Appendix A

MATLAB Fundamentals

The MATLAB provides the extensive tutorials for all its products in various format.
This appendix briefly introduces the MATLAB fundamentals based on the
MATLAB Primer. You can observe the primitive MATLAB capabilities in short.
Once you click the MATLAB icon, you will get the following screen as Fig A.1.
We can observe the four window spaces as below.
– Current Folder: Your working directory and files.
– Command Window: Enter commands for immediate execution.
– Workspace: Variables and data.
– Editor-Untitled: Once click the New button, the editor window provides the
script space for programming.
Most of the icon buttons and ribbon menus are intuitive and self-explanatory for
its functions. We can execute the MATLAB commands by
– Type the command in command window and put the <enter> key
– Write the program in the editor window and put the <ctrl>+<enter> key
The script program also can be executed by clicking the run button. One big
advantage of the MATLAB programming is the worry-free variables. If we need the
variables, just create and use them. Below example creates the a and b variable with
initial values in command window. As soon as we type the ‘a=1’ and <enter>, the
a variable is created with initial value one. When we compute the value without the
assignment, the output value is allocated at the temporary variable ans. See the ‘a+b’
command in below. All created variables can be seen in the workspace window.

© Springer Nature Singapore Pte Ltd. 2021 645

K. Kim, Conceptual Digital Signal Processing with MATLAB,
Signals and Communication Technology,
https://doi.org/10.1007/978-981-15-2584-1
646 Appendix A: MATLAB Fundamentals

Fig. A.1 MATLAB screen

Prog. A.1 Variable creation and initialization

>> a = 1
a=
1
>> b = 2
b=
2
>> a+b
ans =
3
>>

To clear the command window, type the clc. The second MATLAB advantage is
the strong supports for the linear algebra operations based on the matrix and array.
Just like variable, we can create the matrix and array as below.
Appendix A: MATLAB Fundamentals 647

Prog. A.2 Creation, initialization, and arithmetic of matrix variables

>> a = [1 2 3]
a=
1 2 3
>> b = [4, 5, 6]
b=
4 5 6
>> c = [7; 8; 9]
c=
7
8
9
>> d = [1 2 3; 4 5 6; 7 8 9]
d=
1 2 3
4 5 6
7 8 9
>> e = d*a
Error using *
Incorrect dimensions for matrix mulplicaon. Check that the number of
columns in the first matrix matches the number of rows
in the second matrix. To perform elementwise mulplicaon, use '.*'.
>> e = a*d
e=
30 36 42
>>

The most arithmetic operators in MATLAB perform the linear algebra opera-
tions which require to match the dimensions between the matrices. Table A.1
shows the arithmetic operator list. The addition and subtraction are working for the
constant, matrix, and element-wise operation.
Other operators for logic and relation are listed below as Table A.2.
The matrix can be generated by the MATLAB functions. The part of functions is
shown below in Table A.3.
Below example demonstrates the matrix generation from the function. The first
command rng(‘default’) places the random number generator setting to their default
values so that they produce the same random numbers.
648 Appendix A: MATLAB Fundamentals

Table A.1 Arithmetic operator list of MATLAB

Symbol Role
+ Addition
− Subtraction
* Matrix multiplication
/ Matrix right division
\ Matrix left division
^ Matrix power
‘ Complex conjugate transpose
.* Element-wise multiplication
./ Element-wise right division
.\ Element-wise left division
.^ Element-wise power
.’ Transpose

Table A.2 Logical and relational operator list of MATLAB

Symbol Role
¼= Equal to
*= Not equal to
> Greater than
>= Greater than or equal to
< Less than
<= Less than or equal to
& Logical AND
| Logical OR
&& Logical AND (with short-circuiting)
|| Logical OR (with short-circuiting)
* Logical NOT

Table A.3 Special matrix generation functions

Function Role
zeros Create array of all zeros
ones Create array of all ones
rand Uniformly distributed random numbers
eye Identity matrix
Appendix A: MATLAB Fundamentals 649

Prog. A.3 Matrix generation example from function

>> rng('default')
>> aa = rand(3,3)
aa =
0.8147 0.9134 0.2785
0.9058 0.6324 0.5469
0.1270 0.0975 0.9575
>> bb = ones(3,2)
bb =
1 1
1 1
1 1
>>

The matrix is also produced by the concatenation which joins the arrays to make
larger ones. See below example.

