Chapter 3.

Fourier Series Representation of

Periodic Signals
• 3.0 Introduction
• Representation of signals as linear combinations of shifted impulses (Chapter 2)
à An LTI system’s response: convolution of an input with unit impulse response
• Representation of periodic signals as linear combination of complex
exponentials (Chapter 3) à Fourier series

• Representation of aperiodic, finite energy signals as linear combination of

complex exponentials (Chapter 4) à Fourier transform
• 3.1 A Historical Perspective

• 3.2 The Response of LTI Systems to Complex Exponentials

• Exponential functions are eigenfunctions.
• The system output as a constant times the input is referred to as an
eigenfunction of the system, and the amplitude factor is referred to as the
system’s eigen value (c.f. eigenvector and eigenvalue in linear algebra)
 Continuous-Time: est H(s)est
LTI system
Discrete-Time: zn H(z)zn

• Derivation in C-T:

• H(s) for a specific value of s is the eigenvalue associated with the eigenfunction est.

• Derivation in D-T:

• H(z) for a specific value of z is the eigenvalue associated with the eigenfunction zn.
• The LTI system’s response to a linear combination of complex exponentials can
also be represented by a linear combination of complex exponentials.
• Continuous-Time:

est H(s)est
LTI system

• Discrete-Time:

zn H(z)zn
LTI system
• In Fourier analysis, we consider purely imaginary values of s = jω à est = ejωt in
continuous-time, and we consider z having unit magnitude z = ejω in discrete-
time.

• Example 3.1: consider an LTI system relating the input x(t) and output y(t) as:

If the input is given by x(t) = ej2t, then

, which is in the form of y(t) = H(s)est at s = j2.

• The above result can be confirmed by considering the system’s impulse response:
h(t) = 𝛿(t-3) and calculating H(s) as:

H(j2) = e-j6 y(t) = e-j6 ej2t

• Example 3.1 (continued): consider the input signal x(t) = cos(4t) + cos(7t).

• It is straightforward to think of the output:

• The above result can also be obtained by the following steps

1. By Euler’s formula

2. Using the input-output relation:

H(s)est = e-3s est at s = j4.

3.3 Fourier Series Representation of
Continuous-Time Periodic Signals
• 3.3.1 Linear Combination of Harmonically Related Complex Exponentials
• A periodic signal with a period of T satisfies:
• Fundamental period: the minimum positive, nonzero T

• Fundamental frequency: 𝜔! = 2𝜋/𝑇

• Harmonically related complex exponentials

• A linear combination of harmonically related complex exponentials:

Fourier series representation of a periodic C-T signal x(t)

• Fourier series representation:

• The term for k=0 (a0): a constant (DC component)

• The terms for k = +1 (a1) and k = -1 (a-1): fundamental component or first harmonics

• The terms for k = +2 (a2) and k = -2 (a-2): second harmonics

• The terms for k = +N (aN) and k = -N (a-N): N-th harmonic components

• Example 3.2
x(t) is a real periodic signal ß ak = a-k
• Suppose a real periodic signal x(t), then x(t) = x*(t)
𝑥 𝑡 = ∑$
!"#$ 𝑎! 𝑒
%!&! ' = 𝑥 ∗ 𝑡 = ∑$ ∗ #%!&! '
!"#$ 𝑎! 𝑒 = ∑$ ∗
!"#$ 𝑎#! 𝑒
%!&! '

∗
𝑎! = 𝑎#! ßà 𝑎!∗ = 𝑎#! Replacing k by -k

• Alternative forms of the Fourier series

polar form rectangular form

• 3.3.2 Determination of the Fourier Series Representation of a Continuous-
Time Periodic Signal
• A procedure for determining the Fourier coefficients ak

∴ or Orthogonality of
𝜙! (𝑡) = 𝑒 "!#!$

integrating over any interval of length T

• Fourier series pair:

synthesis form

analysis form

• The meaning ak (Fourier series coefficients or the spectral coefficients):

• The portion of the signal x(t) at each harmonic component

• a0: the DC component of x(t); the average value of x(t) over one period

• Example 3.3
• Example 3.4
• Example 3.5

When T = 4T1:
3.4 Convergence of the Fourier Series
• Consider an approximation of x(t) by 𝑥𝑁 𝑡 = ∑( 𝑐
'( ) 𝑒 *)+! , , and the following

minimization problem:

min 𝐸( = ∫- |𝑥 𝑡 − ∑( 𝑐
).'( ) 𝑒 *)+! , |/ 𝑑𝑡 w.r.t. c
k

• The solution of the above problem is always given by:

0
𝑐) = - ∫- 𝑥(𝑡)𝑒 '*)+! , 𝑑𝑡 = 𝑎)

• If x(t) has a Fourier series representation, then lim 𝐸( = 0

(→2
• Not all periodic signals have their corresponding Fourier series representations.
• For some periodic x(t), ak is infinite or x(t) does not converge even with finite ak for all k.

à lim 𝐸) ≠ 0
)→$

• One condition that guarantees the existence of the Fourier series representation:

∫- |𝑥 𝑡 |/𝑑𝑡 < ∞ → lim 𝐸( = 0

(→2
Finite energy over one period

• Note: lim 𝐸( = 0 does not imply that x(t) and its Fourier series representation
(→2
are equal at every value of t, but it means there is no energy in their difference.
• An alternative set of conditions (Dirichlet conditions) guarantees existence of the
Fourier series representation of a periodic signal x(t).
• Condition 1: x(t) is absolutely integrable over one period, i.e.,

• Condition 2: there are only a finite number of maxima and minima during any single period.

