9/22/2021

Conjugate Symmetry of X(ejω) for Real Properties of DTFT (I)

x[n]
❑ For real x[n]; X(e-jω) =X*(ejω)
❑ If ( )
g n = G e j and hn = H e j ( )
( )  xne
X e − j =

j n then g[n] + h[n]  G (e j ) + H (e j )
n = −

❑ But X(ejω) is generally complex;   

( ) ( )
X e j = X e j e jX (e )
j
{gn + hn}e
n = −
− jn
=   g ne − jn +   g ne − jn
n =1 n =1
❑ Due to conjugate symmetry of X(e-jω), it follows that
( ) ( )
X e j = X e − j
X (e ) = −X (e )
j − j

i.e. the amplitude spectrum is an even function of ω and the

= G(e j ) + H (e j )
phase is an odd function of ω.

1 2

Properties of DTFT (II) Properties of DTFT (III)

❑ If xn is real
❑ Time reversal: leads to frequency reversal in the DTFT

then ( )
X e j = X * e − j ( ) If g [ n ]  G e j ( )
Then g[ − n]  G (e ) − j

❑ If xn is real Conjugate-

F {g − n} =


 g− ne − jn

then X (e ) = − X (e )
Antisymmetric Function
j * − j n = −

=  g− ne
n = −
− j ( − n )

= G(e − j )

3 4

Properties of DTFT (IV) Properties of DTFT (V)

❑ Multiplication by n: Frequency Differentiation
❑ Time-Shifting Property → Phase Change (Similar to CTFT)
If ( )
g [ n ]  G e j
If g [ n ]  G e j ( )
dG (e ) j
Then ng n   j Then g[n − nd ]  G (e )e j − jnd 
, nd  I
d Delaying a signal by nd units does not change its amplitude spectrum,

F {ng n} =  ngne − jn
but the phase spectrum is changed by –ndω. i.e. phase is a linear
n = − function of ω with slope –nd.
1  de − jn 
= 
− j n =−
g n
d
F [ g (n − nd )] =  gn − n e
n = −
d
− jn

= j
d 
 gne− jn = j d G(e j )
d =  gne − j ( n + nd )

d n=− n = −

❑ Can’t take derivative w.r.t. n in DT, but can take derivative w.r.t. = e − jnd  gne − jn
= e − jnd G(e j )
ω in the frequency domain. n = −

5 6

1
 9/22/2021

Properties of DTFT (VI) Properties of DTFT (VII)

❑ Frequency-Shifting (Modulation) Property → Signal Modulation
❑ Time and Frequency Convolution Property (Similar to CTFT)
(Similar to CTFT)
Multiplying a sequence by a complex exponential correspond to If g[n]  G e j ( ) and h[n]  H e j ( )
shifting its DTFT in the frequency domain.
If g [ n ]  G e j ( ) Then
g n* hn  G e j H e j ( ) ( )
G (e ( ) )
j s n j  − s ‘delay’
Then g[n]e  in
and
g nhn 
1
G e j * H e j ( ) ( )
frequency 2
This property is the dual of the time-shifting property. where 
 g n* hn =  g mhn − m
F {e j s n
g n} = e j s n
g ne − jn
and m = −
n = −
 ( ) ( )
G e j * H e j =  G (u )* H e j ( −u ) du ( )
=  gne − j ( − s ) n 2

n = − ❑ Since n is discrete while ω is continuous, there is no dulity

property for DTFT
= G(e j ( −s ) )

7 8

Parseval’s Theorem
❑ Parseval’s Theorm: relates total energy in a sequence to its DTFT.

❑ If x[n]  X e j ( )
2

 X (e ) d


 x(n )
1  2
Then Ex = = j

n = − 2 2

❑ Proof:
2

 x n 2   X (e  )e  d 
 
 1 
Ex =  x(n )
n = −
=
n = −
*
2
j j n

  
  X (e )  x ne

1 
=  d
j * j n

2 2
n = − 

  X (e )X (e ) ( )
2
1 1
2 2
 
= j * j
d = X e j d
2 2

9 10

Example
nxn = j
( )
dX e j
=j
d  1 
 =
 e − j
− j 
❑ Determine the DTFT of y[n]
d d  1 − e  1 − e − j ( ) 2

yn = (n + 1) un,   1 n

yn = n nun +  nun,   1  1   e 

− j
( )
Y e j =  +
− j   =
1
 1 − e   1 − e − j ( )
2
(
 1 − e − j )
2

x[n]

yn = nxn + xn

Apply time-
differentiation
( )
X e j =
1
1 −  e − j
property of DTFT
11 12

2
 9/22/2021

Alternatively, one could use the given z-transform pair

13 14