Prog. A.4 Matrix generation example by concatenation

>> rng('default')
>> bb = randi(10,2,3)
bb =
9 2 7
10 10 1
>> cc = [bb bb]
cc =
9 2 7 9 2 7
10 10 1 10 10 1
>> dd = [bb, bb]
dd =
9 2 7 9 2 7
10 10 1 10 10 1
>> ee = [bb; bb]
ee =
9 2 7
10 10 1
9 2 7
10 10 1
>>
650 Appendix A: MATLAB Fundamentals

Use the indexing, we can access or modify selected elements of an array. The
most general method is to indicate row and column number which starts from one.
A single column-wise numbering can specify the matrix position as well. The
matrix element can be modified by assigning the specific value to the element.

Prog. A.5 Matrix indexing and modification

>> bb = randi(100,4,4)
bb =
23 54 11 82
92 100 97 87
16 8 1 9
83 45 78 40
>> bb(2,4)
ans =
87
>> bb(14)
ans =
87
>> bb(2,4)=0
bb =
23 54 11 82
92 100 97 0
16 8 1 9
83 45 78 40
>> bb(2,1:3)
ans =
92 100 97
>> bb(:,2)
ans =
54
100
8
45
>> bb(end,:)
ans =
83 45 78 40
>> cc = 0:5:30
cc =
0 5 10 15 20 25 30
>> dd = 1:10
dd =
1 2 3 4 5 6 7 8 9 10
>>
Appendix A: MATLAB Fundamentals 651

The colon operator allows to refer the multiple elements in the matrix as the
start:end form. The colon alone indicates all of the elements in that dimension. The
end indicator denotes the last row or column number. The general colon operator
creates an equally spaced vector using the form start:step:end. Missing the step
value indicates the default step value of one.
MATLAB provides a large number of functions which perform various tasks.
The input and output argument format can be found by using the help or doc
function as below. Note that doc function opens the new window for information.

Prog. A.6 MATLAB help

>> help max

max Largest component.
...
[Y,I] = max(X) returns the indices of the maximum values in vector I.
If the values along the first non-singleton dimension contain more
than one maximal element, the index of the first one is returned.
...
Example:
X = [2 8 4; 7 3 9]
max(X,[],1)
max(X,[],2)
max(X,5)
...
>>

One example of function usage is to find the minimum, negative, and zero
locations of the given function below.

y ¼ x2  1 for  10  x  10

Find the x values which satisfy the following conditions.

min x2  1
x
x  1\0
2

x2  1 ¼ 0

The MATLAB numerically computes the values based on the discretized input
values for the given range. The continuous input is divided into the discrete input
with 0.1 step by using the colon operator. The output is calculated by the
element-wise operation. The min and find function perform the designated tasks.
Note that the fine-grain results can be derived from the reduced step size for input.
652 Appendix A: MATLAB Fundamentals

In general, the function supports the various argument styles. See help for further
information. Below example is written by the script editor.

Prog. A.7 Numerically solve the equation

x = -10:0.1:10; % Input values

y = x.^2-1; % Output equaon

[k1,I1] = min(y) % Find the min value at k1 and locaon I1

[k2] = find(y<0) % Find the locaons for negave output
y(k2) % Corresponding negave values
[k3] = find(y==0) % Find the zero locaons
y(k3) % Verify the zero outputs

k1 =
-1
I1 =
101
k2 =
92 93 94 95 96 97 98 99 100 101
102 103 104 105 106 107 108 109 110
ans =
Columns 1 through 13
-0.1900 -0.3600 -0.5100 -0.6400 -0.7500 -0.8400 -
0.9100 -0.9600 -0.9900 -1.0000 -0.9900 -0.9600 -0.9100
Columns 14 through 19
-0.8400 -0.7500 -0.6400 -0.5100 -0.3600 -0.1900
k3 =
91 111
ans =
0 0

The third advantage of the MATLAB is the powerful visualization tools. The
impressive 2-D and 3-D plots are available with various decoration functions to
display extensive information. Labeling, legend, range, and etc. are specified by the
MATLAB functions. Using the TeX markup, Geek letters and special characters
can be added to the plot. Some of the list are arranged below in Table A.4.
Appendix A: MATLAB Fundamentals 653

Table A.4 Special text markup for special characters

Char. sequence Symbol Char. sequence Symbol
^{ } Superscript \delta d
_{ } Subscript \epsilon

\bf Bold font \theta h
\it Italic font \pi p
\alpha a \omega x
\beta b \Omega X
\gamma c \rightarrow !

Below example shows the line and stem plots.

