Fourier Series Representation of Periodic

signals (ch.3)
 The response of LTI systems to complex exponentials
 Fourier series representation of continuous periodic signals
 Convergence of the Fourier series
 Properties of continuous-time Fourier series
 Fourier series representation of discrete –time periodic signals
 Properties of discrete FS
 Fourier series and LTI systems
The response of LTI systems to complex exponentials
Recall Chapter 2
 Objective: characterization of a LTI system

𝑥(𝑡) LTI 𝑦(𝑡)

 𝑥(𝑡) is considered as linear combinations of a basis signal 𝛿(𝑡)

∞ ∞
𝑥 𝑡 = න 𝑥 𝜏 𝛿(𝑡 − 𝜏) 𝑑𝜏 → 𝑦 𝑡 = න 𝑥 𝜏 ℎ(𝑡 − 𝜏) 𝑑𝜏
−∞ −∞
 𝛿(𝑡) is not the only one. In general, a basic signal should satisfy
• It can be used to construct a broad and useful class of signals
• The response of an LTI system to the basic signal is simple
1
The response of LTI systems to complex exponentials
Continuous-time

𝑒 𝑠𝑡 LTI 𝑦 𝑡 =?

∞ ∞
𝑦 𝑡 = න ℎ 𝜏 𝑒 𝑠(𝑡−𝜏) 𝑑𝜏 = 𝑒 𝑠𝑡 න ℎ 𝜏 𝑒 −𝑠𝜏 𝑑𝜏
−∞ −∞
∞
Let න ℎ 𝜏 𝑒 −𝑠𝜏 𝑑𝜏 = 𝐻(𝑠) → 𝑦 𝑡 = 𝐻(𝑠)𝑒 𝑠𝑡
−∞

• 𝑒 𝑠𝑡 is an eigenfunction of the system

• For a specific value s, 𝐻(𝑠) is the corresponding eigenvalue 2
The response of LTI systems to complex exponentials
Continuous-time
∞
𝑒 𝑠𝑡 LTI න ℎ 𝜏 𝑒 −𝑠𝜏 𝑑𝜏 𝑒 𝑠𝑡 = 𝐻(𝑠)𝑒 𝑠𝑡
−∞

If 𝑥 𝑡 = 𝑎1 𝑒 𝑠1 𝑡 + 𝑎2 𝑒 𝑠2 𝑡 +𝑎3 𝑒 𝑠3 𝑡 𝑦 𝑡 =?

𝑦 𝑡 = 𝑎1 𝐻 𝑠1 𝑒 𝑠1 𝑡 + 𝑎2 𝐻(𝑠2 )𝑒 𝑠2 𝑡 +𝑎1 𝐻(𝑠3 )𝑒 𝑠3 𝑡

Generally, if 𝑥 𝑡 = ෍ 𝑎𝑘 𝑒 𝑠𝑘𝑡
𝑘

𝑦 𝑡 = ෍ 𝑎𝑘 𝐻 𝑠𝑘 𝑒 𝑠𝑘𝑡
3
𝑘
The response of LTI systems to complex exponentials
Discrete-time

𝑧𝑛 LTI 𝑦[𝑛] =?

∞ ∞

𝑦 𝑛 = ෍ ℎ[𝑘]𝑧 𝑛−𝑘 = 𝑧 𝑛 ෍ ℎ[𝑘]𝑧 −𝑘

𝑘=−∞ 𝑘=−∞
∞
Let 𝐻 𝑧 = ෍ ℎ[𝑘]𝑧 −𝑘 → 𝑦 𝑛 = 𝐻[𝑧]𝑧 𝑛
𝑘=−∞

• 𝑧 𝑛 is an eigenfunction of the system

• For a specific value z, 𝐻[𝑧] is the corresponding eigenvalue 4
The response of LTI systems to complex exponentials
Discrete-time
∞

𝑧𝑛 LTI ෍ ℎ[𝑘]𝑧 −𝑘 𝑧 𝑛 = 𝐻[𝑧]𝑧 𝑛

𝑘=−∞

If 𝑥[𝑛] = ෍ 𝑎𝑘 𝑧𝑘𝑛
𝑘

𝑦 𝑛 = ෍ 𝑎𝑘 𝐻 𝑧𝑘 𝑧𝑘𝑛
𝑘

5
The response of LTI systems to complex exponentials
Examples
For a LTI system 𝑦 𝑡 = 𝑥(𝑡 − 3), determine 𝐻 𝑠
Solution 1:
let 𝑥 𝑡 = 𝑒 𝑠𝑡 , 𝑦 𝑡 = 𝑒 𝑠(𝑡−3) = 𝑒 −3𝑠 𝑒 𝑠𝑡
∴ 𝐻 𝑠 = 𝑒 −3𝑠

