University of Kufa

Faculty of Engineering
Electronics and Communications Department

im
Preliminaries: Foundational Mathematical Concepts for
Advanced Wireless Communications

lK
Assist. Prof. Dr. -Eng. Haider Al Kim

A
August 31, 2025

d er
ai
Introduction

H
This chapter introduces the mathematical concepts essential for understanding advanced topics in
wireless communications. We will cover complex numbers, linear algebra, probability theory, Fourier
and Laplace transforms, and differential equations. Each topic includes examples that demonstrate
g.
their relevance to wireless communication systems. Mastery of these concepts will enable students
to analyze and design modern communication systems.
En

Contents
1 Complex Numbers and Phasors 3
.-

1.1 Definition of Complex Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.2 Complex Conjugate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Magnitude and Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Dr

1.4 Phasors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2 Linear Algebra 5
2.1 Finite Field (Galois field) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
f.

2.2 Vector Spaces and Subspaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2.3 Basis and Dimension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
o

2.4 Orthogonality and Projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Pr

2.5 Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.6 Matrix Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.7 Invertible and Nonsingular Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.8 Eigenvalues and Eigenvectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
t.

2.9 Matrix Decompositions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

sis

3 Probability Theory 11
3.1 Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.2 Log-Likelihood Ratio (LLR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
As

1
 3.3 Difference between Probability and LLR . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.4 Random Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.5 Probability Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

im
3.6 Expectation and Variance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.7 Common Distributions: more reading as H/W . . . . . . . . . . . . . . . . . . . . . . 17
3.8 Joint Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

lK
4 Fourier and Laplace Transforms 18
4.1 Fourier Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.2 Convolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

A
4.3 Linear Time Invariance (LTI) System . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.4 Laplace Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

er
5 Differential Equations 21
5.1 Basic Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

d
5.2 First-Order Linear Differential Equations . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3 Second-Order Linear Differential Equations . . . . . . . . . . . . . . . . . . . . . . . 21

ai
A Example of a Subspace of R2 23

H
B First-Order Linear Differential Equations 23

C Sin Cos and Exponential Formulas 25

g.
En
.-
Dr
o f.
Pr
t.
sis
As

2
 1 Complex Numbers and Phasors
1.1 Definition of Complex Numbers

im
A complex number z is defined as:
z = a + jb
√

lK
where a is the real part, b is the imaginary part, and j = −1.
Example: Consider the complex number z = 3 + 4j. Here, a = 3 and b = 4.

A
1.2 Complex Conjugate
The complex conjugate of z is denoted by z ∗ and is defined as:

er
z ∗ = a − jb

d
Example: For z = 3 + 4j, the complex conjugate is z ∗ = 3 − 4j.

ai
1.3 Magnitude and Phase

H
The magnitude of a complex number z is:
p
|z| = a2 + b2
g.
The phase (or argument) of z is:  
b
En
θ = arctan
a
Example: For z = 3 + 4j, p
|z| = 32 + 42 = 5
.-

The phase θ = arctan 3 ≈ 53.13◦ . This means that the vector representing z = 3 + 4j makes an
4

angle of about 53.13◦ with the positive real axis.

Dr

More Details on arctan

The arctan function returns angles in radians by default, but it can be converted to degrees by
◦
multiplying the result by 180
f.

π .
The arctan function only gives correct phase angles in the range − π2 to π2 (or −90◦ to 90◦ ).
o

Therefore, care must be taken when determining the phase of complex numbers in different quad-
Pr

rants of the complex plane.

For instance:
ˆ First Quadrant: a > 0, b > 0 (as in our example, so no adjustments are necessary).
t.

ˆ Second Quadrant: a < 0, b > 0 (need to add π or 180◦ to the result of arctan).
sis

ˆ Third Quadrant: a < 0, b < 0 (add π or 180◦ ).

ˆ Fourth Quadrant: a > 0, b < 0 (no adjustment needed, but you may get a negative angle
which can be adjusted to a positive equivalent by adding 360◦ ).
As

3
 In practical applications, the magnitude and phase of complex numbers are crucial for under-
standing sinusoidal signals, impedance in AC circuits, and signal modulation in wireless communi-
cations.

im
1.4 Phasors

lK
Phasors provide a powerful way to represent sinusoidal signals in a simplified form, making them
easier to manipulate, especially when dealing with linear systems and circuits.
Consider a general sinusoidal signal in the time domain:

A
x(t) = A cos(ωt + θ)

where:

er
ˆ A is the amplitude,

d
ˆ ω is the angular frequency, and

ai
ˆ θ is the phase angle.

H
To convert this sinusoidal signal into its phasor form, we follow these steps:

Step 1: Express the Cosine as a Complex Exponential

g.
Using Euler’s formula 1 , we know that a cosine function can be expressed as the real part of a
complex exponential:
En
 
cos(ωt + θ) = ℜ ej(ωt+θ)

Thus, the time-domain signal becomes:

.-

 
x(t) = ℜ Aej(ωt+θ)
Dr

Step 2: Factor Out the Time-Dependent Component

In phasor notation, we focus on the part of the expression that does not vary with time: the complex
constant Aejθ . The time-dependent component ejωt is typically factored out, leaving us with:
f.

X = Aejθ
o
Pr

Here, X is the phasor representation of the sinusoidal signal. It encapsulates the amplitude A
and the phase θ, while the time dependency has been separated into a standard exponential term
ejωt .
1 Euler’s formula states that ejθ = cos(θ) + j sin(θ). It is a fundamental relation in complex analysis, showing the
t.

deep connection between trigonometric functions and exponential functions. Additionally, Euler’s formula allows us
to express sinusoidal functions as real and imaginary parts of complex exponentials:
sis

cos(θ) = ℜ(ejθ ), sin(θ) = ℑ(ejθ )

As

4
 Step 3: Phasor Interpretation
The phasor X = Aejθ can be interpreted as a vector in the complex plane2 , where:

im
ˆ The magnitude |X| = A represents the amplitude of the sinusoid.