Prog. A.8 2-D plot example

x = 0:pi/100:2*pi; % Input
y1 = sin(x); % sine funcon
y2 = cos(x); % cosine funcon

figure,
plot(x,y1,x,y2), grid % Line plot for sine and cosine
tle(‘Example 1’) % Title
legend('sin(\omega)','cos(\omega)'); % Legend with special char
xcks([0 pi/2 pi 3*pi/2 2*pi]) % Tick posion
xcklabels({'0','\pi/2','\pi','3\pi/2','2\pi'}) % Tick label
xlim([0 2*pi]) % x display range
ylim([-1.5 1.5]) % y display range
xlabel('\omega') % x label
ylabel('Magnitude') % y label

figure,
subplot(211), stem(x,y1), grid % Subplot with stem
tle(‘Example 2’)
ylabel('sin(\omega)')
xcks([0 pi/2 pi 3*pi/2 2*pi])
xcklabels({'0','\pi/2','\pi','3\pi/2','2\pi'})
xlim([0 2*pi])
subplot(212), stem(x,y2), grid
ylabel('cos(\omega)')
xcks([0 pi/2 pi 3*pi/2 2*pi])
xcklabels({'0','\pi/2','\pi','3\pi/2','2\pi'})
xlim([0 2*pi])
xlabel('\omega')
654 Appendix A: MATLAB Fundamentals

Example 1
1.5
sin( )
cos( )
1

0.5
Magnitude

-0.5

-1

-1.5
0 /2 3 /2 2

Fig. A.2 Output of Prog. A.8. Part 1

The simple sine and cosine plots are displayed for 2p range as below (Fig A.2).
The subplot function allows us to place the more than one plot in the single
figure as Fig A.3.
The MATLAB also gives the versatile 3-D plot features. Below example pre-
sents the 3-D Gaussian plot according to the following equation.

z ¼ f ðx; yÞ ¼ eððxx0 Þ þ ðyy0 ÞÞ

Appendix A: MATLAB Fundamentals 655

Example 2
1

0.5
sin( )

-0.5

-1
0 /2 3 /2 2

0.5
cos( )

-0.5

-1
0 /2 3 /2 2

Fig. A.3 Output of Prog. A.8. Part 2

Prog. A.9 3-D plot example

[X,Y] = meshgrid(-2:0.1:2); % Cartesian grid

Z = exp(-(X.^2+Y.^2)); % 2-D Gaussian funcon Z = f(X,Y)

figure,
surf(X,Y,Z) % 3D surface plot
tle('3-D Normal Distribuon') % Title
xcks([-2 -1 0 1 2]) % Tick posion
xcklabels({'-2','-1','x_0','1','2'}) % Tick label
xlabel('x') % x axis label
ycks([-2 -1 0 1 2]) % Tick posion
ycklabels({'-2','-1','y_0','1','2'}) % Tick label
ylabel('y') % y axis label
zlabel('e^{-((x-x_0)^2+(y-y_0)^2)}') % z axis label
colormap('gray') % Color bar color configuraon
colorbar % Color bar
656 Appendix A: MATLAB Fundamentals

3-D Normal Distribution

1

0.9

1
0.8

0.8 0.7
+(y-y ) )
2
0

0.6 0.6
2
e -((x-x 0 )

0.4 0.5

0.4
0.2
0.3
0
2 0.2
1 2
y0 1 0.1
x0
-1
-1
y -2 -2
x

Fig. A.4 Output of Prog. A.9

Prior to create the 3-D plot, we must generate the input coordinates for the 3-D
equation. The meshgrid function returns 2-D grid coordinates based on the coor-
dinates contained in input vector(s). The X and Y are the square grid coordinates
with single input grid size. The element-wise operation completes the equation
calculation followed by the 3-D plot function. Below figure shows the 3-D surface
plot. We can change the viewpoint by clicking the rotate 3D button and dragging
the figure (Fig A.4).
We can create the MATLAB function by generating the new script for function
as shown in Fig A.5.
The created editor includes the function format as below. Decide function name,
write function code, and update the comment. Finally save the function as the
function name with m extension.

Prog. A.10 Created new function format

funcon [outputArg1,outputArg2] = untled2(inputArg1,inputArg2)

%UNTITLED2 Summary of this funcon goes here
% Detailed explanaon goes here
outputArg1 = inputArg1;
outputArg2 = inputArg2;
end
Appendix A: MATLAB Fundamentals 657

Fig. A.5 New script for function

The simple sum function is written for example as below.