Solution 2:
∞ ∞
𝐻 𝑠 = න ℎ 𝜏 𝑒 −𝑠𝜏 𝑑𝜏 = න 𝛿 𝜏 − 3 𝑒 −𝑠𝜏 𝑑𝜏 = 𝑒 −3𝑠
−∞ −∞

6
The response of LTI systems to complex exponentials
Examples
For a LTI system 𝑦 𝑡 = 𝑥(𝑡 − 3)
If 𝑥 𝑡 = cos(4𝑡) + cos(7𝑡), 𝑦 𝑡 =?
Solution 1: 𝑦 𝑡 = cos 4 𝑡 − 3 + cos 7 𝑡 − 3
1 𝑗4𝑡 1 −𝑗4𝑡 1 𝑗7𝑡 1 −𝑗7𝑡
Solution 2: 𝑥 𝑡 = 𝑒 + 𝑒 + 𝑒 + 𝑒
2 2 2 2
1 𝑗4𝑡
1 −𝑗4𝑡
1 1
𝑦 𝑡 = 𝐻 𝑗4 𝑒 + 𝐻 −𝑗4 𝑒 + 𝐻 𝑗7 𝑒 + 𝐻 −𝑗7 𝑒 −𝑗7𝑡
𝑗7𝑡
2 2 2 2
−3𝑠 1 −𝑗12 𝑗4𝑡 1 𝑗12 −𝑗4𝑡 1 −𝑗21 𝑗7𝑡 1 𝑗21 −𝑗7𝑡
𝐻 𝑠 =𝑒 = 𝑒 𝑒 + 𝑒 𝑒 + 𝑒 𝑒 + 𝑒 𝑒
2 2 2 2
1 𝑗4(𝑡−3) 1 −𝑗4(𝑡−3) 1 𝑗7(𝑡−3) 1 −𝑗7(𝑡−3)
= 𝑒 + 𝑒 + 𝑒 + 𝑒
2 2 2 2 7
Fourier Series Representation of Periodic
signals (ch.3)
 The response of LTI systems to complex exponentials
 Fourier series representation of continuous periodic signals
 Convergence of the Fourier series
 Properties of continuous-time Fourier series
 Fourier series representation of discrete –time periodic signals
 Properties of discrete
 FS Fourier series and LTI systems
 Fourier series representation of C-T periodic signals
Recall
𝑥(𝑡) LTI 𝑦(𝑡)
𝑒 𝑠𝑡 𝐻(𝑠)𝑒 𝑠𝑡

 Decompose 𝑥(𝑡) into linear combinations of basis signals, which should satisfy
• It can be used to construct a broad and useful class of signals
• The response of an LTI system to the basic signal is simple

 Complex exponentials are eigenfunctions of a LTI system

 Can we represent 𝑥(𝑡) as linear combinations of complex exponentials?

9
 Fourier series representation of C-T periodic signals
Linear combination of harmonically related complex exponentials
 Harmonically related complex exponentials (consider 𝑒 𝑠𝑡 with 𝑠 purely imaginary)
∅𝑘 𝑡 = 𝑒 𝑗𝑘𝜔0 𝑡 = 𝑒 𝑗𝑘(2𝜋/𝑇)𝑡 , 𝑘 = 0, ±1, ±2, …
2𝜋 𝑇
For any 𝑘 ≠ 0, fundamental frequency |𝑘|𝜔0 ; fundamental period =
|𝑘|𝜔0 |𝑘|

 Linear combination of ∅𝑘 𝑡 is also periodic

∞ ∞
𝑥 𝑡 =෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡 = ෍ 𝑎𝑘 𝑒 𝑗𝑘(2𝜋/𝑇)𝑡
𝑘=−∞ 𝑘=−∞
 Representation of a periodic signal by Linear combination of ∅𝑘 𝑡 is referred
to as Fourier Series representation, 𝜔0 is the fundamental frequency
 For 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡 , 𝑘 = 0: DC component; 𝑘 = ±1: fundamental (first harmonic)
components; 𝑘 = ±𝑁: 𝑁th harmonic components 10
 Fourier series representation of C-T periodic signals
Linear combination of harmonically related complex exponentials
 An example
3
If 𝑥 𝑡 = ෍ 𝑎𝑘 𝑒 𝑗𝑘2𝜋𝑡
𝑘=−3
And 𝑎0 = 1, 𝑎1 = 𝑎−1 = 1/4, 𝑎2 = 𝑎−2 = 1/2, 𝑎3 = 𝑎−3 = 1/3
1 𝑗2𝜋𝑡 −𝑗2𝜋𝑡
1 𝑗4𝜋𝑡 −𝑗4𝜋𝑡
1 𝑗6𝜋𝑡
𝑥 𝑡 =1+ 𝑒 +𝑒 + 𝑒 +𝑒 + 𝑒 + 𝑒 −𝑗6𝜋𝑡
4 2 3
1 2
= 1 + cos 2𝜋𝑡 + cos 4𝜋𝑡 + cos 6𝜋𝑡
2 3