ˆ The angle arg(X) = θ represents the phase shift.

lK
Phasors simplify the analysis of AC signals and systems by allowing us to work with algebraic
expressions in the complex plane instead of trigonometric functions. Once the analysis is complete,
we can convert back to the time domain by taking the real part of the product of the phasor and

A
the time-dependent term ejωt .

2 Linear Algebra

er
2.1 Finite Field (Galois field)

d
ai
A (finite) field (also called a Galois field) is a field with a finite number of elements. The most
common example of a finite field is Fp , where p is a prime number, consisting of the integers
0, 1, . . . , p − 1, with addition and multiplication defined modulo p. Finite fields are widely used in

H
error correction codes and cryptography.

Arithmetic Operations Modulo p

g.
Arithmetic operations modulo p are defined as follows for a prime number p:
En

ˆ Addition Modulo p: For two integers a and b, the sum modulo p is given by:

(a + b) mod p = r, where r is the remainder when a + b is divided by p.

.-

ˆ Multiplication Modulo p: For two integers a and b, the product modulo p is given by:
Dr

(a · b) mod p = r, where r is the remainder when a · b is divided by p.

ˆ Subtraction Modulo p: For two integers a and b, the difference modulo p is given by:
f.

(a − b) mod p = r, where r is the remainder when a − b is divided by p.

o

5 5 5
6 6 6
Pr

4 4 4

0 0 0

3 3 3
t.

1 1 1
2 2 2
sis

3 3 + 3 mod 7 = 6 6 + 2 mod 7 = 1
2 Thecomplex plane is a two-dimensional plane where complex numbers are represented. The horizontal axis
represents the real part, and the vertical axis represents the imaginary part. Every complex number corresponds to
As

a unique point in this plane.

5
 2.2 Vector Spaces and Subspaces
A vector space V is a collection of vectors that can be added together and multiplied by scalars.
Definitions and Properties

im
A vector space V over a field F is defined by a set of vectors equipped with two operations: vector
addition and scalar multiplication.
In a vector space, certain operations must satisfy the following axioms. Let V be a vector space

lK
and W be a subspace of V. These operations must hold for vectors u, v, w ∈ V and scalars a, b ∈ R
(or C for complex vector spaces).

ˆ Closure:

A
– The sum of any two vectors in V is also in V.
– Example: If u = (1, 2) and v = (3, 4) are in R2 , then u + v = (1 + 3, 2 + 4) = (4, 6) is

er
also in R2 .

d
ˆ Associativity:

ai
– The sum of vectors is associative: (u + v) + w = u + (v + w).
– Example: Let u = (1, 2), v = (3, 4), w = (5, 6) in R2 . Then (u + v) + w = (4, 6) + (5, 6) =

H
(9, 12) and u + (v + w) = (1, 2) + (8, 10) = (9, 12), both giving the same result.

ˆ Commutativity:
g.
– The sum of vectors is commutative: u + v = v + u.
En
– Example: Let u = (1, 2) and v = (3, 4) in R2 . Then u + v = (4, 6) and v + u = (4, 6),
showing that the sum is commutative.

ˆ Identity:
.-

– There exists an identity element 0 ∈ V such that u + 0 = u for all u ∈ V.

– Example: In R2 , the zero vector is 0 = (0, 0). For any vector u = (1, 2), u + 0 =
Dr

(1 + 0, 2 + 0) = (1, 2), so the identity property holds.

ˆ Distributivity:

– Scalar multiplication distributes over vector addition: a(u + v) = au + av.

f.

– Example: Let a = 2, u = (1, 2), and v = (3, 4) in R2 . Then a(u + v) = 2(4, 6) = (8, 12),
o

and au + av = (2, 4) + (6, 8) = (8, 12), satisfying the distributive property.

Pr

A subspace W of V is a subset of V that forms a vector space under the same operations (see the
Example in section A).
In signal space representation, the space of all possible signal constellations can be represented
as a vector space. For instance, in baseband communication systems, signals are represented as
t.

vectors in a high-dimensional space where each dimension represents a different signal state in the
sis

constellation diagram, or corresponds to a different frequency or time component.

As

6
 2.3 Basis and Dimension
Basis Vectors

im
A basis of a vector space V is a set of linearly independent vectors that span the entire space. Any
vector in V can be expressed as a linear combination of these basis vectors.

lK
Linear Independence
Vectors {v1 , v2 , . . . , vn } are linearly independent if the only solution to the equation

A
c1 v1 + c2 v2 + · · · + cn vn = 0

is c1 = c2 = · · · = cn = 0.

er
Dimensionality of Signal Spaces

d
The dimension of a vector space is the number of vectors in a basis for the space. In wireless com-

ai
munications, the dimension often relates to the number of distinct signals that can be transmitted
simultaneously without interference.

H
2.4 Orthogonality and Projections
Inner Product and Orthogonal Projections
g.
The inner product (or dot product) of two vectors u and v is defined as
En
n
X
⟨u, v⟩ = ui vi ,
i=1
.-

providing a measure of their alignment. Vectors are orthogonal if their inner product is zero.
Orthogonal projections involve projecting a vector onto a subspace such that the difference between
the vector and its projection is orthogonal to the subspace as illustrated in Figure 1.
Dr

Applications in MIMO Systems

In Multiple Input Multiple Output (MIMO) systems, orthogonal signaling allows for the separation
of multiple signal paths, improving data throughput and system capacity.
f.