Prog. A.11 Function example

funcon [z] = Mysum(x,y)

% Compute the sum of two variables or arrays
% z = x+y
% Example
% x = [1 2 3];
% y = [3 5 7];
% z = Mysum(x,y)
% Created by Keonwook Kim
% Jan. 23rd 2019
z = x+y;
end

Save the function as Mysum.m in the working directory. The help function with
Mysum replies the comment in the function as shown below. Execute the comment
example for verification.
658 Appendix A: MATLAB Fundamentals

Prog. A.12 Function help and execution

>> help Mysum

Compute the sum of two variables or arrays
z = x+y
Example
x = [1 2 3];
y = [3 5 7];
z = Mysum(x,y)
Created by Keonwook Kim
Jan. 23rd 2019
>> x = [1 2 3];
y = [3 5 7];
z = Mysum(x,y)
z=
4 7 10
>>

We can execute the part of MATLAB program code as section. Sometime, this
kind of execution is required for debugging. The double percent sign (%%) divides
the program into the sections as shown in Fig A.6. By moving the cursor, the
selected section is highlighted by yellow color as shown below. Typing the <ctrl>
+<enter> executes the chosen section. MATLAB provides the advanced debugging
tools and the section run is the part of them.
In command window, the command history is recalled by clicking the <up> key
and the history is narrowed by typing the character letters as shown in Fig A.7. The
quick access to the history makes the multiple executions easy for various purpose.
Followings are the frequently used commands.
clear all: Clear all variables
clc: Clear the command window
clf: Clear the current figure
close all: Close all figure windows
lookfor ‘keyword’: Search all MATLAB files for keyword
Further information about the MATLAB and toolboxes can be referred from the
MathWorks documentations.
Appendix A: MATLAB Fundamentals 659

Fig. A.6 Section division

Fig. A.7 Recall command history

Appendix B
MATLAB Symbolic Math Toolbox

Symbolic math performs exact computation with variables which have no given
value as symbols. Symbolic math toolbox provides functions for solving and
manipulating symbolic math equations. The toolbox delivers functions in subjects
such as calculus, linear algebra, differential equations, equation simplification, and
equation manipulation. The computations can be realized either analytically or
using variable-precision arithmetic. The MATLAB live script editor presents the
optimal environment to create and run the symbolic math code. We can share the
symbolic work with other users as live scripts or convert them to HTML or PDF for
publication. We can generate the new live script by selecting the New ! Live
Script in the menu as below Fig B.1.
The live script is divided into the text and code area. The text area (white)
describes the background, algorithm, explanation, analysis, etc. with plain text. The
code area (gray) locates the MATLAB code for algorithm execution.
The MATLAB outcome is embedded at the live script immediate right or below
according to your choice as shown in Fig B.2.

© Springer Nature Singapore Pte Ltd. 2021 661

K. Kim, Conceptual Digital Signal Processing with MATLAB,
Signals and Communication Technology,
https://doi.org/10.1007/978-981-15-2584-1
662 Appendix B: MATLAB Symbolic Math Toolbox

Fig. B.1 Generate live script

The buttons under the LIVE EDITOR ribbon menu specify the file, text, code,
and execution related information. The buttons are intuitive and self-explanatory.
Example live script code is below. Note that the below is not the symbolic math
code. Execute the code by clicking <ctrl>+<enter>.
Appendix B: MATLAB Symbolic Math Toolbox 663

Prog. B.1 Example of live script code for MATLAB programming

Title
Heading
Normal
% Code
x = 0:pi/100:2*pi; % Input
y1 = sin(x); % sine funcƟon
y2 = cos(x); % cosine funcƟon
plot(x,y1,x,y2), grid % Line plot for sine and cosine
Ɵtle('Example 1') % Title
legend('sin(\omega)','cos(\omega)'); % Legend with special char
xƟcks([0 pi/2 pi 3*pi/2 2*pi]) % Tick posiƟon
xƟcklabels({'0','\pi/2','\pi','3\pi/2','2\pi'}) % Tick label
xlim([0 2*pi]) % x display range
ylim([-1.5 1.5]) % y display range
xlabel('\omega') % x label
ylabel('Magnitude') % y label

Most type of MATLAB code can be executed in the live script. The graphic
environment of live script gives the powerful presentation for the symbolic math
code which use the mathematical symbols and characters frequently. Now, we will
explore the symbolic math toolbox with live script editor. Below example shows
the symbolic math example.
664 Appendix B: MATLAB Symbolic Math Toolbox