11
 Fourier series representation of C-T periodic signals
Linear combination of harmonically related complex exponentials
 Real signal
∞ ∞ ∞
𝑥 𝑡 =෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡 𝑥∗ 𝑡 = ෍ 𝑎𝑘∗ 𝑒 −𝑗𝑘𝜔0 𝑡 = ෍ ∗
𝑎−𝑘 𝑒 𝑗𝑘𝜔0 𝑡
𝑘=−∞ 𝑘=−∞ 𝑘=−∞
∗
Real ⇒ 𝑥 𝑡 = 𝑥 ∗ 𝑡 ⇒ 𝑎𝑘 = 𝑎−𝑘 , or 𝑎𝑘∗ = 𝑎−𝑘 （Conjugate symmetry）
 Alternative form of Fourier Series for real signal
∞
𝑥 𝑡 = 𝑎0 + ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡 + 𝑎−𝑘 𝑒 −𝑗𝑘𝜔0 𝑡
𝑘=1
∞ ∞
= 𝑎0 + ෍ 2ℛ𝑒 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡 = 𝑎0 + 2 ෍ 𝐴𝑘 cos 𝑘𝜔0 𝑡 + 𝜃𝑘
𝑘=1 𝑘=1
𝑎𝑘 = 𝐴𝑘 𝑒 𝑗𝜃𝑘 12
 Fourier series representation of C-T periodic signals
Determine the Fourier Series Representation
𝑇 𝑇 ∞
න 𝑥 𝑡 𝑒 −𝑗𝑛𝜔0 𝑡 𝑑𝑡 = න ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡 𝑒 𝑗−𝑛𝜔
−𝑗𝑛𝜔0 𝑡 𝑑𝑡
0 0 𝑘=−∞ 𝑇, 𝑘 = 𝑛
=ቊ = 𝑇𝛿[𝑘 − 𝑛]
0, 𝑘 ≠ 𝑛
∞ 𝑇
=෍ 𝑎𝑘 න 𝑒 𝑗(𝑘−𝑛)𝜔0 𝑡 𝑑𝑡 = 𝑇𝑎𝑛
𝑘=−∞ 0

1 𝑇
∴ 𝑎𝑛 = න 𝑥 𝑡 𝑒 −𝑗𝑛𝜔0 𝑡 𝑑𝑡
𝑇 0

1
𝑎𝑘 = න 𝑥 𝑡 𝑒 −𝑗𝑘𝜔0 𝑡 𝑑𝑡
𝑇 𝑇
13
 Fourier series representation of C-T periodic signals
Fourier Series pair
∞
𝑥 𝑡 =෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡 Synthesis equation
𝑘=−∞

1
𝑎𝑘 = න 𝑥 𝑡 𝑒 −𝑗𝑘𝜔0 𝑡 𝑑𝑡 Analysis equation
𝑇 𝑇

 𝑎𝑘 : Fourier Series coefficients or spectral coefficients of 𝑥 𝑡

1
𝑎0 = න 𝑥 𝑡 𝑑𝑡
𝑇 𝑇
14
 Fourier series representation of C-T periodic signals
Determine the Fourier Series Representation
 Examples: determine the FS coefficients of 𝑥 𝑡

𝑥 𝑡 = sin 𝜔0 𝑡

1 𝑗𝜔 𝑡 1 −𝑗𝜔 𝑡
sin 𝜔0 𝑡 = 𝑒 0 − 𝑒 0
2𝑗 2𝑗
1 1
∴ 𝑎1 = 𝑎−1 =− 𝑎𝑘 = 0, for 𝑘 ≠ ±1
2𝑗 2𝑗