2.5 Matrices
o

Matrices represent linear transformations between vector spaces.

Pr

 
2 x
Example: Consider the vector space R . A vector in this space can be represented as v = .
y
A matrix can transform a vector. For example, a rotation matrix R(θ) rotates vectors by an
angle θ:  
t.

cos(θ) − sin(θ)
R(θ) =
sin(θ) cos(θ)
sis

Then:   
cos θ − sin θ x
R(θ) · v =
sin θ cos θ y
As

7
 Vector Projections Idea

y
Vector Projections Algebra

pW v

im
v−

lK
v
v
pW

A
W

er
x

d
Figure 1: In the notation pW , the letter “p” stands for “projection,” and the subscript W indicates

ai
that the projection is onto the subspace W. In mathematical terms, pW v denotes the orthogonal
projection of the vector v onto the subspace W. The projection operation finds the point in W

H
that is closest to v. It effectively “drops” v onto W such that the difference between v and its
projection is orthogonal to W. g.
The result is:
En
 
x cos θ − y sin θ
R(θ) · v =
x sin θ + y cos θ
.-

2.6 Matrix Operations

Key operations on matrices include:
Dr

ˆ Addition: A + B

ˆ Multiplication: AB
f.

ˆ Transpose: AT
o

ˆ Inversion: A−1
Pr

Key operations on matrices include:

ˆ Addition: A + B, e.g.:    
1 2 5 6
t.

A= , B=
3 4 7 8
sis

   
1+5 2+6 6 8
A+B= =
3+7 4+8 10 12
As

8
 ˆ Multiplication3 : AB or A · B, e.g.:
   
1 2 5 6
A= , B=

im
3 4 7 8
   
(1 · 5 + 2 · 7) (1 · 6 + 2 · 8) 19 22
A·B= =
(3 · 5 + 4 · 7) (3 · 6 + 4 · 8) 43 50

lK
ˆ Transpose:  
1 2
A=
3 4

A
 
1 3
AT =
2 4

er
 
a b
ˆ Inversion: for any matrix A = , the inverse of A is given by:

d
c d

ai
 
−1 1 d −b
A =
det(A) −c a

H
For more details and examples check the following section.
g.
2.7 Invertible and Nonsingular Matrices
En
A matrix A is said to be invertible (or nonsingular) if there exists another matrix B such that:

AB = BA = I
.-

where I is the identity matrix. If such a matrix B exists, it is called the inverse of A and is
denoted A−1 . A matrix that does not have an inverse is called singular.
Dr

ˆ Invertible Matrix: A matrix A is invertible if its determinant is non-zero, i.e., det(A) ̸= 0.

ˆ Nonsingular Matrix: This term is often used interchangeably with invertible. A matrix is
nonsingular if it has an inverse.
f.

Example:
Consider the matrix:
o
Pr

 
1 2
A=
3 4
To determine if A is invertible, we first compute its determinant:
t.

det(A) = (1 · 4 − 2 · 3) = −2
sis

Since det(A) ̸= 0, matrix A is invertible. To find the inverse A−1 :

3 The multiplication is defined as the dot product, notated by “·” of the rows of A with the columns of B. It is
As

also called scalar product.

9
  
−1 1 d −b
A =
det(A) −c a

im
where a = 1, b = 2, c = 3, d = 4. Substituting these values:
   
1 4 −2 −2 1
A−1 = =

lK
−2 −3 1 1.5 −0.5
Thus, matrix A is invertible, and its inverse is:
 
−2 1

A
−1
A =
1.5 −0.5

er
2.8 Eigenvalues and Eigenvectors
Given a square matrix A, an eigenvector v and its corresponding eigenvalue λ satisfy the equation:

d
Av = λv

ai
 
4 1
Example: For the matrix A = , one of the eigenvalues is λ = 5, with the corresponding

H
  2 3
1
eigenvector v = .
1
g.
Eigenvalues and eigenvectors are critical in wireless communications for analyzing MIMO (Mul-
tiple Input Multiple Output) systems and the behavior of communication channels. They are also
En
critical in assessing system stability and performance in various communication algorithms.

2.9 Matrix Decompositions

.-

SVD (Singular Value Decomposition)

Singular Value Decomposition (SVD) of a matrix A is expressed as
Dr

A = UΣVT ,

where U and V are orthogonal matrices and Σ “Sigma” is a diagonal matrix containing the singular
values. In Multiple Input Multiple Output (MIMO) systems, SVD is used to decompose the channel
f.

matrix into orthogonal modes of transmission. This allows for sending multiple independent data
streams, improving the capacity and reliability of the wireless link SVD is also widely used for
o

dimensionality reduction and data compression.

Pr

QR Decomposition
QR decomposition factors a matrix A into

A = QR,
t.

where Q is an orthogonal matrix and R is an upper triangular matrix. This decomposition is useful
for solving linear systems and least squares problems.
sis

Applications in Wireless Communication Algorithms

Matrix decompositions such as SVD and QR are integral to various algorithms in wireless commu-
As

nications, including channel estimation, signal processing, and optimization techniques.

10
 3 Probability Theory
3.1 Probability

im
Probability is a measure of the likelihood of a given event occurring. For a discrete random variable
X with possible values x1 , x2 , . . . , xn , the probability of a specific value xk is defined as:

lK
P (X = xk ) = pk
Pn
where 0 ≤ pk ≤ 1 and k=1 pk = 1.

