Lecture: Introduction to

DSP simulations in MATLAB

Konstantin Rykov

konstantin.rykov@tut.fi

SGN-1158 Introduction to Signal Processing, short
version
Why youre at this lecture/lab?
Do not fear MATlab. Its your friend
MATlab is a tool
Where I can use MATlab? Examples
Im afraid of program languages

THE MAIN IDEA OF THE LECTURE

11.9.2012 2
ontents
BASICS OF MATLAB
Mainwindow. How to make m-file? How to save m-file?
Some basic hints
Main MATLAB objects (commands, variables)
Main operation symbols
Operation symbols
MATLAB IS AN ADVANCED CALCULATOR
Complex numbers
HELP
Vectors
Matrices
2D GRAPHS
Main MATLAB functions for plotting graphs
General rules of forming graphs
Main tools of staging graphs
Controlling graph properties
LineSpec parameters
OUTER FUNCTIONS IN MATLAB
11.9.2012 3
11.9.2012 4
ontents
DISCRETE SIGNALS IN MATLAB
Sequences
Unit sample sequence, unit step sequence, discrete exp
Discrete complex harmonic signal
Functions max, sum and prod
Generation of signals: rectpuls, tripuls, gauspuls, sinc, square, sawtooth, diric
Functions rand(1,N) and randn(1, N)



TASK: Open MATLAB



The main MATLAB window

11.9.2012 5
BASICS OF MATLAB
Some basic hints
help <name> (for example: >> help cos)
; blocks automatically output of the variables
% makes a comment
to comment a few rows hold Ctrl+R
to uncomment a few rows Ctrl+T
Always use: clc,clear all; close all;
TASK
Type in Editor:
===============
My MATlab Crib
===============
Use CTRL+R to comment it
clc; clear all; close all;


11.9.2012 6
Main MATLAB objects
Commands (clc, help, demo)
Constants (10, -17.28, 5+3j, 1e-6, 10^2)
Standard const (pi, 1i, eps)
Variables MATlab object, which might change its value during
simulation. All variables are MATRIXES in MATlab
Functions (sin(X), exp(X), log10(X), sqrt(X), abs(X),
real(X), imag(X))
Expressions is a sum of constants, functions, variables, which are
summed by operational symbols (x+sin(a)-sqrt(pi);)
11.9.2012 7
Main operation symbols
Symbol Operation
+ Summation
- Difference
* Multiplication of matrixes
.* Multiplication of elements
/ Right division
. Transposing
11.9.2012 8
Complex numbers
11.9.2012 9
MATLAB IS AN ADVANCED
CALCULATOR
Use MATLAB as calculator to find answers
Use help to find what these commands do
abs
angle
exp
conj
11.9.2012 10
Type and simulate
z=3+4i
r=abs(z)
fii=angle(z)
r*exp(i*fii)
zk=conj(z)
z*zk-r^2

What the command format does?
11.9.2012 11
Vectors
Type a=[2 4 5 7] and b=[-1 4 -2 1]
Find a+b, 2*a-2*b
What happens if you type a and b
a*b; a*b; a*b; a*b;
-1:10; 0:2:100; 1:-0.25:-2
Form vectors a=(7,8,9,,22); b=(0,2,4,,100);
c=(100,95,90,,35)
What did you get a(3)? a([3 5 7])? a(3:7)? a(3:end)?
11.9.2012 12
Matries



11.9.2012 13
A=[-7 5 -9; 2 -1 2; 1 -1 2];

Task
Calculate: 3A-5C, 7A+2B, CA, CD
Find out commands: zeros(n), zeros(m,n), ones(n), ones(m,n),
size(D), zeros(size(D)), diag([1 2 3 4]), eye(n)
What happens [A,B] and [A;B]?
Try to find an easy way to build a 7*8-matrix whose other entries are
zeros, but in its diagonal and its last column are 5s



