Lecture 8

Signal and Systems

Fourier Series
 Fourier series
Baron Jean Baptiste Joseph
Fourier (1768−1830) introduced the idea
that any periodic function can be
represented by a series of sines and cosines
which are harmonically related.

A Fourier series is a way of representing

a periodic function as a (possibly infinite)
sum of sine and cosine functions.
A Fourier series is a periodic function
composed of harmonically related sinusoids,
combined by a weighted summation.

Function s(x) (in red) is a sum of six sine functions of different

amplitudes and harmonically related frequencies
Periodic signals as a sum of complex exponential
signals.
A Fourier series is a way of representing
a periodic function as a (possibly infinite)
sum of sine and cosine functions.

1 j t 1  jt
cos(t )  e  e
2 2
1 j t 1  jt
sin(t )  e  e
2j 2j

So mathematically, we can represent any periodic function as sum of sine and cosine functions or
as sum of complex exponential
Periodic signals as a sum of complex exponential
signals.
1 j t 1  jt So mathematically, we can represent any periodic function as
cos(t )  e  e sum of sine and cosine functions or as sum of complex
2 2 exponential with ak being the complex co-efficient.
1 j t 1  jt
sin(t )  e  e 
2j 2j x(t )  ae
k 
k
jk  t

 1
cos(t )   ak e jkt   ak e jt  a1e jt  a0e j 0t  a1e  jt
k  k 1
1 𝑗 𝜔𝑡 1 − 𝑗𝜔 𝑡
.= 𝑒 +0+ 𝑒
2 2
1 1
.= 𝑒 𝑗 𝜔 𝑡 + 𝑒− 𝑗 𝜔 𝑡
2 2
1
.=𝑎 1=𝑎 −1= , otherwise 0
2
Periodic signals as a sum of complex exponential
signals.
 1
cos(t )   ak e jkt   ak e jt  a1e jt  a0e jt  a1e  jt
k  k 1

1 𝑗 𝜔𝑡 1 − 𝑗 𝜔𝑡 1 𝑗 𝜔𝑡 1 − 𝑗𝜔 𝑡
𝑒 + 0+ 𝑒 = 𝑒 + 𝑒
2 2 2 2
1
𝑎1 =𝑎 −1= , otherwise 0
2

1.5
Cosine signal in Frequency domain using complex exp: 1.5

1.0
1.0

0.5

1 j t 1  jt
cos(t )  e  e
0.5

0.0

-0.5
2 2 0.0

-0.5
-2 -1 0 1 2 K
 
Periodic signals as a sum of complex exponential
signals. B

x(t )  1  cos(2t ) 1

 a2e j 2t  a0e j 0t  a2e  j 2t

1 𝑗 2𝜔𝑡 1 − 𝑗2𝜔𝑡
𝑥 ( 𝑡 ) =1 + 𝑒 + 𝑒
2 2 0
-2 -1 0 1 2 K

a0 is average/mean/DC value of signal

1 j t 1  jt
cos(t )  e  e x(t )  1  2cos(2t )
2 2 j 2 t j 0 t  j 2 t
1 j t 1  jt  a 2 e  a0 e  a 2 e
sin(t )  e  e . =1 + 𝑒
𝑗2𝜔𝑡
+ 𝑒
− 𝑗 2𝜔𝑡
2j 2j
Advantages
Performing operations on sinusoidal or exponential signals is very easy.

Fourier representation can identify various frequency components into signal.

For some system, if it’s complex exponential signal response is known then output of system for any input signal
(represented in F.S. form) can be found easily.

Used to analyze LTI systems to sinusoidal/complex exponential input signals (El: ckts)

FS applicable to sea waves, radio waves as well.

Fourier series is used for CT and DT periodic signals. Power

and phase content.
Response of LTI system due to complex
exponential input signal Now for multiple exponential input signals, there are multiple
x(t )  e j t output signals

input x(t )   ak e jkt
 
k 

y (t )   x( )h(t   )d   h( ) x(t   )d
 
y (t )   ak H ( jk )e jk  t

k 
 
H(jw) is the frequency response of the system
 h( ) e  h( ) e
j ( t   ) j ( t   )
y (t )  d  d ∞
𝐻 ( 𝑗 w )= ∫ h ( 𝜏 ) 𝑒
− 𝑗𝜔𝜏
  𝑑𝜏
e j t −∞
 are known as eigen functions of system
y (t )  e j t  h( ) e j d  H ( j ) e jt
Response of an LTI system due to complex exponential signal is
 same complex exponential signal with change in magnitude by
a complex factor H(jw).
Example to find H(jw) and then output of system (method 01)
y (t )  x(t  2) system
h(t )   (t  2) impulse response
 

H ( jw)   h( ) e d    (  2) e d
  jw  jw
Frequency response
 

H ( jw)   (2  2) e  jw2
e  jw2

 3
j t j t
x(t )  2 e 4
e 4
for different exponential input signal, output will be
 3
 j t 3 j t
y (t )  H ( jw) e jwt
 2H ( j )e 4
 H ( j )e 4
4 4
  3 3   3  3
j j t j t j t j( t ) j( t )
y (t )  2e 2
e 4
e 2
e 4
 2e 4 2
e 4 2
Slide # 10
Example to find H(jw) and then output of system (method 02)
• y (t )  x(t  2) system
• 
j t j
3
t input
• x(t )  2 e 4
e 4

 3
• j ( t  2) j ( t  2) output
y (t )  2 e 4
e 4

• 3rd Method is convolution (Do it yourself)

• Output will be same
• Keep h(t) first (integration will be solved easily)

