Frequency Domain

Spectrum and Bandwidth

Time domain concepts: viewed as a function of

time
• Can be analog or digital
• analog signal
– varies in a smooth way over time
• digital signal
– maintains a constant level then changes to another constant level
• periodic signal
– pattern repeated over time
• aperiodic signal
– pattern not repeated over time
we now consider signals/waveforms that are
composed of multiple sinusoids having different
amplitudes, frequencies, and phases
 Sine Wave - 2

A sine wave can be represented as:

s(t) = A sin(2ft + ) t
where:
A – peak amplitude
f – frequency
t – time
 - phase angle
 - phase = Degrees Radians
00 =0 00 = 
450 = /4
900 = /2
1350 = 3/4
1800 =
 Varying Sine Waves
s(t) = A sin(2ft + )

(a) A=1, f =1, ø = 00 (b) A=0.5, f =1, ø = 00

(c) A=1, f =2, ø = 00

(d) A=1, f =1, ø = Π/4 = 450

Wavelength ()
is distance occupied by one cycle
between two points of corresponding phase in
two consecutive cycles
assuming that the signal velocity is vs
then, we have:  = vsT
or
 = vs / f (since T = 1/ f)
 and so, vs = f 
 especially when vs = c
c = 3*108 ms-1 (speed of light in free space)
 Jean Baptiste Joseph Fourier (1768-1830)
had crazy idea (1807):
Any periodic function
can be rewritten as a
weighted sum of sines
and cosines of different
frequencies.
Don’t believe it?
• Neither did Lagrange,
Laplace, Poisson and
other big wigs
• Not translated into
English until 1878!
But it’s true!
• called Fourier Series
 Frequency Domain Concepts

signal are made up of many frequencies

components are sine waves
Fourier Analysis can shown that any signal is
made up of component sine waves
can plot frequency domain functions

Fourier Analysis:
The complex wave at the top can
be decomposed into the sum of
the three simple sine waves
shown here.
 Addition of Frequency
Components (T=1/f)
graph c is sum of graphs a & b

S(t) = A * sin [ (2) f t ]

(a) = 1*sin[ (2) f t ]

(b) = 1/3 * sin[ (2) (3f) t ]

(c) = (4/) * { 1*sin[ (2) f t ]

+ 1/3 * sin[ (2) (3f) t ] }

The scaling factor of (4/) is used to produce

the cumulative wave whose maximum
peak amplitude is close to 1.
 Frequency Domain
Representations
Frequency domain function of Fig 3.4c t

S(f)

f
S(t) t
 Frequency Domain Representations

• Frequency domain function

of single square pulse

S(t) S(f)

1 Fourier transform

- X/2 +X/2 t

S(t) = 1 for time period -X/2 to +X/2 S(f) = 1 X sin( f X)

= 0 elsewhere Fourier transform -----------------
fX
reference Page-839 William Stallings
Frequency Domain Representations
(Java Demo)
Fourier series representation of any given
periodic signals
• Can be represented as a sum of sine and cosine waveforms
known as Fourier series.
• reference Page-836 William Stallings

Fourier Series Applets in Java.

• http://www.falstad.com/fourier/index.html
• http://www.falstad.com/fourier/
Spectrum & Bandwidth
spectrum
• range of all frequencies contained in signal

absolute bandwidth
• width of spectrum

effective bandwidth
• often just bandwidth
• narrow band of frequencies containing most energy

DC Component
• component of zero frequency
 Effect of noise

Signal

Noise
Logic
Threshold Signal+Noise

Sampling times
0 1 1 1 1 0 0 0 0 1 Data Received
0 1 0 1 1 0 0 1 0 1 Original data

Bit error
24
A sum of sines
Our building block:
Asin(x   
Add enough of them to get
any signal f(x) you want!

How many degrees of

freedom?

What does each control?

Which one encodes the

coarse vs. fine structure of
the signal?
Fourier Transform
We want to understand the frequency of our signal. So,
let’s reparametrize the signal by  instead of x:

f(x) Fourier F()

Transform

For every  from 0 to inf, F() holds the amplitude A

and phase of the corresponding sine Asin(x   
• How can F hold both? Complex number trick!