Fig. B.2 Live editor ribbon menu

Prog. B.2 Example of live script code for symbolic math

Create symbolic variables

clear all
syms a % Symbolic variable declaraƟon
a = sym(pi) % Symbolic pi
a=π
sin(a) % Symbolic sin(pi) result
ans = 0
sin(pi) % Numeric sin(pi) result
ans = 1.2246e-16
A = sym('b',[1 10]) % MulƟ symbolic variable declaraƟon
A=
syms f(x,y) % Symbolic funcƟon declaraƟon
f(x,y)=x^2*y % Define funcƟon
f(x, y) =
f(3,2)
ans = 18
B = sym('C%d%d',[3 3]) % Declare the 3X3 matrix
B=

inv(B) % Square matrix inversion

ans =
Appendix B: MATLAB Symbolic Math Toolbox 665

The symbolic math on MATLAB requires to declare the symbolic variable

before use. The sym creates numbered symbolic variables or symbolic variables.
The syms command is shorthand for the sym syntax. The above code demonstrates
the various situations for symbolic variable declarations. The sinðpÞ computation
with numeric and symbolic p results in the different outcome due to the precision
limitation. We can define the multi-variable function by using the symbolic rep-
resentation.

Prog. B.3 Symbolic computation example

Perform symbolic calculus

clear all
syms x y
f1 = sin(x)^2+cos(y)^2
f1 =
diff(f1) % Differenate
ans =
diff(f1,y) % Differenate over y
ans =
int(f1) % Integral (indefinite)
ans =

int(f1,y) % Integral (indefinite) over y

ans =

int(f1,-2,2) % Integral (definite) over x from -2 to 2

ans =

The above code performs the symbolic computations for simple calculus. The
symbolic math toolbox supports extensive calculus functions such as differentia-
tion, integral, vector operations, series, limits, transformation, and etc.
666 Appendix B: MATLAB Symbolic Math Toolbox

Prog. B.4 Solving the polynomials

Solve equations
clear all
syms x y z
z1 = x^3-6*x^2
z1 =
z2 = 6-11*x
z2 =
solve(z1==z2)
ans =

solve(z1-z2==0)
ans =

z3 = 6*x^2 - 6*x^2*y + x*y^2 - x*y + y^3 - y^2

z3 =
solve(z3==0,y)
ans =

[x,y,z] = solve(z == 4*x, x == y, z == x^2 + y^2)

x=

y=

z=
Appendix B: MATLAB Symbolic Math Toolbox 667

We can solve the multi or single variable polynomials as above. The symbolic math
toolbox provides the equation solvers for polynomials and differential equations.

Prog. B.5 Manipulation of polynomials

Manipulation of polynomials
syms x y
f1 = (x^2- 1)*(x^4 + x^3 + x^2 + x + 1)*(x^4 - x^3 + x^2 - x + 1)
f1 =
expand(f1) % Expand polynomial
ans =
f2 = x^3 + 6*x^2 + 11*x + 6
f2 =
factor(f2) % Factor polynomial
ans =
f3 = x^5 + x^4 + x^3 + x^2 + x
f3 =
horner(f3) % Nested representaon for polynomial
ans =
f4 = 2*x^2 - 3*x + 1
f4 =
aa = subs(f4,1/3) % Substute the x with 1/3
aa =

eval(aa) % Compute the value

ans = 0.2222
f5 = x^2*y + 5*x*sqrt(y)
f5 =
subs(f5,x,3) % Substute the x with 3
ans =

Manipulations for polynomial can be performed as shown above. The toolbox

presents the extensive simplification methods for instance expand, decomposition, fac-
torization, and etc. If we do not need a particular form of expressions, use simplify
function to shorten mathematical expressions. The simplify works universally; otherwise,
we apply the specific manipulations on the equation to derive the specific form.
668 Appendix B: MATLAB Symbolic Math Toolbox

Prog. B.6 FIR filter design by using symbolic math

Symbolic math example

clear all
syms w n
f1 = int(exp(1j*w*n),w,-pi/4,pi/4)/(2*pi)
% Inverse DTFT (1 for -pi/4 ~ pi/4) h[n]
f1 =

f2 = simplify(f1) % Simplify the equaon h[n]

f2 =

f3 = subs(f2,n,(-20:20)+eps);
% Substute n with -20:20 with eps to avoid the 0/0
f4 = eval(f3); % Compute the substuon
f4(21) % Approximated h[0]
ans = 0.2500
temp = limit(f2,n,0) % Verify real h[0]
temp =

stem((-20:20),f4), grid % Stem plot

xlabel('n')
ylabel('h[n]')

The above code performs the inverse DTFT to design the LPF with cutoff
frequency p=4. The corresponding impulse response is displayed as stem plot. Note
that the simplify function makes the equation in simple form. The substitution is
executed to replace the n symbolic variable with range of integer values. In zero
Appendix B: MATLAB Symbolic Math Toolbox 669

replacement condition, the equation represents the NaN (Not-a-Number) outcome.