A
In a wireless communication system, consider the probability of a received signal Y being de-
coded as a binary 1 or binary 0. Let P (Y = 1) = 0.8 and P (Y = 0) = 0.2. The higher probability
of receiving a 1 indicates that the system is biased toward sending a binary 1 more often.

er
3.2 Log-Likelihood Ratio (LLR)

d
The Log-Likelihood Ratio (LLR) is a measure used in decision-making processes in digital commu-

ai
nication, particularly in decoding algorithms like those used in error-correcting codes. The LLR
compares the likelihood of two hypotheses. For binary variables, the LLR for a received signal Y

H
is defined as:
P (Y |X = 1)
LLR(Y ) = log
P (Y |X = 0)
g.
where P (Y |X = 1) is the likelihood of receiving Y given that the transmitted bit was a 1, and
P (Y |X = 0) is the likelihood of receiving Y given that the transmitted bit was a 0.
En
In the same wireless communication system, suppose the likelihoods of receiving a certain signal
Y given that a binary 1 or binary 0 was transmitted are P (Y |X = 1) = 0.9 and P (Y |X = 0) = 0.3.
The LLR would be:
0.9
.-

LLR(Y ) = log ≈ 1.585

0.3
A positive LLR value indicates that it is more likely that a binary 1 was transmitted than a binary
Dr

0.

3.3 Difference between Probability and LLR

1. Probability measures the absolute likelihood of a specific event occurring, such as receiving
f.

a binary 1 or 0.
o

2. LLR compares the relative likelihood of two competing hypotheses, such as whether a binary
Pr

1 or binary 0 was transmitted.

t.
sis
As

11
 Example: Comparison between Probability and LLR

In a mobile communication system, suppose the probability of receiving a binary 1 is P (Y =

im
1) = 0.7, and the probability of receiving a binary 0 is P (Y = 0) = 0.3.
The LLR, however, focuses on comparing how much more likely it is to receive a binary 1
over a binary 0, using likelihood ratios. If the likelihood of receiving a certain signal Y given

lK
that a binary 1 was sent is 0.8, and given that a binary 0 was sent is 0.2, then:
0.8
LLR(Y ) = log = log 4 ≈ 1.386
0.2

A
A higher LLR value indicates a stronger likelihood that a binary 1 was transmitted, even
though both probabilities and likelihoods help describe the transmission.

er
3.4 Random Variables

d
A random variable X assigns a real number to each outcome in a sample space.

ai
Consider an experiment where we roll a fair six-sided die. The sample space S for this experiment
is:

H
S = {1, 2, 3, 4, 5, 6}.
Now, let the random variable X represent the outcome of the die roll. In this case, X assigns a
real number to each outcome in the sample space. For simplicity, we will assume that the random
g.
variable X just takes the value of the die roll itself. Therefore:
En
ˆ If the outcome is 1, X=1

ˆ If the outcome is 2, X=2

..
.-

ˆ If the outcome is 6, X=6

Dr

So, the random variable X maps each outcome in the sample space S to a real number as stated in
Figure 2. There are two types:
ˆ Discrete Random Variables: Take on countable values (e.g., the outcome of a dice roll as
f.

explained above).
o

ˆ Continuous Random Variables: Take on any value within a range (e.g., signal amplitude
as shown in Figure 3).
Pr

3.5 Probability Distributions

t.

1. Discrete: A discrete random variable takes on a finite or countable number of distinct

values. The Probability Mass Function (PMF) gives the probability of each possible
sis

value of the discrete random variable. The sum of all the probabilities in the PMF is equal
to 1. Thus we state the following definition:
As

12
 Sample Space S

1 2 3 4 5 6

im
lK
X=1 X=2 X=3 X=4 X=5 X=6

A
. . . . . . . . . . . . . . . . . . . . . Random Variable X . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

er
Figure 2: Sample space S of six elements that are mapped, by the random variable X, into a real
1
number with an equal probability of (equi-likely).

d
6

ai
Amplitude X(t)

H
Max X(t) = A
High X(t)
Probability Mass Function
g.
Time
En
.-

Min X(t) = −A Low X(t)

Figure 3: Visualization of a continuous random variable X(t), representing the amplitude of a

Dr

signal over time. The signal can take any value within the continuous range [−A, A]. This smooth
waveform illustrates that X(t) can assume any amplitude along the curve, showing the continuous
nature of the random variable.
f.

Definition: The Probability Mass Function (PMF)

o
Pr

Let X be a discrete random variable with range RX = {x1 , x2 , x3 , . . . } (finite or

countably infinite). The function PX (xk ) = P (X = xk ) for k = 1, 2, 3, . . . is called the
probability mass function (PMF) of X.
t.

Consider the experiment of tossing a fair coin twice. The sample space is S = {HH, HT, T H, T T },
where H represents heads and T represents tails. Define the random variable X as the number
sis

of heads observed. The possible values of X are RX = {0, 1, 2}, where:

ˆ X = 0 corresponds to observing no heads (outcome T T )

As

13
 Probability Density Function

ˆ X = 1 corresponds to observing one head (outcomes HT or T H)

ˆ X = 2 corresponds to observing two heads (outcome HH)

im
The PMF of X is given by:
1 2 1 1
P (X = 0) = , P (X = 1) = = , P (X = 2) = .
4 4 2 4

lK
Figure 4 depicts the PMF of a fair coin twice [1].

Probability P (X = k)

A
er
1
2
1

d
4

ai
Value of X
0 1 2

H
Figure 4: Visualization of the Probability Mass Function (PMF) of the discrete random variable
X, representing the number of heads observed when tossing a fair coin twice. The PMF gives the
probabilities P (X = 0) = 14 , P (X = 1) = 24 , and P (X = 2) = 14 .
g.
2. Continuous: The Probability Density Function (PDF) represents the distribution of proba-
En
bilities across a continuum of values.