NOTE: Transpose of a matrix is obtained with command
row with A(i,:) and column with A(:,j)
11.9.2012 14
Determine whether the given sets of vectors are linearly
independent/dependent:
W1=[1 2 3], W2=[2 1 5], W3=[-1 2 -4], W4=[0 2 -1]
Use MATLAB to to choose randomly three three column vectors in
The MATLAB commands to choose these vectors are:
y1=rand(3,1)
y2=rand(3,1)
y3=rand(3,1)

HINT check the command rref
11.9.2012 15
2D GRAPHS
Function Meaning
plot (x1, y1, x2, y2,)
Linear graphics
stem
Sequence graphs
stairs
Stairs graphs
loglog
Both Logarithmic axis Im and Re
semilogx
semiloxy
Logarithmic Re axis
Logarithmic Im axis
11.9.2012 16
Main MATLAB functions for plotting graphs
General rules of forming graphs
figure making a new window for a graph
subplot (n,m,p) drawing a few graphs in one window:
n colum, m row, p ordinal number of the graph
hold on plotting another graph at the same picture
hold off




For more information help graph2d

11.9.2012 17
Generate x=[1 20 3 15 18];
Use functions and tell what is the difference:
plot
stem
Generate x1=0:pi/8:8*pi. What we have done? Generate
y(t)=sin(x). Use functions to plot graphs:
plot
stem
stairs

HINT: use command figure or function subplot(n,m,p)
11.9.2012 18
Use semilogx, semilogy, loglog to plot graphs of the following
functions:
1. y=3x^5
2. y=3^(5x-2)
3. y=log10(3x^4)

Use subplot command into 3*3-subplot as described bellow

case (1) semilogx case (1) semilogy case(1) loglog
case (2) semilogx case (2) semilogy case(2) loglog
case (3) semilogx case (3) semilogy case(3) loglog

Consider again y=3x^5. Use plot(x,log10(y)) and compare
its plot with semilogy plot. What is the difference and similarity
between them?
11.9.2012 19
Main tools of staging graphs
Function
grid
title(<text>)
xlabel (<text>)
ylabel (<text>)
Legend (<funct1>,<funct2>,..,Pos)
axis([XMIN XMAX YMIN YMAX])
xlim ([XMIN XMAX])
ylim ([YMIN YMAX])
11.9.2012 20
Pos (-1, 0, 1,,4) TRY THEM!
Generate x1=0:pi/8:8*pi; y1=sin(x1);
Form 3 graph in 1 window.
1
st
graph: plot a discrete signal y(x)
2
nd
graph: plot a discrete signal. Use axis([0 10 -1 1])
3rd graph: do the same, but limit Re axis and Im axis by using
xlim([-15 15]) and ylim([-1.5 1.5])
For all graphs: make a grid, title and give names for both axis

Generate y2=0.5*sin(2*x1);
plot(x1,y1,x1,y2),legend(sin(x1),0.5sin(2x1));
!!!HINT: use hold on!!!
11.9.2012 21
Controlling grpah properies
Each function has different properties.
plot(x1,y1,,LineSpec,PropertyName,PropertyValue,);
stem(x1,y1,,LineSpec,fill,MarkerSize,3);
PropertyName is divided into:
LineWidth line width;
MarkerEdgeColor marker color ;
MarkerFaceColor color by which the marker is filled;
MarkerSize size of the marker , give a value (default - 7).
Let us divide LineSpec parameters into 3 groups: s1, s2, s3.
11.9.2012 22
LineSpec parameters
S1 S2 S3
r Red - +
b Blue : *
g Green -. s Square
w White -- d Diamond
k Black (none) v
y Yellow ^
m Magenta <
c Cyan >
p Pentagram
h Hexagram
11.9.2012 23
Form a vector y = [0 1 2 3 4 5 6 7 8 9]; line width is 2,
use squared black markers, dotted line
x1=0:pi/8:8*pi; y1=sin(x1); - line width 3, dashdot line,
filled green markers, marker size 5
y1=sin(x1); y2=0.2*cos(5x1); - one line is dashed, another
is solid; one line is red, another is green; markers, different sizes

