Fourier Analysis

Instructor: Mohammed Farrag

Dept. of Electrical Engineering
Assiut University

[email protected]
Course Outline

• Time domain analysis

• Signals and systems in continuous and discrete time

11/17/2013
• Convolution: finding system response in time domain
• Frequency domain analysis
• Fourier series

Signal Processing
• Fourier transforms
• Frequency responses of systems
• Generalized frequency domain analysis
• Laplace and z transforms of signals
• Tests for system stability
• Transfer functions of linear time-invariant systems

11 -
2
 Periodic Signals
• Complex Exponential Fourier Series

 T0 2

f t   f t  e  j 2mf 0t dt
1
 cm e j 2mf 0t
 cm  

11/17/2013
m   T0 T0 2

Signal Processing
• For all t, f(t + T) = f(t)
• f(t) is a periodic signal
• Smallest value of T is the fundamental period
• Fundamental frequency 1/T

• Periodic signals have a Fourier series representation

• Fourier series coefficient cm quantifies the strength of the component
of f(t) at frequency m/T
12 - 3
• Fourier transforms are for both periodic and aperiodic signals
 Periodic Signals
• Complex Exponential Fourier Series

 T0 2

f t   f t  e  j 2mf 0t dt
1
 cm e j 2mf 0t
 cm  

11/17/2013
m   T0 T0 2

Signal Processing
• For any LTI system h(t), the output y(t) is

y (t )  T [ f t ]   cm T[e j m0t
]
m  


y (t )  T [ f t ]  c m e j m0t
H (m0 )
m  
where
12 - 4



 j t
H ( )  h (t ) e dt and   2f 0
t  
Trigonometric and Harmonic Fourier Series

• General representation f t   a0   am cosm0t   bm sin m0t 
of a periodic signal m 1

11/17/2013
 f t dt  c
1 T0
a0  0
T0 0

 f t cosm t dt  c
2 T0
am   c m

Signal Processing
FS coefficients T0 0
0 m

 f t sin m t dt  j (c

2 T0
bm  0 m  c m )
T0 0


f t   d 0   d m cosm0t   m 
• Harmonic Form Fourier series m 1

where d 0  a0 , d m  am2  bm2 , and 11 -

 bm  5
 m  tan 1  
 am 
Existence of the Fourier Series
• A periodic signal f(t) has a Fourier series representation if it is
absolutely integrable over any period

11/17/2013
Proof: FS representation exists if,

Signal Processing
T0 2

f t  e  j 2mf 0 t dt
1
cm   where cm 
T0 
T0 2

T0 2

f t  e  j 2mf 0 t dt  
1
cm 
T0 
T0 2

T0 2 11 -
 f t  dt  
T0 2
6
Power content of a periodic signal

• The average power of a periodic signal f( t ) over any period is

T0 2

11/17/2013
 f t 
1
P
2
dt
T0 T0 2

• If f( t) is represented by the complex exponential Fourier series,

Signal Processing
then it can be
T0 2 
f t  dt 
1
P   cm
2 2

T0 T0 2 m  

• This equation is called Parseval's identity (or Parseval's theorem)

for the Fourier series.
12 -
7
  2 At  0.5  t  0.5
Example #1 f (t )  
2 A  2 At 0.5  t  1.5


f t   a0   an cos(π n t )  bn sin π n t 
n 1

11/17/2013
f(t)
a0  0 (by inspection of the plot)
A an  0 (because it is odd symmetric)

Signal Processing
1/ 2

1 0 1 bn   2 A t sin(π n t ) dt 
1 / 2
-A
3/ 2

 (2 A  2 A t ) sin(π n t ) dt
1/ 2

• Fundamental period 
 0
• T0 = 2  8 A
n is even
• Fundamental frequency bn   2 2 n  1,5,9,13,  11 -
 n 8
• f0 = 1/T0 = 1/2 Hz  8 A n  3,7,11,15, 
• 0 = 2/T0 =  rad/s  n 2 2
Fourier Analysis

Instructor: Mohammed Farrag

Dept. of Electrical Engineering
Assiut University

[email protected]
 From Fourier Series to Fourier Transform:

• Let x(t) be a nonperiodic signal of finite

duration, that is,

x(t )  0 t  T1

11/17/2013
• Let xT (t ) be a periodic signal formed by

Signal Processing
0
repeating x(t) with period T0.
• If we let T0  ∞, we have

lim xT0 (t )  x(t )

T0 

• The complex exponential FS of xT0 (t ) is given

 T0 2 12 -
xT0 t    x t  e
1
c
m  
m e jm0t
where cm 
T0
T0
 jm0t
dt 10
T0 2
From Fourier Series to Fourier Transform:

• Since xT0 (t )  x(t ) for t  T0 2 and also since x(t )  0 outside

this interval,

11/17/2013
T0 2 
cm 
1
 x t  e  jm0t
dt 
1
 x t  e  jm0t
dt
T0 T0 2
T0 

Signal Processing
• Let us define X ( ) as

X ( )   x t  e  j t
dt


• Then

12 -
11
From Fourier Series to Fourier Transform:

11/17/2013
• Let

Signal Processing
• Then

• The sum on the right-hand side can be viewed as the area under the
function X ( )e jt . Therefore, we obtain
12 -
12
Fourier Transform pair

 
X     xt  e dt  xt   X   e j  t d
1

 jt
 2 

11/17/2013
• Conditions for Fourier transform of x(t) to exist
• x(t) is absolutely integrable

Signal Processing
X ( )  


X ( )   xt  e  jt dt  



 xt  dt  

12 - 13
Fourier Transform

• What system properties does it possess?

 Memoryless (in fact requires infinite memory)

11/17/2013
 Causal
 Linear

Signal Processing
 Time-invariant (doesn’t apply)

• What does it tell you about a signal?

• Answer: Measures frequency content

• What doesn’t it tell you about a signal?

• Answer: When those frequencies occurred in time

12 - 14
Fourier Transform Pairs
  t   jt  /2
F     rect e dt   e  jt dt

   / 2

11/17/2013
     
sin   sin  
2    

j
e 
1  j / 2
 e j / 2  2 

2 
 
  
  sinc


   2 

Signal Processing
 2 

f(t) F() 
1 F

t
-/2 0 /2 6 4 2  0 2 4 6
     
12 - 15
 Fourier Transform Pairs
• From the sifting property of the Dirac delta,

F d t    d t  e  j  t dt  e  j 0  1


11/17/2013
• Consider a Dirac delta in the Fourier domain

F d     d   e
1 jt 1 j 0t 1
1
d  e 

Signal Processing
2  2 2
• Using linearity property, F{ 1 } = 2 d()

x(t) = 1 X() = 2  d()

1 F
(2)

t  12 - 16
0 0

(2) means that the area under the Dirac delta is (2)
Fourier Transform Pairs
1 
F d   0     1 j 0 t

j t
1
d    e d   e
2 2
0
 

e  d   0  or e j0t  2 d   0 
1 j 0 t

11/17/2013
2

Since cos0t   e j0t  e  j0t
1


Signal Processing
2
cos0t    d   0   d   0 

f(t) F()
() ()
F
t  12 - 17
0 0 0 0
Properties of the Fourier Transform

11/17/2013
Signal Processing
12 -
18
Common Fourier Transforms Pairs

11/17/2013
Signal Processing
12 -
19
 Thanks

Digital Signal Processing 20

013
11/17/2