15
 Fourier series representation of C-T periodic signals
Determine the Fourier Series Representation
 Examples: determine the FS coefficients of 𝑥 𝑡
𝜋
𝑥 𝑡 = 1 + sin 𝜔0 𝑡 + 2 cos 𝜔0 𝑡 + cos 2𝜔0 𝑡 +
4
1 𝑗𝜔 𝑡
𝑥 𝑡 =1+ 𝑒 0 − 𝑒 −𝑗𝜔0 𝑡 + 𝑒 𝑗𝜔0 𝑡 + 𝑒 −𝑗𝜔0 𝑡
2𝑗
1 𝑗 2𝜔 𝑡+𝜋/4
+ 𝑒 0 + 𝑒 −𝑗 2𝜔0 𝑡+𝜋/4
2
1 𝑗𝜔 𝑡 1 −𝑗𝜔 𝑡 1 𝑗𝜋/4 𝑗2𝜔 𝑡 1 −𝑗𝜋/4 −𝑗2𝜔 𝑡
∴𝑥 𝑡 =1+ 1+ 𝑒 0 + 1− 𝑒 0 + 𝑒 𝑒 0 + 𝑒 𝑒 0
2𝑗 2𝑗 2 2
𝑎0 𝑎1 𝑎−1 𝑎2 𝑎−2
16
 Fourier series representation of C-T periodic signals
Determine the Fourier Series Representation
 Examples: determine the FS coefficients of 𝑥 𝑡
1 𝑗𝜔 𝑡 1 −𝑗𝜔 𝑡 1 𝑗𝜋/4 𝑗2𝜔 𝑡 1 −𝑗𝜋/4 −𝑗2𝜔 𝑡
𝑥 𝑡 =1+ 1+ 𝑒 0 + 1− 𝑒 0 + 𝑒 𝑒 0 + 𝑒 𝑒 0
2𝑗 2𝑗 2 2
𝑎0 𝑎1 𝑎−1 𝑎2 𝑎−2

17
 Fourier series representation of C-T periodic signals
Determine the Fourier Series Representation
 Examples: determine the FS coefficients of 𝑥 𝑡

1 𝑇/2 1 𝑇1 2𝑇1
𝑎0 = න 𝑥 𝑡 𝑑𝑡 = න 1 𝑑𝑡 =
𝑇 −𝑇/2 𝑇 −𝑇1 𝑇
1 𝑇1 −𝑗𝑘𝜔 𝑡 1 𝑇 2 𝑒 𝑗𝑘𝜔0 𝑇1 − 𝑒 −𝑗𝑘𝜔0 𝑇1
𝑎𝑘 = න 𝑒 0 𝑑𝑡 = − 𝑒 −𝑗𝑘𝜔0 𝑡 | 1 =
𝑇 −𝑇1 𝑗𝑘𝜔0 𝑇 −𝑇1 𝑘𝜔0 𝑇 2𝑗
sin 𝑥
sinc 𝑥 =
2 sin 𝑘𝜔0 𝑇1 sin 𝑘𝜔0 𝑇1 2𝑇1 sin 𝑘𝜔0 𝑇1 𝑥
= = = ,𝑘 ≠ 0
𝑘𝜔0 𝑇 𝑘𝜋 𝑇 𝑘𝜔0 𝑇1 18
 Fourier series representation of C-T periodic signals
Determine the Fourier Series Representation
𝑇 = 4𝑇1
 Examples: determine the FS
coefficients of 𝑥 𝑡

2 sin 𝑘𝜔0 𝑇1 sin 𝑘𝜔0 𝑇1

𝑎𝑘 = = 𝑇 = 8𝑇1
𝑘𝜔0 𝑇 𝑘𝜋
2𝑇1 sin 𝑘𝜔0 𝑇1
= ,𝑘 ≠ 0
𝑇 𝑘𝜔0 𝑇1
𝑇 = 16𝑇1

19
Fourier Series Representation of Periodic
signals (ch.3)
 The response of LTI systems to complex exponentials
 Fourier series representation of continuous periodic signals
 Convergence of the Fourier series
 Properties of continuous-time Fourier series
 Fourier series representation of discrete –time periodic signals
 Properties of discrete FS
 Fourier series and LTI systems
 Convergence of the Fourier series
History
 Using “trigonometric sum” to describe periodic signal can be
tracked back to Babylonians who predicted astronomical events
similarly.
 L. Euler (in 1748) and Bernoulli (in 1753) used the “normal mode”
concept to describe the motion of a vibrating string; though JL
Lagrange strongly criticized this concept.
 Fourier (in 1807) had found series of harmonically related
sinusoids to be useful to describe the temperature distribution
through body, and he claimed “any” periodic signal can be
represented by such series.
 Dirichlet (in 1829) provide a precise condition under which a
periodic signal can be represented by a Fourier series.