• Condition 3: there are only a finite number of discontinuities in any finite interval of time,
and each of these discontinuities is finite.
• Fortunately, signals not satisfying the Dirichlet conditions do not typically arise in
practical contexts.

• Hence, the question of the convergence of Fourier series is not significant to us.

• For a periodic signal without discontinuities: lim 𝑥) 𝑡 = 𝑥(𝑡) ∀ 𝑡

)→$

• For a periodic signal with a finite number of discontinuities in each period:

+ '!" ,+('!# )
lim 𝑥) 𝑡 = 𝑥(𝑡) ∀ 𝑡 except points of discontinuity (t0), where the series converges to
)→$ /

à lim 𝐸) = lim ∫0 |𝑥 𝑡 − 𝑥) 𝑡 |/ 𝑑𝑡 = 0
)→$ )→$
Gibb’s phenomenon
3.5 Properties of C-T Fourier Series
• 3.5.1 Linearity: For x(t) and y(t) with the same fundamental period,

• 3.5.2 Time shifting

ℱ4 ℱ4
𝑥 𝑡 𝑎) 𝑥 𝑡 − 𝑡5 𝑏) = 𝑒 '*)+! ,! 𝑎) (|𝑏) | = 𝑎) )

Derivation: =

τ = t-t0
• 3.5.3 Time reversal
ℱ$ ℱ$
𝑥 𝑡 𝑎% 𝑥 −𝑡 𝑏% = 𝑎&%

Derivation:

• 3.5.4 Time scaling

ℱ$
𝑥 𝑡 𝑎% 𝑥 𝛼𝑡 = ∑(
%'&( 𝑎% 𝑒
)%(+,! ).

• 3.5.5 Multiplication: For x(t) and y(t) with the same fundamental period,
(
ℱ#
𝑥 𝑡 𝑦 𝑡 ℎ$ = 𝑎$ ∗ 𝑏$ = < 𝑎% 𝑏$'%
%&'(

ℱ#
• Derivation: Let z(t)=x(t)y(t), and z 𝑡 ℎ$

% %
• ℎ$ = ∫& ∑' 𝑎' 𝑒 (')! * ∑+ 𝑏+ 𝑒 (+)! * 𝑒 ,($)! * 𝑑𝑡 = ∫& ∑' ∑+ 𝑎' 𝑏+ 𝑒 (('.+,$))! * 𝑑𝑡 = ∑' 𝑎' 𝑏$,'
& &
• 3.5.5 Multiplication (continued)

ℱ#
= 𝑥 𝜏 𝑦 𝑡 − 𝜏 𝑑𝜏 𝑇𝑎$ 𝑏$
)

Periodic convolution

• 3.5.6 Conjugation and Conjugate Symmetry

• For real x(t), (conjugate symmetric; |𝑎#! | = |𝑎! | and ∡𝑎#! = −∡𝑎! )

• For real and even x(t), 𝑎#! = 𝑎! (Fourier series coefficients are real and even)

• For real and odd x(t), 𝑎#! = −𝑎! (Fourier series coefficients are imaginary and odd)
• 3.5.7 Parseval’s relation for C-T periodic signals

1 1 1
• Derivation: P$ = ∫0 |𝑥 t |/ dt = ∫0 𝑥 ∗(t)𝑥(t)dt = ∫0 ∑3 𝑎3 𝑒 #%3&!' ∑4 𝑎4 𝑒 %4&!' 𝑑𝑡
2 0 0

1
= ∑3 ∫0 |𝑎3 |/ 𝑑𝑡 = ∑3 |𝑎3 |/
0

• The average power of the k-th harmonic component:

• Hence, the total average power in a periodic signal equals the sum of the average powers in
all of its harmonic components.
• 3.5.8 Summary of properties of the C-T Fourier series

• Differentiation
ℱ6 7+ ℱ6
𝑥 𝑡 𝑎! 7'
𝑏! = 𝑗𝑘𝜔8 𝑎!

• Integration

9$
ℱ6 ℱ6 , k≠0
%!&!
𝑥 𝑡 𝑎! (𝑎8 = 0) g(t) = ∫ x(𝜏)d𝜏 𝑏! = ?1
∫ g t dt, k = 0
0 0
• Example 3.6: Determine the Fourier series coefficients of g(t)

g(t) = x(t-1) - ½ with T = 4 and T1 = 1

ℱ#
𝑥 𝑡 𝑎$
. 1, k = 0
𝑏$ = 𝑎$ 𝑒 '*+,!-! − 𝛿0 where 𝛿+ = D
ℱ# / 0, k ≠ 0
g 𝑡 𝑏$

1
,k = 0 0, k=0
/
From example 3.5, a3 = ?:;< 3=// b3 = D :;< 3=//
,k ≠ 0 𝑒 #%3=// , k ≠ 0
3= 3=
• Example 3.6: Determine the Fourier series coefficients of x(t)

𝑑𝑥(t)
g(t) =
𝑑𝑡
ℱ#
𝑥 𝑡 e$ /2" /345 +1//
d$ = 𝑗k𝜔0 e$ =
*+1
e+ , k ≠ 0 e+ = = 𝑒 '*+1// , k ≠ 0
ℱ# / *+1 *(+1)#
g 𝑡 d$ e0 =
.
∫ x t dt = /
.
) )