of ω with a
period 2p:

j (0 2 k )   j (0 2 k ) n

X (e )   x[ n].e
n  
  j0 n  j 2 kn  j0 n
  x[n].e .e   x[n].e   X (e j0 )
n   n  
17
 
• Therefore X (e )   x[n].e
j  jn

n 

represents the Fourier transform

• As a result, the Fourier coefficients x[n] can
be computed from using the Fourier
integral
1  j j n
x[n]   X (e ).e d
2   18
Convergence:
• An infinite series of the form

X (e )   x[n].e
j  jn

n  Adobe Acrobat

Document

may or may not converge

K
Let X ( e j ) 
K  x[
n  K
n ].e  jn

19
• Then for uniform convergence of X (e j ) ,
j j
lim X (e
K 
)  X K (e ) 0

• Now, if x[n] is an absolutely summable

sequence, i.e., if


 x[n]  
n 

20
• Then
 
j
X (e )   x[n]. e
n 
 jn
  x[n]  
n 

for all values of ω

• Thus, the absolute summability of x[n] is a
sufficient condition for the existence of the
DTFT X (e j )

21
• Example: The sequence x[n]  [n] for
n

  1 is absolutely summable as
 
1

n 
 [n]   
n

n 0
n

1 


and its DTFT X (e j ) therefore converges

to 1 /(1  e  j ) uniformly

22
• To represent a finite energy sequence x[n]
that is not absolutely summable by a DTFT
j
X (e )
j
X (consider
, it is necessary to e ) a mean-
square convergence
 of 2
:
lim X (e
j j
)  X K (e ) d 0
K   

K
where X K (e j )   x[n]. e
n  K
- jn

23
• Here, the total energy of the error
j j
X (e )  X K (e )
must approach zero at each value of ω as
K goes to 
• In such a case, the absolute value of the
error X (e j )  X (e j ) may not go to
K
zero as K goes to  and the DTFT is no
longer bounded 24
• There are a number of important properties
of the DTFT that are useful in signal
processing applications
• These are listed here without proof
• Their proofs are quite straight forward
• We illustrate the applications of some of the
DTFT properties

25
26
27
28
• Example: Determine the DTFT of

Let
We can therefore write

The DTFT of x[n] is given by

29
• Using the differentiation property of the
DTFT, we observe that the DTFT of nx[n]
is given by

• Next using the linearity property of the

DTFT we arrive at

30
• Example: Determine the DTFT of
the sequence v[n] defined by

The DTFT of is 1
Using the time-shifting property of the
DTFT we observe that the DTFT of
is and the DTFT of v[n − 1] is

31
• Using th linearity property we then obtain
the frequency-domain representation of
d0v[n]+ d1v[n −1] = p0δ[n]+ p1δ[n −1]
as
d0V (e jω) + d1e− jωV (e jω) = p0 + p1e− jω
• Solving the above equation we get

32
33