1.

4 Fourier Analysis
The subject of Fourier analysis is essential for describing certain types of systems and their
properties in frequency domain. It is concerned with representing a signal as weighted
superposition of complex sinusoids. The complex sinusoids are basic signals that can be used to
construct a broad and useful class of signals. There are four distinct Fourier representations:

- Continuous-time Fourier series,

- Discrete-time Fourier series,
- Continuous-time Fourier transform,
- Discrete-time Fourier transform.
Fourier Series (FS) applies to continuous-time periodic signals. Discrete Time Fourier Series (DTFS)
applies to discrete-time periodic signals. Nonperiodic signals have Fourier Transform (FT) representation.
Discrete Time Fourier Transform (DTFT) applies to a signal that is discrete in time and non-periodic.

EE316 Engineering Analysis 39

 1.4.1 Fourier Series (FS)
Fourier series converts the time-domain signals into frequency-domain (or spectral)
representations. Fourier series is a means to represent a periodic signal as an infinite sum of
sine wave components. A periodic signal is just a signal that repeats its pattern at some period.
The primary reason that we use Fourier series is that we can better analyze a signal in another
domain rather in the original domain. A continuous-time signal x(t) to be periodic if there is a
positive nonzero value of T for which

The fundamental period To of x(t) is the smallest positive value of T for which above Eq. is
satisfied, and l/To = fo is referred to as the fundamental frequency. Two basic examples of
periodic signals are the real sinusoidal signal and the complex exponential signal

𝒋𝝎𝟎 𝒕
𝟎

where ωo= 2π/ To = 2π fo is called the fundamental angular frequency.

Review Example/
What are the magnitude, frequency and phase of the following sinusoidal signal?
x(t)=10 cos (10πt+π/4)?

Solution: A=10 ; f = 5 Hz , π/4,

EE316 Engineering Analysis 40
 Complex Exponential Fourier Series Representation:
The complex exponential Fourier series representation of a periodic signal x(t) with
fundamental period To is given by

𝒋𝒌𝝎𝒕
𝒌
𝒌
where ck are known as the complex Fourier coefficients and are given by
𝑻/𝟐
𝒋𝒌𝝎𝒕
𝒌
𝑻/𝟐

𝑻
Setting k=0;
𝟎
𝟎

which indicates that 𝟎 equals the average value of x (t) over a period. When x(t) is real, then
∗
𝒌 𝒌
where the asterisk indicates the complex conjugate.

To find the Fourier coefficients of periodic x(t):

1. - For a given k, multiply x(t) by 𝒋𝒌𝝎𝒕 , and take the area under the curve (dividing by T ).
2. - Repeat step (1) for all k.

EE316 Engineering Analysis 41

 Example:- Fourier Series
Example : Determine the complex exponential Fourier series representation for the following
signal: x(t)= cos(ωt)
Solution : we can write the cos signal as below and compare it with general equation

The complex Fourier coefficients for cos(ωt) are

𝟏 𝟏 , 𝒌 for k ≠ ±1

Example : Find FS for x(t)= sin(3πt+ π/2)

𝒋𝒌𝝎𝒕
Compare with 𝒌
𝒌

Note that : x(t)=sin(3πt+π/2)=cos(3πt).

EE316 Engineering Analysis
42
 Trigonometric Fourier Series:
The trigonometric Fourier series representation of a periodic signal x(t) with fundamental
period T , is given by
𝟎
𝒌 𝒌
𝒌 𝟏

where ak , and bk , are the Fourier coefficients given by

𝑻 𝑻 𝑻
𝟎 𝒌 𝒌
𝟎 𝟎 𝟎

The coefficients ak , and bk , and the complex Fourier coefficients ck , are related by
𝒂𝟎
𝟎 , 𝒌 𝒌+ 𝒌 𝒌 𝒌 − 𝒌
𝟐

From these equations we can obtain

𝒌 𝒌 − 𝒌 𝒌 𝒌 + 𝒌

When x(t) is real, then ak , and bk , are real and we have

𝒌 𝒌 𝒌 𝒌]

EE316 Engineering Analysis 43

 Fourier Series for Even and Odd Periodic Signals
Even and Odd Signals:
If a periodic signal x(t) is even, then bk = 0 and its Fourier series contains only cosine terms:

𝟎
𝒌
𝒌 𝟏
If x(t) is odd, then ak = 0 and its Fourier series contains only sine terms:

𝒌
𝒌 𝟏

Example: Find the Fourier Series for x(t)

Solution : x(t) is even, then bk = 0

T=2π , =1

𝝅/𝟐
𝟏 𝒔𝒊𝒏( )
𝒌 => 𝒌 𝟏
𝝅/𝟐 𝟐

EE316 Engineering Analysis 44

 Fourier Series - Example
Example: Find the Complex Exponential FS coefficients for x(t) shown
𝑻/𝟐 1
Solution: 𝒌
𝒋𝒌𝝎𝒕
𝑻/𝟐