Slide # 11
Example
y (t )  x(t  3) ∞
𝐻 ( 𝑗 w )= ∫ h ( 𝜏 ) 𝑒 So now if you have any x(t) then just multiply H(jw) to find y(t).
− 𝑗𝜔𝜏
𝑑𝜏
h(t )   (t  3) −∞  3
j t j t
 x(t )  2 e 4
e 4


 j
H ( j )  h (  3) e d  3
j t j t

 y (t )  H ( j )[2 e 4
e 4
]

 j
H ( j )   (  3) e d 
 j t 3 j 34 t

Frequency response H(jw) only exist for t=3 because impulse  2H ( j )e 4
 H ( j )e
exist at this value, while rest of the values from negative infinity 4 4
to infinity will be zero, therefore   3 3
j 3 j t j 3 j t

H ( j )   (3  3)e  j 3
y (t )  2e 4
e 4
e 4
e 4

 3 3 9
j ( t ) j( t )
H ( j )  1.e  j 3  2e 4 4
e 4 4
 ∞
𝐻 ( 𝑗 w )= ∫ h ( 𝜏 ) 𝑒
− 𝑗𝜔𝜏
𝑑𝜏
Example −∞

j(

t
3
) j(
3
t
9
) y (t )  x(t  3)
y (t )  2e 4 4
e 4 4

j t j
3
t
x(t )  2 e 4
e 4

 3
 3 j ( t  3) j ( t  3)

y (t )  2e
j ( t  3)
4
e
j
4
( t  3)
y (t )  2 e 4
e 4
 ∞
𝐻 ( 𝑗 w )= ∫ h ( 𝜏 ) 𝑒
− 𝑗𝜔𝜏
𝑑𝜏
Example −∞

y (t )  x(t  3) So now if you have any x(t) then just multiply H(jw) to find y(t).
h(t )   (t  3)
 x(t )  cos(4t )  cos(7t )

 j
H ( j )  h (  3) e d

y (t )  H ( j )[cos(4t )  cos(7t )]

1 j 4t 1 j 7t
H ( j )    (  3)e d  j
y (t )  H ( j )[ (e  e )  (e  e )]  j 4t  j 7t

 2 2
Frequency response H(jw) only exist for
 1 1
t=3 because impulse exist at this value, y (t )  e [ (e  e )  (e j 7 t  e  j 7 t )]
 j 3 j 4 t  j 4 t
while rest of the values from negative
infinity to infinity will be zero, therefore
2 2
 j 3 1  j 4t  j 3 1  j 7t
H ( j )   (3  3)e  j 3
y (t )  e (e  e )  e
j 4t
(e  e )
j 7t

 j 3 2 2
H ( j )  1.e 1  j12 j 4t 1 j12  j 4t 1  j 21 j 7 t 1 j 21  j 7 t
y (t )  e e  e e  e e  e e
2 2 2 2
 1 j 4( t  1  j 4( t  1 j 7( t  1  j 7( t 
y (t )  e 3)
 e 3)
 e 3)
 e 3)

2 2 2 2
1 j t 1  jt
cos(t )  e  e
So now x(t) as below then just multiply H(jw) to find y(t). 2 2
x(t )  sin(4t )  sin(7t ) 1 j t 1  jt
sin(t )  e  e
2j 2j

1 j 4( t  1  j 4( t  1 j 7( t  1  j 7( t 
y (t )  e 3)
 e 3)
 e 3)
 e 3)

2j 2j 2j 2j
Determine periodic signal which is expressed in complex exponential Fourier Series

16
 
x(t )  
k 
ak e jk0 t
T
 T  2
0 x(t )e dt  k ak   e e
 jn0 t jk 0 t  jn0 t
dt  , T
 0  
T   T

0 x(t )e dt  k ak   e
 jn0 t j0 t ( k  n )
dt 
0 
Fourier

series
T


T T

Coefficients
 e dt    cos(k  n) t dt  j  sin(k  n) t dt
j0 t ( k  n )
0 0
 0  0 0

 T j0 t ( k  n )  T k  n
 e dt   
0 k  n
0 
T T
1
 ak   x(t )e  jn0 t dt
 jn0 t
x (t )e dt  akT T
T0 1
0 a0   x(t )dt
T0
Determining Fourier series coefficients
• Multiply both sides with exponential signal

Slide # 20
Determining Fourier series coefficients

 T j0 t ( k  n )  T k  n 
 e dt    
0   0 k  n

Slide # 21
Determining Fourier series coefficients

a0 is average/mean/DC value of signal

Slide # 22
Determining Fourier series coefficients- Another approach

Slide # 23
Find F.S coefficients, their magnitude and phase spectrum

Slide # 24
• Magnitude
• Square root of (real part square + imaginary part square)

1.11
1

0.5

Slide # 25
• Phase
• Tangent inverse (value of imaginary part/value of real part)

450

26.560

00

26.560
Note: Plots of the amplitudes and45phases of the F.S. coefficients (harmonics)
0

versus frequencies (k) are considered as frequency spectrum.

Individually for amplitude and phase, it can be called as amplitude and phase
spectrum respectively.
Slide # 26
Find and plot coefficients for T=4T1, T=8T1, T=16T1
• Signal is periodic and will keep repeating same as x(t)

Slide # 27
Slide # 28
• Infinite frequency components

Slide # 29
• (a) T=4T1

• (b) T=8T1

• (c) T=16T1

• If T is same, pulse duration is decreasing

• Frequency components are increasing, and amplitude of such components is
Slide # 30
Slide # 31
Taken from mx.nthu.edu.tw/

Slide # 32
Taken from mx.nthu.edu.tw/

Slide # 33
Taken from mx.nthu.edu.tw/

Slide # 34
Taken from mx.nthu.edu.tw/

Slide # 35
Taken from mx.nthu.edu.tw/

Slide # 36