11.9.2012 24
OUTER FUNCTIONS IN MATLAB
Function file is a M-file, which generates outer function









DO NOT PUT ; after function row
After function there is a function body
Put ; everywhere in the body to prevent undesirable output
Good programming means good comments
11.9.2012 25
If you have a few parameters
function [z, p] = F1(x,y)
% Sum of cubes z
% Square root p
z=x.^3+y.^3;
p=sqrt(abs(z));
end
If you have one parameter
function z = F2(x,y)
% Sum of cubes z
z=x.^3+y.^3;
end
After making and saving function-file you can use it in other M-files (script
files).
Actual/Real parameters a=4, b=3, [d,c]=F1(a,b) => saved in
Workspace
Formal parameters 3+5-sqrt(9) => not saved in Workspace
11.9.2012 26
Number of input and output parameters can be formed by commands:
nargin(<function name>)
nargout(<function name>)
Listing of the function is formed by command:
type <name of function-file>
If you need commends of the function file:
help <name of function-file>
If you need to exit compulsory from the body of the outer function use
operator:
return
11.9.2012 27
Let us remake function F1 to F3 with controlling negative argument of
the square root and appropriation p=0 in this case:

function [z, p] = F3(x,y)
% Sum of cubes z
% Square root p
z=x.^3+y.^3;
if z<0
p=0;
return
else
p=sqrt(z);
end
end

11.9.2012 28
DESCRETE SIGNALS IN MATLAB
What is a discrete signal? How does it look like?
How to make a discrete signal:
Matrix x=[0 -1; -4 7; 3 2];
Vector y=[1 20 3 15 18];
Pair of vectors n1=0:12; x1=n.^2;
Vector+Matrix n2=0:2; x2=[0 -1; -4 7; 3 2];




11.9.2012 29
Sequences
11.9.2012 30
1 1.5 2 2.5 3
-5
0
5
10
n
1 2 3 4 5
0
5
10
15
20
n
0 5 10 15
0
50
100
150
n
0 0.5 1 1.5 2
-5
0
5
10
n
Unit sample sequence, unit step sequence, discrete exp,
Form a unit sample sequence, unit step sequence and discrete exp.
The length of the sequence is N=11. Plot graphs.
11.9.2012 31
Discrete complex harmonic signal
is presented as
or , where
Fs=1/T. Real and imaginary parts of x(n) are calculated
by functions real and imag. Absolute value and angle/phase can be
hound with the use of abs and angle

Now, present 32 samples of DCHS x(n), if C=2 and
w=pi/8. Plot real, imaginary parts of the signal.
Present absolute value and the phase of x(n).
11.9.2012 32
Functions max, sum and prod
We can work only with vectors/matrices, which have the same
length/dimentions.
Generate 3 signals x1=(0.8.^n1), x2=cos(w*n2) and
x3=sin(w*n3) with
vector length correspondingly N1=16, N2=24, N3=32 and w=pi/8.
N=max([N1 N2 N3]) to find the maximum value of the vector
length. To add the needed number of zeros to the signal:
y1=[x1 zeros(1, (N-N1))];
Use sum to summate signals and prod to multiply signals. Check
commands if needed in help.
11.9.2012 33
Generation of signals: rectplus, triplus, gausplus, sinc,
square, sawtooth, diric
There is a number of functions for generating signals in the folder
Signal Processing Toolbox.
y=rectpuls(t,w);
y=tripuls(t,w,s);
y=gauspuls(t,fc, bw);
y=sincpuls(t);
y=squarepuls(t,d);
y=sawtoothpuls(t,width);
y=diricpuls(x,N);




11.9.2012 34
Functions rand(1, N) and randn(1, N)
RAND is a uniformly distributed pseudorandom number.
RANDN is a normally distributed pseudorandom numbers.
(1, N) number of rows and columns.

Form additive mixture (sum) of sequence x(n)=sin(wn) with the length N=32
with white noise: uniformly distributed and normally distributed.

11.9.2012 35
Thanks for attention!
Questions?
11.9.2012 36