F ( )  R( )  iI ( )
 1 I ( )
A  R( )  I ( )
2 2
 tan
R( )
We can always go back:

F() Inverse Fourier f(x)

Transform
Time and Frequency
example : g(t) = sin(2pf t) + (1/3)sin(2p(3f) t)
Time and Frequency
example : g(t) = sin(2pf t) + (1/3)sin(2p(3f) t)

= +
Frequency Spectra
example : g(t) = sin(2pf t) + (1/3)sin(2p(3f) t)

= +
Frequency Spectra
Usually, frequency is more interesting than the phase
Frequency Spectra

= +

=
Frequency Spectra

= +

=
Frequency Spectra

= +

=
Frequency Spectra

= +

=
Frequency Spectra

= +

=
Frequency Spectra
 The Convolution Theorem
The greatest thing since sliced (banana) bread!

• The Fourier transform of the convolution of two

functions is the product of their Fourier transforms
F[ g  h] F[ g ] F[h]
• The inverse Fourier transform of the product of two
Fourier transforms is the convolution of the two
inverse Fourier transforms
1 1 1
F [ gh] F [ g ]  F [h]
• Convolution in spatial domain is equivalent to
multiplication in frequency domain!
Power Spectral Density & Bandwidth
Absolute bandwidth of any time-limited signal is
infinite.
Most power in a signal is concentrated in finite
band.
Effective bandwidth is the spectrum portion
containing most of the power.
Power spectral density (PSD) describes power
content of a signal as a function of frequency.
 DFT

1/M
Example 2
Sampling
Fourier Transform (1-D)


A
Fu   f x exp j2ux  dx  sin uX e  juX

u
sin uX 
 Fu  AX
uX 

55
Discrete
Fourier
Transform
(Properties)

Sampling 1-D
The Fourier transform and the convolution theorem provide
the tools for a deeper analytic study of sampling problem. In
particular, we want to look at the question of how many samples
should be taken so that no information is lost in the sampling
process. Expressed differently, the problem is one of the
establishing the sampling conditions under which a continuous
image can be recovered fully from a set of sampled values. We
begin the analysis with the 1-D case.
As a result, a function which is band-limited in frequency
domain must extend from negative infinity to positive infinity in time
domain (or x domain).
60
 Discrete Fourier
Transform
Sampling
1-D
(Properties)

f(x) : a given function

F(u): Fourier
Transform of f(x)
which is band-limited
s(x) : sampling
function
S(u): Fourier
Recovered f(x) from sampled data
Transform of s(x)
G(u): window for
recovery of the main
function F(u) and f(x).
∆X ≤1/2w 61
Advanced 2-D
Home work2,
For the signal x(n)={4, 2, 6, 6, 1} for n=0,1,2,3,4
Find :
- the FT X(u)
- The Fourier spectrum .
END
Frequency Spectra


1
= A sin(2 kt )
k 1 k
 Analog Signaling
represented by sine waves
amplitude (volts)

1 cycle

phase
difference
time
(sec)

frequency (hertz)
= cycles per second
68
 Digital Signaling
represented by square waves or pulses
amplitude (volts)

1 cycle

time
(sec)

frequency (hertz)
= cycles per second
69
Continuous & Discrete Signals

70
Periodic Signals

71
Fourier Series
Any periodic signal can be represented as sum
of sinusoids, known as Fourier Series


A0
x(t )     An cos( 2nf 0t )  Bn sin( 2nf 0t )
2 n 1
T T
2 2
A0  
T 0
x(t )dt An  
T 0
x(t ) cos(2nf 0t )dt

If A0 is not 0, 2
T

x(t) has a DC Bn  x(t ) sin( 2nf 0t )dt

T 0
component
Fourier Series
Amplitude-phase representation


C0
x(t )    Cn cos(2nf 0t   n )
2 n 1

C0  A0 Cn  An2  Bn2

  Bn 
 n tan 
1

 An 
 Fourier Series Representation of
Periodic Signals - Example
x(t)
1

-3/2 -1 -1/2 1/2 1 3/2 2

-1
T

Note that x(-t)=x(t) => x(t) is an even function

T 2 1 1/ 2 1
2 2
A0  x(t )dt  x(t )dt 2 x(t )dt 2 1dt  2  1dt 1  1 0
T 0 20 0 0 1/ 2
 Fourier Series Representation of
Periodic Signals - Example
T T /2 1
2 4
An  x(t ) cos( 2nf 0t )dt  x(t ) cos( 2nf t )dt 2x(t ) cos( 2nf t )dt
0 0
T0 T 0 0