21
 Convergence of the Fourier series
Convergence problem
 Approximate periodic signal 𝑥 𝑡 by 𝑥𝑁 𝑡 = σ𝑁 𝑎
𝑘=−𝑁 𝑘 𝑒 𝑗𝑘𝜔0 𝑡

 How good the approximation is?

𝑁
𝑒𝑁 (𝑡) = 𝑥 𝑡 − 𝑥𝑁 𝑡 = 𝑥 𝑡 − ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡 𝐸𝑁 = න 𝑒𝑁 (𝑡) 2 𝑑𝑡
𝑘=−𝑁 𝑇
1
• When 𝑎𝑘 = ‫𝑥 ׬‬ 𝑡 𝑒 −𝑗𝑘𝜔0 𝑡 𝑑𝑡, 𝐸𝑁 is minimized; 𝑁 → ∞ ⇒ 𝐸𝑁 → 0
𝑇 𝑇
 Problem:
• 𝑎𝑘 may be infinite
Convergence problem!
• 𝑁 → ∞, 𝑥𝑁 𝑡 may be infinite 22
 Convergence of the Fourier series
Two different classes of conditions
 Condition 1: Finite energy condition
If ‫|)𝑡(𝑥| 𝑇׬‬2 𝑑𝑡 < ∞, 𝑥(𝑡) can be represented by a FS
• Guarantees no energy in their difference; FS is not equal to 𝑥(𝑡)
 Condition 2: Dirichlet condition
(1) Absolutely integrable ‫∞ < 𝑡𝑑|)𝑡(𝑥| 𝑇׬‬
An example: a periodic signal
1
𝑥 𝑡 = ,0 < 𝑡 ≤ 1
𝑡
is not absolutely integrable. 23
 Convergence of the Fourier series
Two different classes of conditions
 Condition 2: Dirichlet condition

(2) In any finite interval of time, x(t) is of bounded variation; finite maxima
and minima in one period
An example: a periodic signal
2𝜋
𝑥 𝑡 = sin ,0 < 𝑡 ≤ 1
𝑡
meets (1) but not (2).

24
 Convergence of the Fourier series
Two different classes of conditions
 Condition 2: Dirichlet condition

(3) In any finite interval of time, only a finite number of finite discontinuities

An example: a periodic signal meets (1) and (2) but not (3).

• Dirichlet condition guarantees 𝑥 𝑡 equals

its Fourier Series representation, except
for discontinuous points.
• Three examples are pathological in nature
and do not typically arise in practical
contexts. 25
 Convergence of the Fourier series
Example
 𝑥 𝑡 is a square wave

𝑁
𝑥𝑁 𝑡 = ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡
𝑘=−𝑁

lim 𝑥𝑁 𝑡1 = 𝑥 𝑡1
𝑁→∞

26
Fourier Series Representation of Periodic
signals (ch.3)
 The response of LTI systems to complex exponentials
 Fourier series representation of continuous periodic signals
 Convergence of the Fourier series
 Properties of continuous-time Fourier series
 Fourier series representation of discrete –time periodic signals
 Properties of discrete FS
Fourier series and LTI systems
 Properties of continuous-time FS

 Use the notation

ℱ𝒮
𝑥 𝑡 𝑎𝑘

to signify the paring of a periodic signal with its FS coefficients.

 Linearity: if 𝑥 𝑡 and 𝑦 𝑡 are periodic signals with the same period T

ℱ𝒮
𝑥 𝑡 𝑎𝑘
ℱ𝒮 𝑐 = 𝐴𝑎 + 𝐵𝑏
ℱ𝒮 ⇒ 𝑧 𝑡 = 𝐴𝑥 𝑡 + 𝐵𝑦 𝑡 𝑘 𝑘 𝑘
𝑦 𝑡 𝑏𝑘

28
 Properties of continuous-time FS

 Time shifting
ℱ𝒮 ℱ𝒮
𝑥 𝑡 𝑎𝑘 ⇒ 𝑥 𝑡 − 𝑡0 𝑒 −𝑗𝑘𝜔0 𝑡0 𝑎𝑘

 Proof
𝑡 − 𝑡0 = 𝜏
1 −𝑗𝑘𝜔 𝑡
1
න 𝑥 𝑡 − 𝑡0 𝑒 0 𝑑𝑡 = න 𝑥 𝜏 𝑒 −𝑗𝑘𝜔0 (𝜏+𝑡0) 𝑑𝜏
𝑇 𝑇 𝑇 𝑇
−𝑗𝑘𝜔 𝑡
1
=𝑒 0 0 න 𝑥 𝜏 𝑒 −𝑗𝑘𝜔0 𝜏 𝑑𝜏
𝑇 𝑇