Definition: The Probability Density Function (PDF)

.-

Consider a continuous random variable X with an absolutely continuous CDF FX (x).

The function fX (x) defined by
Dr

dFX (x) ′
fX (x) = = FX (x),
dx
if FX (x) is differentiable at x, is called the probability density function (PDF) of X.
o f.

A continuous random variable takes on an infinite number of possible values within a given
Pr

range. The Probability Density Function (PDF) represents the distribution of probabilities
across a continuum of values. Unlike the Probability Mass Function (PMF) for discrete random
variables, the PDF does not give the probability of a specific value but rather the density of prob-
ability in a small interval.
Example: The PDF is the Standard Normal Distribution or what is known (Gaussian Distribu-
t.

tion).
sis

Consider a continuous (Gaussian) random variable X that follows a standard normal distribu-
tion. The PDF of X is given by:
1 x2
fX (x) = √ e− 2
As

2π

14
 This function describes the well-known bell curve4 , where the highest probability density is at the
mean (0) and the density decreases symmetrically as you move away from the mean.

im
Remark
The probability is represented by the area under the normal distribution curve, with the
entire area summing to one.

lK
Density fX (x)

A
√1
2π

er
Value of X
-3 -2 -1 0 1 2 3

d
Figure 5: Visualization of the Probability Density Function (PDF) of a standard normal distribu-

ai
tion. The bell-shaped curve represents the distribution of probability density across the continuum
of values, with the highest density at the mean (0).

H
3.6 Expectation and Variance
g.
Expectation (Mean)
En
The expectation or mean of a random variable X, denoted E[X], is the long-run average value of
the random variable over many independent realizations of the experiment. For a discrete random
variable with probability mass function P (X = xk ), the expectation is defined as:
.-

X
E[X] = xk · P (X = xk )
k
Dr

For a continuous random variable with probability density function fX (x), the expectation is given
by: Z ∞
E[X] = x · fX (x) dx
−∞
f.

Example: the Mean

o

Suppose you roll a fair six-sided die. Let X be the outcome of the roll. The expectation
Pr

E[X] is calculated as:

6 6
X X k 1+2+3+4+5+6
E[X] = k · P (X = k) = = = 3.5
6 6
t.

k=1 k=1
sis

Hence, the expected value of the outcome of a fair die roll is 3.5.

4 The bell curve, or Gaussian distribution, is a symmetric, bell-shaped graph where data points are most frequent
As

near the mean, gradually decreasing in frequency as they move away from the center.

15
 Properties of Expectation
1. Linearity: The expectation operator is linear, which means that for any random variables X
and Y , and constants a and b:

im
E[aX + bY ] = aE[X] + bE[Y ]

lK
2. Expectation of a Constant: If c is a constant, then:

E[c] = c

A
Variance
The variance of a random variable X, denoted Var(X), measures the spread or dispersion of the

er
random variable around its mean. It is defined as:

Var(X) = E[(X − E[X])2 ]

d
For a discrete random variable, the variance is:

ai
X
Var(X) = (xk − E[X])2 · P (X = xk )

H
k

For a continuous random variable, the variance is:

Z ∞
g.
Var(X) = (x − E[X])2 · fX (x) dx
−∞
En

Example: the Variance

Continuing with the example of rolling a fair six-sided die, the variance Var(X) is calculated
.-

as:
6
1 (1 − 3.5)2 + (2 − 3.5)2 + · · · + (6 − 3.5)2
Dr

X 2
Var(X) = (k − 3.5) · = = 2.92
6 6
k=1

Thus, the variance of the outcome of a fair die roll is 2.92.

o f.

Properties of Variance
Pr

1. Non-Negativity: Variance is always non-negative:

Var(X) ≥ 0

2. Variance of a Constant: If c is a constant, then:

t.

Var(c) = 0
sis

3. Scaling: If a is a constant, then:

Var(aX) = a2 Var(X)
As

16
 Relation Between Expectation and Variance
Variance is closely related to expectation. In fact, an alternative formula for variance is:

im
Var(X) = E[X 2 ] − (E[X])2
This formula can sometimes be easier to use when calculating the variance, as it requires computing
only the expectations E[X 2 ] and E[X].

lK
Example: the Mean and the Variance

For a Gaussian distribution with mean µ = 0 and variance σ 2 = 1, the expectation is

A
E[X] = 0 and the variance is Var(X) = 1.

er
3.7 Common Distributions: more reading as H/W Joint (PMF)

d
In wireless communications, the following distributions are commonly encountered:
ˆ Gaussian Distribution: Models noise in communication systems.

ai
ˆ Rayleigh Distribution: Used to model multipath fading in wireless channels.

H
ˆ Rician Distribution: Used to model multipath fading in wireless channels with a dominant
line-of-sight (LOS) path.
g.
ˆ Exponential Distribution: Describes the time between events in a Poisson process5 , often
used in modeling delay or inter-arrival times.
En
Example: A Rayleigh distribution models the amplitude of a signal in a multipath environment,
Joint (CDF)
where there is no direct line of sight between the transmitter and receiver. This topic will be
extensively discussed later in this course.
.-

3.8 Joint Probability

Dr

In probability theory, the concept of joint probability refers to the probability of two events
occurring together. If X and Y are two random variables, their joint probability distribution
captures the likelihood of specific combinations of values for X and Y .
f.