To avoid the NaN output, the integer values are added with very small value eps
(2.220410−16) which is the distance from 1.0 to the next larger double-precision
number. The zero input result is verified with the limit function which computes the
output for infinitesimal approach to given input value.

Prog. B.7 Plot in symbolic math

Plot the symbolic math

syms x y
f1 = x^3 - 6*x^2 + 11*x - 6
f1 =
fplot(f1,[-5 10]) % 2D plot of symbolic funcon
xlabel('x')
ylabel('y')
tle(texlabel(f1)) % Title the TeX format equaon
grid on

f2 = x^2 + y^2
f2 =
fsurf(f2) % Surface plot of symbolic funcon (auto range)
xlabel('x')
ylabel('y')
tle(texlabel(f2))
grid on
670 Appendix B: MATLAB Symbolic Math Toolbox

Again, there is no given value to the symbolic variables. This condition is also
true for drawing the plot in symbolic math. We only assign the symbolic function to
the designated plot functions to illustrate the plot as shown above.

Prog. B.8 Assumptions on variables

Assumptions
clear all
syms n x
f1 = sin(2*n*pi) % Given funcon1
f1 =
simplify(f1) % Simplify funcon1
ans =
assume(n,'integer') % Assume that n is integer
simplify(f1) % Re-simplify funcon1
ans = 0
f2 = x^4==1 % Given funcon2
f2 =
solve(f2,x) % Solve funcon2
ans =

assume(x,'real') % Assume that x is real

solve(f2,x) % Re-solve funcon2
ans =

assumeAlso(x>0) % Assume again x is real and posive

solve(f2,x) % Re-solve funcon2
ans = 1
assumpons % Show all assumpons
ans =
assume([x n],'clear') % Clear assumpons on x and n
assumpons % Show all assumpons again
ans =
Empty sym: 1-by-0
Appendix B: MATLAB Symbolic Math Toolbox 671

When the symbolic variables are declared, there is no assigned conditions.

Therefore, the variable could be any value and number set. We can apply certain
conditions on the variables to specify the output range. See above code for example.
Index

Symbols Block convolution, 577, 581

( ), 1 Bohman window, 304
[ ], 1 Bounded Input and Bounded Output (BIBO)
2’s complement, 504 stability, 135
,1 Brick wall specification, 276
Butterworth IIR filter, 395
A
Absolute summability, 236 C
Additivity, 51 Canonic form, 559
Aliasing, 65, 489 Cascade connection, 526
Allpass filter, 384, 479 Cauchy residue theorem, 540
Alternation theorem, 344 Causality, 129, 149
App, 596 Causal system, 129
Arguments, 1 Ceiling function, 517
Arithmetic mean, 513 Chebyshev approximation, 344
Auto-completion, 611 Chebyshev polynomial, 343
Autocorrelation, 461 Chebyshev window, 304
Auto-listing, 611 Coefficient quantization, 521
Autoregressive (AR) model, 446 Commutative law, 53
Autoregressive Moving Average (ARMA) Complex and nonlinear-phase equiripple
model, 447 design, 380
Complex Chebyshev approximation, 380
B Compressed, 49
Band Pass Filter (BPF), 186, 192 Conjugate transpose (Hermitian conjugate),
Band Pass Filter (BPF) - FIR type III, 329 354
Band Stop Filter (BSF), 186, 195 Constrained least square design, 376
Bartlett window, 304 Continuous frequency distriution, 83
Base, 505 Convolution sum, 12, 53, 123
BIBO stability, 236 Convolution sum in frequency domain, 138
BIBO stability condition, 135 Convolution sum to multiplication property,
Bilinear transformation, 399, 413, 426, 444, 228
493 Coupled form for SOS structure, 532
Binary number, 504 Cutoff frequency, 274
Bit-reversed order, 572, 577 Cyclic frequency, 274
Blackman window, 304

© Springer Nature Singapore Pte Ltd. 2021 673

K. Kim, Conceptual Digital Signal Processing with MATLAB,
Signals and Communication Technology,
https://doi.org/10.1007/978-981-15-2584-1
674 Index