= 𝑒 −𝑗𝑘𝜔0 𝑡0 𝑎𝑘
29
 Properties of continuous-time FS
 Time reversal
ℱ𝒮 ℱ𝒮
𝑥 𝑡 𝑎𝑘 ⇒ 𝑦 𝑡 = 𝑥 −𝑡 𝑏𝑘 = 𝑎−𝑘

 Proof
∞ ∞ ∞
𝑥 𝑡 = ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡 ⇒ 𝑥 −𝑡 = ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 (−𝑡) = ෍ 𝑎𝑘 𝑒 𝑗(−𝑘)𝜔0 𝑡
𝑘=−∞ 𝑘=−∞ 𝑘=−∞
∞

= ෍ 𝑎−𝑚 𝑒 𝑗𝑚𝜔0 𝑡
𝑚=−∞

 If 𝑥 𝑡 even, 𝑎−𝑘 = 𝑎𝑘 , if 𝑥 𝑡 odd, 𝑎−𝑘 = −𝑎𝑘

30
 Properties of continuous-time FS
 Time scaling
ℱ𝒮 ℱ𝒮
𝑥 𝑡 𝑎𝑘 ⇒ 𝑦 𝑡 = 𝑥 𝛼𝑡 𝑏𝑘 = 𝑎𝑘
 Proof
∞ ∞ ∞
𝑥 𝑡 = ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡 ⇒ 𝑥 𝛼𝑡 = ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝛼𝑡 = ෍ 𝑎𝑘 𝑒 𝑗𝑘(𝛼𝜔0 )𝑡
𝑘=−∞ 𝑘=−∞ 𝑘=−∞

FS coefficients the same, but fundamental frequency changed.

31
 Properties of continuous-time FS

 Multiplication
ℱ𝒮 ∞
𝑥 𝑡 𝑎𝑘 ℱ𝒮
𝑦 𝑡
ℱ𝒮
𝑏𝑘
⇒ 𝑧 𝑡 =𝑥 𝑡 𝑦 𝑡 ℎ𝑘 = ෍ 𝑎𝑙 𝑏𝑘−𝑙
𝑙=−∞
 Proof ∞ ∞ ∞ ∞ 𝑙+𝑚 =𝑘

𝑥 𝑡 𝑦 𝑡 = ෍ 𝑎𝑙 𝑒 𝑗𝑙𝜔0 𝑡 ෍ 𝑏𝑚 𝑒 𝑗𝑚𝜔0 𝑡 = ෍ ෍ 𝑎𝑙 𝑏𝑚 𝑒 𝑗(𝑙+𝑚)𝜔0 𝑡

𝑙=−∞ 𝑚=−∞ 𝑙=−∞ 𝑚=−∞
∞ ∞ ∞ ∞

= ෍ ෍ 𝑎𝑙 𝑏𝑘−𝑙 𝑒 𝑗𝑘𝜔0 𝑡 = ෍ ෍ 𝑎𝑙 𝑏𝑘−𝑙 𝑒 𝑗𝑘𝜔0 𝑡

𝑙=−∞ 𝑘=−∞ 𝑘=−∞ 𝑙=−∞
∞
ℎ𝑘
= ෍ ℎ𝑘 𝑒 𝑗𝑘𝜔0 𝑡
32
𝑘=−∞
 Properties of continuous-time FS

 Conjugation and conjugate symmetry

ℱ𝒮 ℱ𝒮
𝑥 𝑡 𝑎𝑘 ⇒ 𝑧 𝑡 = 𝑥∗ 𝑡 ∗
𝑏𝑘 = 𝑎−𝑘
 Proof ∞ −𝑘 = 𝑚 ∞
∞

𝑥 𝑡 = ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡 ∴ 𝑥 ∗ 𝑡 = ෍ 𝑎𝑘∗ 𝑒 −𝑗𝑘𝜔0 𝑡 = ෍ 𝑎−𝑚