Joint Probability Mass Function (PMF)

o

For discrete random variables, the joint probability mass function (PMF) is defined as:
Pr

pX,Y (x, y) = P (X = x, Y = y)
The PMF gives the probability that X takes the value x and Y takes the value y simultaneously.
The PMF satisfies the following properties:
t.

ˆ pX,Y (x, y) ≥ 0 for all x, y.

ˆ
sis

P P
x y pX,Y (x, y) = 1.
5 A Poisson process is a type of stochastic process where events occur continuously and independently of each

other at a constant average rate. It is often used in modeling events such as the arrival of calls at a call center or
As

the occurrence of natural phenomena like earthquakes.

17
 Joint Cumulative Distribution Function (CDF)
For both discrete and continuous random variables, the joint cumulative distribution function (CDF)
is given by:

im
FX,Y (x, y) = P (X ≤ x, Y ≤ y)
This function represents the probability that X is less than or equal to x and Y is less than or

lK
equal to y. The joint CDF has the following properties:
ˆ 0 ≤ FX,Y (x, y) ≤ 1 for all x, y.
ˆ FX,Y (x, y) is non-decreasing in both x and y.

A
ˆ As x or y → ∞, FX,Y (x, y) → 1.

er
Example
Consider two discrete random variables X and Y that represent the outcomes of rolling two six-sided

d
dice. The joint PMF can be defined as:

ai
1
pX,Y (x, y) = P (X = x, Y = y) = , x, y ∈ {1, 2, 3, 4, 5, 6}
36

H
The joint CDF can be calculated by summing the probabilities of the events where X ≤ x and
Y ≤ y.
g.
ˆ For example, FX,Y (3, 4) is the sum of all probabilities where X ≤ 3 and Y ≤ 4.
En

4 Fourier and Laplace Transforms

.-

4.1 Fourier Transform

Dr

The Fourier transform of a time-domain signal x(t) can also be denoted by F(x(t)), and is defined
as: Z ∞
X(f ) = F(x(t)) = x(t)e−j2πf t dt
−∞
f.

This operation converts a time-domain signal into its frequency-domain representation, which is Fourier Transform Duality

crucial for analyzing bandwidth, filtering, and modulation.

o

Similarly, the inverse Fourier transform is denoted by F −1 (X(f )) and allows you to recover the
Pr

time-domain signal x(t) from its frequency-domain representation X(f ):

Z ∞
x(t) = F −1 (X(f )) = X(f )ej2πf t df
−∞
t.

The inverse transform is essential for reconstructing the original signal from its frequency compo-
nents. Note that we call 2πf as ω which is the angular6 frequency.
sis

6 In signal processing and physics, ω represents angular frequency and is measured in radians per second. It is

related to the standard frequency f (measured in Hertz) by the formula ω = 2πf . Angular frequency is commonly
As

used in the context of continuous-time signals and oscillatory phenomena such as sine and cosine waves.

18
 Example: For a rectangular pulse signal x(t) of duration T , the Fourier transform is:

X(f ) = T sinc(f T )

im
This illustrates the trade-off between time-domain duration and frequency-domain bandwidth.
To illustrate the Fourier transform of a rectangular pulse signal x(t) of duration T , we can create
a figure (Figure 6) with two plots: one showing the rectangular pulse in the time domain and the

lK
other showing its Fourier transform in the frequency domain.

x(t) Fourier Transform 1 X(ω)

1

A
Sinc function
Inverse Fourier Transform
T

er
ω
t

d
−10 10
−1 1

ai
Figure 6: Fourier transform and inverse Fourier transform of a rectangular pulse signal x(t) of

H
duration T and Sinc signal X(ω) of angular frequency ω.
g.
4.2 Convolution Convolution Animation
En
Convolution is a fundamental operation in signal processing and linear systems, used to deter-
mine the output of a system when the input and the system’s impulse response are known. The
convolution of two signals x(t) and h(t) is denoted by x(t) ∗ h(t) and is defined by the following
integral:
.-

Z ∞
y(t) = x(t) ∗ h(t) = x(τ )h(t − τ )dτ
Dr

−∞

In simple terms, convolution “slides” one function over another, multiplies them point by point,
and integrates the result to produce the output signal.
f.

Example of Convolution
o

Consider an example where the input x(t) is a rectangular pulse of width 2 units, and the impulse
response h(t) is an exponentially decaying function. The convolution of these two functions will
Pr

result in the output signal y(t), which is a smoothed version of the rectangular pulse.
Figure 7 illustrates the convolution of a rectangular pulse x(t) with an exponentially decaying
impulse response h(t). The resulting output signal y(t) is shown as well.
t.

4.3 Linear Time Invariance (LTI) System

sis

Before introducing the Laplace transform in the next section, we will explain Linear Time-
Invariant (LTI). It is a signal processing and control theory system that satisfies two key proper-
As

ties: linearity and time invariance.

19
 x(t) h(t) y(t)

im
x(t) t t t

lK
(a) Input Signal x(t) (b) Impulse Response h(t) (c) Output Signal y(t) = x(t) ∗ h(t)

Figure 7: Illustration of the convolution in a time domain of a rectangular pulse x(t) with an
exponentially decaying impulse response h(t) to give the convoluted signal y(t).

A
1. Linearity: The system obeys the principles of superposition. If x1 (t) and x2 (t) are inputs,

er
and their corresponding outputs are y1 (t) and y2 (t), then for any constants a and b, the input
ax1 (t) + bx2 (t) will produce the output ay1 (t) + by2 (t).

d
2. Time Invariance: The behavior of the system does not change over time. If the input x(t)
produces an output y(t), then shifting the input by some time t0 (i.e., x(t−t0 )) will shift the output

ai
by the same amount of time (i.e., y(t − t0 )).