1/ 2 1
4 n
2 cos(2nf 0t )dt  2  cos(2nf 0t )dt  sin
0 1/ 2
n 2
T T /2
2 2
Bn  x(t ) sin( 2nf 0t )dt  x(t ) sin( 2nf 0t )dt
T0 T  T /2
0 T /2
2 2
 x(t ) sin(2nf 0t )dt  x(t ) sin(2nf t )dt
0 Replacing t by –t
T  T /2 T 0 in the first integral
2
T /2
2
T /2
sin(-2nf t)=

T 
0
x (  t ) sin( 2nf 0 t ) dt 
T x(t ) sin(2nf 0t )dt
0 - sin(2nf t)
 Fourier Series Representation of
Periodic Signals - Example

Since x(-t)=x(t) as x(t) is an even function, then

Bn = 0 for n=1, 2, 3, …

A0
x(t )     An cos( 2nf 0t )  Bn sin( 2nf 0t )
2 n 1

4 n
x(t )  sin cos nt
n 1 n 2
4 4 4 4
x(t )  cos t  cos 3t  cos 5t  cos 7t
 3 5 7
4 1 1 1 
x(t )   cos t  cos 3t  cos 5t  cos 7t 
 3 5 7 
 Another Example

x1(t)
1

-2 -1 1 2

-1
T

Note that x1(-t)= -x1(t) => x(t) is an odd function

Also, x1(t)=x(t-1/2)
4  1 1  1 1  1 1  1 
x1(t )   cos   t    cos 3 t    cos 5 t    cos 7 t  
  2 3  2 5  2 7  2 
 Another Example

4   1  3  1  5  1  7  
x1(t )   cos  t    cos  3t    cos  5t    cos  7t   
  2 3  2 5  2 7  2 

4 1 1 1 
x1(t )   sint  sin 3t  sin 5t  sin 7t 
 3 5 7 

   3 
cos  t   sin t cos  3t    sin 3t
 2  2

 5   7 
cos  5t   sin 5t cos  7t    sin 7t
 2  2
Fourier Transform
For a periodic signal, spectrum consists of
discrete frequency components at
fundamental frequency & its harmonics.
For an aperiodic signal, spectrum consists of a
continuum of frequencies.
• Spectrum can be defined by Fourier transform
• For a signal x(t) with spectrum X(f), the following relations
hold

 
x(t )  X ( f ) e j 2ft
df X ( f )  x(t ) e  j 2ft
dt
 
Fourier Transform Example

x(t)
A

 2  2


X ( f )  x (t ) e  j 2ft
dt

 /2
 j 2ft A  j 2ft  / 2
X(f ) 



/2
Ae dt 
j 2f
e
 /2
Fourier Transform Example

2A  e j 2f / 2  e  j 2f / 2  2 A  2f   sin( 2f / 2) 

      
2f  2j  2f  2   2f / 2 

sin( 2f / 2) sin(f )

X ( f )  A  A
2f / 2 f

 e j  e  j   e j  e  j 
sin    cos  
 2 j   2 
Signal Power
A function x(t) specifies a signal in terms of
either voltage or current
Instantaneous power of a signal is related to
2
Average power of a time limited signal is x(t )

2
x(t )
t2
1 2
t1  t 2  x (t ) dt
For a periodic signal,t1the average power in one
period is
T
1 2
T 
0
x (t ) dt
Power Spectral Density & Bandwidth
For a continuous valued function S(f), power
contained in a band of frequencies f1<f<f2

f2

P 2 S ( f ) df
f1

For a periodic waveform, the power through the

first j harmonics is
j
1
P C02   Cn2
2 n 1
Power Spectral Density & Bandwidth -
Example
Consider the following signal

 1 1 1 
x(t )  sint  sin 3t  sin 5t  sin 7t 
 3 5 7 
The signal power is

1 1 1 1 
Power  1     0.586 watt
2 9 25 49 
end