∗ 𝑒 𝑗𝑚𝜔0 𝑡

𝑘=−∞ 𝑘=−∞ 𝑚=−∞

 If 𝑥 𝑡 real, 𝑎𝑘∗ = 𝑎−𝑘 (conjugate symmetry) ⇒ |𝑎𝑘 | = |𝑎−𝑘 |

• 𝑥 𝑡 real and even (𝑎−𝑘 = 𝑎𝑘 ) ⇒ 𝑎𝑘 = 𝑎𝑘∗ ⇒ 𝑎𝑘 real and even

• 𝑥 𝑡 real and odd (𝑎−𝑘 = −𝑎𝑘 ) ⇒ 𝑎𝑘 = −𝑎𝑘∗ ⇒ 𝑎𝑘 pure imagery and odd
• 𝑎0 =?
33
 Properties of continuous-time FS

 Differentiation and Integration

ℱ𝒮
𝑑𝑥(𝑡)/𝑑𝑡 𝑗𝑘𝜔0 𝑎𝑘
ℱ𝒮
𝑥 𝑡 𝑎𝑘 ⇒ ൞ 𝑡 ℱ𝒮
න 𝑥 𝜏 𝑑𝜏 𝑎𝑘 /(𝑗𝑘𝜔0 )
 Proof −∞

∞ ∞
𝑑𝑥(𝑡) 𝑑(𝑒 𝑗𝑘𝜔0 𝑡 )
= ෍ 𝑎𝑘 = ෍ 𝑎𝑘 𝑗𝑘𝜔0 𝑒 𝑗𝑘𝜔0 𝑡
𝑑𝑡 𝑑𝑡
𝑘=−∞ 𝑘=−∞

𝑡 ∞ 𝑡 ∞

න 𝑥 𝜏 𝑑𝜏 = ෍ 𝑎𝑘 න 𝑒 𝑗𝑘𝜔0 𝜏 𝑑𝜏 = ෍ 𝑎𝑘 /(𝑗𝑘𝜔0 )𝑒 𝑗𝑘𝜔0 𝑡

−∞ 𝑘=−∞ −∞ 𝑘=−∞
34
 Properties of continuous-time FS

 Frequency shifting
ℱ𝒮 ℱ𝒮
𝑥 𝑡 𝑎𝑘 ⇒ 𝑒 𝑗𝑀𝜔0 𝑡 𝑥 𝑡 𝑎𝑘−𝑀
 Proof
∞ ∞

𝑒 𝑗𝑀𝜔0 𝑡 𝑥 𝑡 = 𝑒 𝑗𝑀𝜔0 𝑡 ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡 = ෍ 𝑎𝑘 𝑒 𝑗(𝑘+𝑀)𝜔0𝑡

𝑘=−∞ 𝑘=−∞
∞
𝑘+𝑀 =𝑙
= ෍ 𝑎𝑙−𝑀 𝑒 𝑗𝑙𝜔0 𝑡
𝑙=−∞

35
 Properties of continuous-time FS
 Periodic convolution
ℱ𝒮
𝑥 𝑡 𝑎𝑘 ℱ𝒮
𝑦 𝑡
ℱ𝒮
𝑏𝑘
⇒ න 𝑥 𝜏 𝑦 𝑡 − 𝜏 𝑑𝜏
𝑇
𝑇𝑎𝑘 𝑏𝑘

 Proof ∞ ∞

න 𝑥 𝜏 𝑦 𝑡 − 𝜏 𝑑𝜏 = න ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝜏 ෍ 𝑏𝑚 𝑒 𝑗𝑚𝜔0 𝑡−𝜏

𝑑𝜏
𝑇 𝑇 𝑘=−∞ 𝑚=−∞
∞ ∞

=න ෍ ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝜏 𝑏𝑚 𝑒 −𝑗𝑚𝜔0 𝜏 𝑒 𝑗𝑚𝜔0 𝑡 𝑑𝜏

𝑇 𝑘=−∞ 𝑚=−∞
𝑇𝛿[𝑘 − 𝑚]
∞ ∞ ∞

= ෍ 𝑎𝑘 ෍ 𝑒 𝑗𝑚𝜔0 𝑡 𝑏𝑚 න 𝑒 𝑗𝑘𝜔0 𝜏 𝑒 −𝑗𝑚𝜔0 𝜏 𝑑𝜏 = ෍ 𝑇𝑎𝑘 𝑏𝑘 𝑒 𝑗𝑘𝜔0 𝑡

𝑘=−∞ 𝑚=−∞ 𝑇 𝑘=−∞ 36
 Properties of continuous-time FS

 Parseval’s relation ∞
1
න |𝑥(𝑡)|2 𝑑𝑡 = ෍ |𝑎𝑘 |2
𝑇 𝑇
𝑘=−∞
 Proof
∞
1 1 1
2 ∗
න |𝑥(𝑡)| 𝑑𝑡 = න 𝑥 𝑡 𝑥 𝑡 𝑑𝑡 = න 𝑥 𝑡 ෍ 𝑎𝑘∗ 𝑒 −𝑗𝑘𝜔0 𝑡 𝑑𝑡
𝑇 𝑇 𝑇 𝑇 𝑇 𝑇
𝑘=−∞
∞
1
= ෍ 𝑎𝑘∗ න 𝑥 𝑡 𝑒 −𝑗𝑘𝜔0 𝑡 𝑑𝑡
𝑇 𝑇
𝑘=−∞
∞ ∞

