ADVANCED SIGNAL

PROCESSING
Dinko Begušić

[email protected]
Textbook:

Signal Processing for Communications

by Paolo Prandoni and Martin Vetterli

available at:
https://www.sp4comm.org/getit.html
 iBook
 paper version
 free online version

https://www.sp4comm.org/download.html
 HTML version
 pdf version
SIGNAL PROCESSING FOR COMMUNICATIONS

Chapter 1 What Is Digital Signal Processing?

Chapter 2 Discrete-Time Signals

Chapter 3 Signals and Hilbert Spaces

Chapter 4 Fourier Analysis

Chapter 5 Discrete-Time Filters

Chapter 6 The Z-Transform

Chapter 7 Filter Design

FREQUENCY IN THE CONTINUOUS AND DISCRETE TIME DOMAIN

Analog sinusiodal signal:

x a (t )  A cos(  t   ),  t 

A – amplitude (depends on the physical quantity),

Ω – angular frequency [rad/s],
θ – phase [rad].
Relationship between frequency variables Ω and f [Hz]:
  2f
Sinusoidal signal in continuous time domain
The properties of the analog sinusoidal signal:

1. For every fixed value of the frequency f, the signal x(t) is periodic with the period T=l/f,

2. Continuous-time sinusoidal signals with different frequecies are themselves distinct.

3. Increasing the frequency f results in an increase of the rate of oscillation (more periods are
included in a given time period).

The same properties apply to complex exponentials:

j ( t   )
x a (t )  Ae  A cos(  t   )  jA sin(  t   )
Sinusoidal signal in the discrete time domain:

x ( n)  A cos( n  ),   n  

n - index, sample number (integer variable),

A – amplitude (depends on the physical quantity),
ω – angular frequency (rad/sample),
θ - phase (rad).

The relationship between the frequency variable ω and relative frequency F (Hz
s/sample):

  2 F
The properties of the discrete-time sinusoidal signals:

1. The signal x(n) is periodic with the fundamental period N only if its frequency F is a
rational number F = k/N.

cos 2f  N  n      cos 2fn   

k
2fN  2k  Ff 
N
2. Discrete-time sinusoids whose frequencies are separated by an integer multiple of 2π
are identical (the frequencies are aliased).

cos   2k n     cos  n  2kn     cos  n   

3. The highest rate of oscillation in a discrete-time sinusiod is attained when:
1
   ili F 
2

The frequency range for discrete-time sinusoids is finite with duration 2π:

0    2 ili   
1 1
0  F 1  F 
2 2

It is called the fundamental range.

The properties of discrete-time sinusoids apply for the discrete-time complex exponentials.
An important consequenca is that there are only N different periodic complex exponentials
which are in a harmonic relationship with the same fundamental period N.

s k ( n )  e j 2 kn / N k  0,1, 2,..., N  1

s k  N ( n )  e j 2n ( k  N ) / N  e
j 2 n
 s k (n)  s k (n)
1
SAMPLING OF ANALOG SIGNALS

The periodic or uniform sampling with the sampling period T = 1/Fs ( Fs is a sampling rate in
samples per second or sampling frequency in hertz):

x a t   x n 
tn f  F
n
t  nT  
FS
Sampling of the analog sinusoidal signal with the sampling rate FS (samples/s):

x a t   A cos 2 ft   

 f 
xa nT   x n   A cos 2 fnT     A cos  2 n   
 FS 

f 1
F     T  2 f
FS FS
 F ω

Relationship between frequency variables

If the range of the frequency variables in the discrete time f and ω is limited to:

1 1
  f 
2 2
    

Then the range of the frequency variables in the continuous time F and ω is limited to:

FS FS
 F 
2 2
 
 
TS TS
The highest values of the frequency variables in the continuous time is called the folding
frequency:

FS 1
Fmax  
2 2T

 max   FS 
T

The sampling may introduce an ambiguity.

The Sampling Theorem
If the highest frequency contained in an analog signal xa(t) is Fmax and the signal is sampled at
a rate Fs > 2 Fmax (Nyquist rate), then the signal xa(t) can be exactly recovered from its sample
values using the intepolation function g(t) as follows:


 n   n 
x a (t )   x a   gt 
 

FS 
n    FS  

sin 2Fmax t
where g t  
2Fmax t

and xa(n/FS) = xa(nT) are the samples of the signal xa(t).

Basic discrete-time signal types
a) b)

c) d)

Basic signals. Impulse (a); unit step (b); decaying exponential (c); real-valued sinusoid
(d).
Discrete-time complex exponential
x[n] = exp(j n) (real and imaginary parts).
Elementary Operators

Shift (a sequence x[n], shifted by an integer k):

y[n] = x [n - k]

Scaling (a sequence x[n] scaled by a factor α C):

y[n] = αx[n]

Sum (the sum of two sequences x[n] and w[n] is their term-by-term sum):
y[n ] = x[n]+ w [n ]

Product (the product of two sequences x[n] and w[n] is their term-by-term
product):
y[n] = x[n]w[n]