D FIR filter transformations, 187

Decibel unit, 268 FIR realization structures, 553
Decimation-in-frequency FFT, 572 First kind Chebyshev polynomial, 406, 416,
Decimation-in-time Fast Fourier transform 424
(FFT), 565 Fixed-point designer, 618
Delta function, 35 Fixed-point number, 511, 619
Denormalize, 507 Fixed-point number computation, 511
DFT for non-periodic signal, 98 Flat top weighted window, 304
DFT length, 105 Flipped transposed FIR filter, 554
DFT length for convolution sum, 141 Floating-point number, 505
DFT method - filter design, 150 Floor function, 512
DFT method without truncation, 356 Flow graph, 570, 577
Differentiator, 337 Fourier transform, 32, 36
Dirac delta function, 35 Fourier transform for discrete time signal, 136
Direct form, 554 Four types of FIR filter with zero group delay,
Direct form I, 558 285
Direct form II, 559 Fraction length, 619
Discrete delta function, 87 Fraction number, 512
Discrete Fourier Transform (DFT), 89 Frame-based processing, 579
Discrete frequency x, 59, 273 Frequency, 2
Discrete-Time Fourier Transform (DTFT), 83 Frequency aliasing, 493
Discrimination factor, 438 Frequency discretization, 83
Double precision, 509 Frequency edge specifications, 274
DTFT method - filter design, 142 Frequency inversion, 199
Duality, 141 Frequency resolution, 102
Dual poles, 244 Frequency response, 234
Dual poles with zeros, 252 Frequency sampling design - type I, 353
Frequency sampling design - type II, 360
E Frequency sampling method, 353
Effective frequency, 65 Frequency shift, 192, 199
Effective radian frequency, 62 Frequency shift - DFT, 114
Elliptic filter, 430 Frequency shift - DTFT, 87
Elliptic rational function, 431 Frequency transformation in IIR filter, 477
Energy, 54 Fundamental frequency, 81, 89
Energy ratio between the main lobe and side
lobe, 279 G
Entirety-based processing, 579 Gaussian window, 304
Equivalent filter, 554 Geometric series, 205
Euler formula, 24, 57, 66, 90 Gibbs phenomenon, 276
Even (odd) function, 27 Group delay, 284, 385
Expanded, 49
Expected value, 513 H
Exponent, 505 Halfpower frequency, 274
Hamming window, 304
F Hann (Hanning) window, 304
Filter, 12 High Pass Filter (HPF), 7, 186, 188
Filtering length, 5 High Pass Filter (HPF) - FIR type IV, 334
Filter order, 274 Hilbert transform, 339
Final value theorem, 212
Finite Impulse Response (FIR) filter, 130 I
FIR filter advantages, 275 IEEE standard 754, 509
FIR filter gain, 632 Ill-condition, 523, 556
Index 675

IIR filter method summary, 494 Minimum attenuation, 271

Impulse invariance, 500 Minimum group delay, 390
Impulse response, 52, 228 Minimum phase filter, 443, 479
Impulse train, 96 Minimum phase filter design, 382
Infinite Impulse Response (IIR) filter, 133 Minimum phase system, 389
Initial values, 232 Modified Bartlett-Hann window, 293
Initial value theorem, 212 Most Significant Bit (MSB), 504
Invalid radian frequency, 62 Moving Average (MA) model, 446
Inverse Z-transform, 213 Multiplication, 234
Multiplication for convolution sum, 137
K Multiplication property, 137
Kaiser window, 304, 312 Multi poles with zeros, 258
Kronecker delta function, 46
Kuhn–Tucker conditions, 377 N
Negative frequency, 24
L Normalized frequency, 273
L∞-norm, 539 Not-A-Number (NaN), 507
L1-norm, 537 Nuttall-defined minimum 4-term
L2-norm, 538 Blackman-Harris window, 304
Lagrangian, 377 Nyquist sampling rate, 45, 59
Laplace transform, 399, 411, 423
Leakage factor, 309 O
Least Significant Bit (LSB), 504 Object-oriented programming, 604
Least squared error frequency domain design— Orthogonality, 32, 71, 91
discrete frequency sampling DFT, 364 Overlap-add method, 584
Least squared error frequency domain design— Overlap-save method, 586
integral squared error approximation
criterion, 372 P
L’Hospital’s rule, 35 Parallel connection, 527
Limit cycle, 563 Parametric methods, 446
Linear and Time Invariant (LTI), 52, 124 Parks-McClellan algorithm, 380
Linear and time invariant system, 14 Parseval’s theorem, 143, 177, 236, 278, 319,
Linearity, 51 322, 327, 331, 334, 337, 372, 376
Linearity - DFT, 113 Parseval’s theorem - DFT, 115
Linearity - DTFT, 87 Parseval’s theorem - DTFT, 88
Linear phase, 275, 284 Partial fraction, 216, 527
Linear prediction filter, 461 Partial fraction expansion, 540
Low Pass Filter (LPF), 142 Parzen window, 304
LPF design method selection, 185 Passband frequency, 274
LPF design procedure, 187 Passband magnitude, 271
Passband ripple, 269
M Peak-to-peak ripple, 396
Magnitude, 268 Period, 72
Magnitude inversion, 188, 197 Periodicity in spectrum, 67
Main lobe, 279 Phase delay, 395
Main lobe width, 291, 305 Phasor, 28
Mantissa, 505 Pole, 220
MATLAB coder, 638 Pole-zero plot, 221
Maxflat filter, 459 Polynomial long division, 214
Maximum deviation, 271 Positive (negative) frequency, 67
Minimum 4-term Blackman-Harris window, Power, 56
293 Prefiltering, 473
676 Index