H
Example of an LTI System
An example of an LTI system is a low-pass filter. In a simple low-pass filter, high-frequency
g.
components of a signal are attenuated, while low-frequency components are allowed to pass through.
En
Impulse Response of an LTI System
The response of an LTI system to an impulse input (a delta function) is called the impulse response
h(t). For an arbitrary input signal x(t), the output y(t) of an LTI system can be found by convolving
.-

the input x(t) with the system’s impulse response h(t):

Z ∞
y(t) = x(t) ∗ h(t) = x(τ )h(t − τ )dτ
Dr

−∞

Figure 8 illustrates the concept of an LTI system, where an input signal x(t) passes through the
LTI system, resulting in an output signal y(t).
o f.

x(t) y(t)
LTI System
Pr

Figure 8: A block diagram representing a linear time-invariant (LTI) system.

t.
sis
As

20
 4.4 Laplace Transform
The Laplace transform is a more general form of the Fourier transform and is particularly useful in
the analysis of linear time-invariant (LTI) systems:

im
Z ∞
X(s) = x(t)e−st dt
0

lK
where s = σ + jω. The Laplace transform is useful for system stability analysis and control theory.
Example: For x(t) = e−at with a > 0, the Laplace transform is:
1
X(s) =

A
s+a

5 Differential Equations

er
5.1 Basic Concepts

d
A differential equation involves an unknown function and its derivatives. In wireless communica-

ai
tions, differential equations model dynamic systems such as filters and oscillators.

H
5.2 First-Order Linear Differential Equations
A first-order linear differential equation takes the form:
g.
dy
+ a(t)y = b(t)
dt
En
dy
Example: Consider the equation dt + 2y = e−t . The solution method involves finding an inte-
grating factor, yielding:
1 −t
y(t) = e + Ce−2t (see B)
.-

5.3 Second-Order Linear Differential Equations

Dr

Second-order differential equations appear frequently in systems with oscillations or resonance:

d2 y dy
+a + by = f (t)
dt2 dt
f.

d2 y
Example: Consider the equation dt2 + 4y = 0, representing a simple harmonic oscillator. The
o

solution is:
y(t) = C1 cos(2t) + C2 sin(2t)
Pr

This is relevant in communication systems involving resonant circuits, filters, and oscillators.

Conclusion
t.

This chapter has introduced the foundational mathematical concepts necessary for understanding
sis

wireless communication systems. These concepts will be used extensively in the following chapters,
where we will analyze and design advanced wireless systems. Mastery of these topics will allow you
As

to analyze complex communication models and optimize system performance.

21
 Assignment
Complete the following exercises to practice the concepts covered in this chapter:

im
1. Complex Numbers: Given z1 = 2 + 3j and z2 = 1 − j, compute z1 + z2 , z1 × z2 , and zz12 .
Find the magnitude and phase of z1 and z2 .

lK
 
2 1
2. Linear Algebra: Let A = . Find the eigenvalues and eigenvectors of A. Interpret
1 2
the significance of these eigenvalues in the context of a wireless MIMO system.

A
3. Probability Theory: Consider a Rayleigh-distributed random variable with parameter σ =
1. Find the mean and variance of the distribution. Plot the PDF and CDF of the distribution.

er
4. Fourier Transform: Find the Fourier transform of a square wave signal with period T . Plot
the magnitude spectrum and discuss the bandwidth implications for a communication system.

d
2
5. Differential Equations: Solve the second-order differential equation ddt2y + 5 dy
dt + 6y = 0.

ai
Find the general solution and discuss the physical meaning in the context of an RLC circuit
in a communication system.

H
Reading
g.
This is a reference to Pishro-Nik’s book [1], Strang’s book [2], and Molisch’s book [3].
For further exploration of the mathematical concepts covered in this chapter, consider the
En
following resources:
ˆ Papoulis, A. Probability, Random Variables, and Stochastic Processes.
.-

ˆ Oppenheim, A. V., and Schafer, R. W. Discrete-Time Signal Processing.

ˆ Bracewell, R. The Fourier Transform and Its Applications.

Dr

ˆ Boyce, W. E., and DiPrima, R. C. Elementary Differential Equations and Boundary Value
Problems.
f.

References
o
Pr

[1] H. Pishro-Nik, Introduction to Probability, Statistics, and Random Processes. Kappa Research
LLC, 2014.

[2] G. Strang, Linear Algebra and Its Applications, 4th ed. Thomson Brooks/Cole, 2006.
t.

[3] A. F. Molisch, Wireless Communications. John Wiley & Sons, 2011.

sis
As

22
 Appendix A: Example of a Subspace of R2
Consider the set:

im
   
1
W = λ |λ∈R
2
 
1

lK
This set consists of all scalar multiples of the vector . It represents a line through the origin in
2
R2 .
Verifying that W is a Subspace:

A
1. Zero vector: If λ = 0, we get the zero vector:
   
1 0
λ =

er
2 0
So, W contains the zero vector.

d
   
1 1

ai
2. Closed under addition: Take two vectors in W , say λ1 and λ2 . Their sum is:
2 2
     

H
1 1 1
λ1 + λ2 = (λ1 + λ2 )
2 2 2
Since (λ1 + λ2 ) is a scalar, the sum is also in W .
g.
 
1
3. Closed under scalar multiplication: If λ is in W and c ∈ R, then:
2
En
    
1 1
c λ = (cλ)
2 2
.-

Since cλ is a scalar, the result is also in W .