EE316 Engineering Analysis 45

 Harmonic Form Fourier Series:
Another form of the Fourier series representation of a real periodic signal x(t) with
fundamental period T is the harmonic form Fourier series of x(t) which is given by

𝟎 𝒌 𝒌
𝒌 𝟏

The term Co is known as the dc component, and the term Co cos(kωt- θk) is referred to as the
kth harmonic component of x(t). The first harmonic component C1 cos(ω t- θ1) is commonly
called the fundamental component because it has the same fundamental period as x(t). The
coefficients Ck and the angles θk are called the harmonic amplitudes and phase angles,
respectively, and they are related to the Fourier coefficients ak , and bk , by

𝟎 𝟏 𝒌
𝟎 𝟐 𝟐 𝒌
𝒌 𝒌 𝒌 𝒌

Note : The FS coefficients are known as a frequency domain representation of x(t) because it
is the coefficient associated with complex sinusoid at frequency kω.

EE316 Engineering Analysis 46

 Amplitude and Phase Spectra of a Periodic Signal
The complex form for Fourier series shown below is more general and usually more convenient

𝒋𝒌𝝎𝒕
𝒌
𝒌

𝑻/𝟐
𝒋𝒌𝝎𝒕
𝒌
𝑻/𝟐

Let the complex Fourier coefficients 𝒌 in above Eq. be expressed as

𝒋𝝋𝒌
𝒌 𝒌

Where a plot of | 𝒌 | versus the angular frequency ω is called the amplitude spectrum of the
periodic signal x (t) ,and a plot of versus ω is called the phase spectrum of x(t). Since the
index k assumes only integers, the amplitude and phase spectra are not continuous curves but
appear only at the discrete frequencies kω . They are therefore referred to as discrete
frequency spectra or line spectra. For a real periodic signal x(t) we have 𝒌 = 𝒌 * . Thus

𝒌 𝒌 𝒌 𝒌

Hence, the amplitude spectrum is an even function of ω, and the phase spectrum is an odd
function of ω for a real periodic signal.
EE316 Engineering Analysis 47
 Amplitude and Phase Spectra of a Periodic Signal
Example Calculate the FS coefficients for the impulse train and plot the magnitude of its
frequency spectrum (which are simply the Fourier Series coefficients).

Solution

EE316 Engineering Analysis 48

 Parseval’s theorem
Power Content of a Periodic Signal:

As mentioned before, the average power of a periodic signal x(t) over any period can be
written as
𝑻/𝟐
𝟐
𝑻/𝟐

If x(t) is represented by the complex exponential Fourier series, then it can be shown that
𝑻/𝟐
𝟐 𝟐
𝒌
𝑻/𝟐 𝒌

Above equation is called Parseval's identity (or Parseval’s theorem) for the Fourier series.

EE316 Engineering Analysis 49

 Example
Example/ Determine the FS coefficients for the signal x(t) depicted in Figure below. Derive
analytical expressions and plot, the double-sided magnitude, phase, and power spectra.
Solution:
Step 1: Determine T and ω.
Step 2: Solve for 𝑪𝒌 .
Step 3: Plot magnitude, phase and power spectrum

T = 2 sec, ω = 2π/T= π rad

𝑻 𝟐 𝟐
𝒋𝒌𝝎𝒕 𝟐𝒕 𝒋𝒌𝝅𝒕 (𝟐 𝒋𝒌𝝅)𝒕
𝒌
𝟎 𝟎 𝟎
𝟐
𝟏 𝒆 (𝟐 𝒋𝒌𝝅)𝒕 𝟏 𝒆 𝟒 𝒋𝟐𝒌𝝅 𝟏 𝟏 𝒆 𝟒 ±𝒋𝟐𝒌𝝅
𝒌 𝟐 𝟐 𝒋𝒌𝝅 𝟐 𝟐 𝒋𝒌𝝅 𝟒 𝒋𝟐𝒌𝝅
𝟎

Analytical expressions for magnitude and phase spectra:

𝟒 𝟒 𝟒
𝒌 𝟐 𝟐

𝟒 𝟒

𝟐 𝟐 𝟐 𝟐

EE316 Engineering Analysis 50

Note that
 Example-
𝟒 𝟒
𝟏 − 𝒆 𝟏−𝒆
𝑴𝒂𝒈𝒏𝒊𝒕𝒖𝒅𝒆 = 𝑪𝒌 = 𝑹𝒆𝟐 + 𝑰𝒎𝟐 = =
𝟒 + 𝒋𝟐𝒌𝝅 𝟒𝟐 + (𝟐𝒌𝝅)𝟐
𝟒 𝟒
𝟏
𝑰𝒎 𝟏
𝒆 𝟐𝝅𝒌 − 𝟐𝝅𝒌 𝟏
−𝟐𝝅𝒌 ∗ 𝟏 = 𝒆 ) 𝟏
𝝅𝒌
𝑷𝒉𝒂𝒔𝒆 = 𝝋𝒌 = 𝒕𝒂𝒏 ( ) = 𝒕𝒂𝒏 = 𝒕𝒂𝒏 ( ) = 𝒕𝒂𝒏 ( )
𝑹𝒆 𝟒 − 𝟒𝒆 𝟒 𝟒 𝟏−𝒆 𝟒 𝟐