Pre-scaling, 592 Sensitivity, 523

Primitive low pass filter, 4 Shifting, 48
Probability density function, 513 Shift pattern for DFT, 111
Prony, 461, 473 Shift property - DFT, 114
Pseudoinverse, 366 Signal periodicity, 83
Signal power, 513
Q Signal-to-Noise Ratio (SNR), 514
QR solver, 471 Signal types and pole locations, 225
Quantization, 511, 548 Sign bit, 504
Quantization effect, 587 Signed fraction number, 511
Quantization error sources, 548 Signed integer, 504
Significand, 505
R Simulink, 595
Radian frequency, 22, 62 Sinc function, 37
Ramp function, 47 Single pole, 244
Random process, 513 Single pole with zero, 247
Random variable, 513 Single precision, 509
Real sequence - DFT, 114 Single quantization noise, 551
Real signal property, 210 Single Sideband (SSB) output, 339
Real-time signal, 1 Speedup, 642
Rectangular window, 291 Stability, 275
Recursive filter, 132 Statistical signal processing, 447
Reflecting, 49 Steiglitz-McBride algorithm, 468
Region of Convergence (ROC), 205 Stopband, 398
Relatioinship between DTFT and DFT, 94 Stopband attenuation, 271
Relatioinship between Z-transform, DTFT, and Stopband frequency, 274
DFT, 218 Substitution method, 134
Relative peak side lobe amplitude, 291 Symmetricity for impulse response, 284
Relative sidelobe attenuation, 309 System object, 604
Remes exchange algorithm, 345
Remes exchange algorithm (Parks and T
McClellan), 342 Taylor window, 304
Restrictions for linear phased FIR filters, 288, Tense, 1
291 Three transformations, 263
Reversal operations, 554 Time delay - Z-transform, 231
Rounding, 512 Time invariance, 51
Rule of thumb for pairing poles and zeros, 527 Time reversal - DFT, 116
Time reversal - DTFT, 88
S Time shift, 25
Sample-based processing, 579 Time shift - DFT, 113
Sampling, 40 Time shift - DTFT, 87
Sampling frequency, 40 Toolbox, 595
Sampling period, 41 Transfer function, 230, 236
Saturation addition, 518 Transformation procedure from LPF, 199
Saturation arithmetic, 518 Transform capability, 230
Scalability, 51 Transposed direct, 554
Scaling, 48 Transposed direct form II, 559
Second moment, 513 Transposition, 554
Second-Order Section (SOS), 526 Transversal, 553
Selective factor, 438 Transversal structure, 553
Index 677

Triangular window, 304 Variance, 513

Truncated impulse response from DTFT, 373 Visual spectral resolution, 108
Truncated LPF, 182
Truncation, 512 W
Tukey (tapered cosine) window, 293 Weight values, 4
Type I, 285 White noise, 461
Type I Chebyshev filter, 406 Wilkinson’s polynomial, 521
Type II, 285 Window Designer from MATLAB, 309
Type II Chebyshev filter, 416 Windowing methods, 291
Type II FIR filter, 316 Window shape, 8
Type II HPF, 323 Word length, 619
Type III, 285
Type III and type IV FIR filter, 334 Y
Type III FIR filter, 328 Yule-walker method - approximated, 447
Type IV, 285
Type IV FIR filter, 332 Z
Z, 204
U Z domain transformation, 477
Unbiased estimation, 514 Zero, 221
Underflow, 520 Zero group delay FIR filter, 285
Unit of frequency, 21 Zero padding, 102, 108
Unit-step function, 47 Zero reflection, 389
Untrimmed LPF, 182 Z-transform, 203
Z-transform properties, 208
V Z-transforms for fundamentals, 206
Vandermonde matrix, 354