   
1
| λ ∈ R is a valid subspace of R2 , representing a line through the origin.
Dr

Thus, W = λ
2

Appendix B: First-Order Linear Differential Equations

f.

A first-order linear differential equation has the general form:

o

dy
+ a(t)y = b(t) (1)
Pr

dt
where y(t) is the unknown function of t, and a(t) and b(t) are known functions of t.
The standard method for solving such an equation involves finding an integrating factor that
simplifies the equation.
t.

Considering the differential equation:

sis

dy
+ 2y = e−t (2)
dt
As

By eq. (1), eq. (2) is a first-order linear differential equation with:

23
 ˆ a(t) = 2

ˆ b(t) = e−t

im
Step 1: Find the Integrating Factor
The integrating factor µ(t) is given by:

lK
R R
a(t)dt 2dt
µ(t) = e =e = e2t

Step 2: Multiply the Differential Equation by the Integrating Factor

Multiplying both sides of the differential equation by e2t gives:

A
dy
e2t + 2e2t y = e2t e−t
dt

er
This simplifies7 to:
d 2t

d
e y = et
dt

ai
Step 3: Integrate Both Sides
Integrating both sides with respect to t:

H
Z
e2t y = et dt = et + C
g.
where C is the constant of integration.
Step 4: Solve for y(t)
En
Now, solve for y(t) by dividing both sides by e2t :
1 −t
y(t) = et · e−2t + Ce−2t = e + Ce−2t
.-

2
Thus, the solution to the differential equation is:
Dr

1 −t
y(t) = e + Ce−2t
2
where C is an arbitrary constant.
f.

7 Theexpression e2t dy + 2e2t y can be rewritten as dtd

e2t y by using the product rule of differentiation. The


dt
product rule states that for two differentiable functions u(t) and v(t), the derivative of their product is given by:
o

d dv(t) du(t)
Pr

(u(t)v(t)) = u(t) + v(t) .

dt dt dt
In our case,
d 2t dy d 2t

e y(t) = e2t


+ y(t) e .
dt dt dt
d
e2t = 2e2t , we get:


Since
t.

dt
d 2t dy
e y(t) = e2t + 2e2t y(t).


dt dt
sis
As

24
 Appendix C: Sin Cos and Exponential Formulas
Trigonometric formulas are key concepts in wireless communications and signal processing. The

im
are:

ˆ sin(θ) = opposite
hypotenuse

lK
ˆ cos(θ) = adjacent
hypotenuse

sin(θ)
ˆ tan(θ) = cos(θ)

A
Euler’s formula, ejθ = cos(θ) + j sin(θ), is fundamental in connecting trigonometric and exponential
representations.

er
ˆ cos2 (θ) + sin2 (θ) = 1

d
ˆ Half-angle formulas:

ai
r
θ 1 − cos θ
sin = ±

H
2 2
r
θ 1 + cos θ
cos = ±
2 2
g.
ˆ Double Angle Formulas:
En
sin(2θ) = 2 sin(θ) cos(θ)
cos(2θ) = cos2 (θ) − sin2 (θ) = 2 cos2 (θ) − 1 = 1 − 2 sin2 (θ)
.-

ˆ Triple Angle Formulas:

Dr

sin(3θ) = 3 sin(θ) − 4 sin3 (θ)

cos(3θ) = 4 cos3 (θ) − 3 cos(θ)
f.

ˆ Half-Angle Formulas:
o
Pr

1 − cos(2θ)
sin2 (θ) =
2
2 1 + cos(2θ)
cos (θ) =
2
t.
sis
As

25
 ˆ Sum and Difference of Angles

sin(θ + ψ) = sin(θ) cos(ψ) + cos(θ) sin(ψ)

im
sin(θ − ψ) = sin(θ) cos(ψ) − cos(θ) sin(ψ)
cos(θ + ψ) = cos(θ) cos(ψ) − sin(θ) sin(ψ)
cos(θ − ψ) = cos(θ) cos(ψ) + sin(θ) sin(ψ)

lK
   
θ+ψ θ−ψ
sin(θ) + sin(ψ) = 2 sin cos
2 2
   
θ+ψ θ−ψ

A
sin(θ) − sin(ψ) = 2 cos sin
2 2
   
θ+ψ θ−ψ
cos(θ) + cos(ψ) = 2 cos cos

er
2 2
   
θ+ψ θ−ψ
cos(θ) − cos(ψ) = −2 sin sin

d
2 2

ai
ˆ Product to Sum Formulas

H
2 cos(θ) cos(ψ) = cos(θ − ψ) + cos(θ + ψ)
2 sin(θ) sin(ψ) = cos(θ − ψ) − cos(θ + ψ)
2 sin(θ) cos(ψ) = sin(θ + ψ) + sin(θ − ψ)
g.
2 cos(θ) sin(ψ) = sin(θ + ψ) − sin(θ − ψ)
En

ˆ Periodic Identities

sin π2 − x
= cos(x) cos π2 − x
= sin(x)

.-

sin π2 + x
= cos(x) cos π2 + x
= − sin(x)

sin 3π 2 − x
= − cos(x) cos 3π2 − x
= − sin(x)
sin 3π 2 + x = − cos(x) cos 3π
Dr

2 + x = sin(x)
sin(π − x) = sin(x) cos(π − x) = − cos(x)
sin(π + x) = − sin(x) cos(π + x) = − cos(x)
sin(2π − x) = − sin(x) cos(2π − x) = cos(x)
sin(2π + x) = sin(x) cos(2π + x) = cos(x)
o f.
Pr
t.
sis
As

26

Email: [Link]@[Link]
Phone: +9647817822030
Website: [Link]