Magnitude spectrum |ck |

Phase spectrum arg{ ck }= tan⁻¹ (ck) = tan⁻¹ (2πk/4)
Power spectrum |ck |2

51
EE316 Engineering Analysis
 Properties of Continuous –time Fourier series:
Properties are important features of transform. They make the transform easier to work with,
and thus extend the transform’s utility. Since transform all do the same thing (i.e. decompose a
general signal into a linear combination of basic signals), properties for different transforms are
often similar. So learning a set of properties for one transform directly contributes to
understanding and employing other transforms.

Symmetry: For real-valued periodic x(t), with FS coefficients ak, we have that

That is, the CT FS coefficients are symmetric

Linearity: Consider two signals, x1(t) and x2(t), each periodic with period T , with CTFS
coefficients ak1, ak2 respectively. Given any two constants c1 and c2, the CTFS pair is

Time Shift (i.e. Delay): Given a signal x(t), periodic with period T , with CTFS coefficients
ak, then for any delay τ, x(t − τ) is also periodic with period T with CTFS pair

Note that here the magnitude of the series coefficients are not changed, only their phases.
EE316 Engineering Analysis 52
 1.4.2 Discrete Time Fourier Series (DTFS)
The discrete Fourier series representation of a periodic sequence x[n] with fundamental period
No (No is the number of data points in the data set) is given by
𝑵𝟎 𝟏
𝒋𝒌Ω𝒏
𝒌
𝒌 𝟎 𝟎

where Ck , are the Fourier coefficients and given by

𝑵𝟎 𝟏
𝒋𝒌Ω𝒏
𝒌
𝟎
𝒏 𝟎

The DTFS coefficients Ck are called frequency-domain representation for x[n] since each
coefficient is associated with a complex sinusoid of a different frequency. Setting k = 0,
𝑵𝟎 𝟏

𝟎
𝟎
𝒏 𝟎

which indicates that c0 equals the average value of x[n] over a period. The Fourier
coefficients Ck are often referred to as the spectral coefficients of x[n]. The Fourier series
coefficients Ck are periodic with fundamental period N0.

𝒌 𝑵𝟎 𝒌

EE316 Engineering Analysis 53

 DTFS
Example: Determine the DTFS for the following signal

Solution
The signal has a period N = 5, hence Ω0 =2π/ 5.
Also the signal has odd symmetry, hence we
can sum over n=-2 to n =2 in Eq. below to get
𝑵𝟎 𝟏 𝟐
𝟐𝝅𝒏
𝒋𝒌Ω𝒏 𝒋𝒌
𝒌 𝟓
𝟎
𝒏 𝟎 𝒏 𝟐

Substituting for x[n] from above Fig., we get

From this equation, one period of the DTFS coefficients ck for k =-2 to k = 2 are
54
EE316 Engineering Analysis
 DTFS- Example

Hence, the magnitude and phase spectrums

of 𝒌 are as shown

EE316 Engineering Analysis 55

 DTFS
Example Determine the Fourier coefficients for the periodic sequence x[n] shown in Figure

𝑵𝟎 𝟏
Solution: 𝒌
𝒋𝒌Ω𝒏
𝟎
𝒏 𝟎

𝟑
𝟏
𝑪𝒌 = 𝒙[𝒏] (−𝒋)𝒌𝒏
𝟒
𝒏 𝟎

EE316 Engineering Analysis 56

 DTFS- Example

EE316 Engineering Analysis 57

 Properties of discrete Fourier series
Periodicity of Fourier Coefficients: The Fourier series coefficients ck are periodic with
fundamental period N0.
𝒌 𝑵𝟎 𝒌

Symmetry or Duality: Symmetry property of the discrete Fourier series is given by

Even and Odd Sequences:

Other Properties: When x[n] is real, then it follows that

EE316 Engineering Analysis 58
 Parseval's Theorem: DTFS
If x[n] is represented by the discrete Fourier series, then it can be shown that

which is called Parseval's identity (or Parseval’s theorem) for the discrete Fourier series.

Example

𝑵𝟎 𝟏 𝑵𝟎 𝟏
𝟐 𝟐
𝒌
𝟎
𝒏 𝟎 𝒏 𝟎

Note that EE316 Engineering Analysis 59