= ෍ 𝑎𝑘∗ 𝑎𝑘 = ෍ |𝑎𝑘 |2
37
𝑘=−∞ 𝑘=−∞
 Properties of continuous-time FS

 Parseval’s relation ∞
1
න |𝑥(𝑡)|2 𝑑𝑡 = ෍ |𝑎𝑘 |2
𝑇 𝑇
𝑘=−∞

1 1
න |𝑎𝑘 𝑒 0 | 𝑑𝑡 = න |𝑎𝑘 |2 𝑑𝑡 = |𝑎𝑘 |2
𝑗𝑘𝜔 𝑡 2
𝑇 𝑇 𝑇 𝑇

 |𝑎𝑘 |2 is the average power in the k-th harmonic component of 𝑥(𝑡)

 Total average power in 𝑥(𝑡) equals the sum of the average powers in all of
its harmonic components
38
 Properties of continuous-time FS

 Summary

39
 𝑔(𝑡)
Properties of continuous-time FS
 Examples FS coefficients (𝑐𝑘 ) of 𝑔 𝑡 ? 𝑥(𝑡)

 Solution
• Let 𝑥 𝑡 = σ∞
𝑘=−∞ 𝛿(𝑡 − 𝑘𝑇)
𝑞(𝑡)

1 𝑇/2 −𝑗𝑘𝜔 𝑡
1
𝑎𝑘 = න 𝛿 𝑡 𝑒 0 𝑑𝑡 =
𝑇 −𝑇/2 𝑇
• Let 𝑞 𝑡 = 𝑥 𝑡 + 𝑇1 − 𝑥 𝑡 − 𝑇1
1 𝑗𝑘𝜔 𝑇 2𝑗 sin 𝑘𝜔0 𝑇1
𝑏𝑘 = 𝑒 𝑗𝑘𝜔0 𝑇1 𝑎𝑘 − 𝑒 −𝑗𝑘𝜔0 𝑇1 𝑎𝑘 = 𝑒 0 1 −𝑒 −𝑗𝑘𝜔0 𝑇1 =
𝑇 𝑇
• 𝑞 𝑡 = 𝑑𝑔 𝑡 /𝑑𝑡 ∴ 𝑏𝑘 = 𝑗𝑘𝜔0 𝑐𝑘
𝑏𝑘 2𝑗 sin 𝑘𝜔0 𝑇1 sin 𝑘𝜔0 𝑇1 2𝑇1
∴ 𝑐𝑘 = = = ,𝑘 ≠ 0 𝑐0 =
𝑗𝑘𝜔0 𝑗𝑘𝜔0 𝑇 𝑘𝜋 𝑇 40
 Properties of continuous-time FS

𝑥(𝑡) sin 𝑘𝜔0 𝑇1

 Examples 𝑇1 = 1, 𝑇 = 4 𝑎𝑘 = ,𝑘 ≠ 0
𝑘𝜋
sin 𝑘𝜋/2
= ,𝑘 ≠ 0
𝑘𝜋

𝑔 𝑡 = 𝑥 𝑡 − 1 − 1/2 FS coefficients of 𝑔 𝑡 ?

 Solution
ℱ𝒮 −𝑗𝑘𝜔 𝑡
𝑥 𝑡−1 ↔𝑒 0 0𝑎
𝑘 = 𝑒 −𝑗𝑘𝜋/2 𝑎𝑘 , 𝑘 ≠ 0

ℱ𝒮
0, 𝑘 ≠ 0 ℱ𝒮 𝑒 −𝑗𝑘𝜋/2 𝑎𝑘 , 𝑘 ≠ 0
−1/2 ↔ ቐ 1 ∴ 𝑥 𝑡 − 1 − 1/2 ↔ ቐ 1
− ,𝑘 = 0 𝑎0 − ,𝑘 =0
2
2 41
 Properties of continuous-time FS
 Examples
Given a signal 𝑥 𝑡 with the following facts, determine 𝑥 𝑡
1. 𝑥 𝑡 is real;
2. 𝑥 𝑡 is periodic with T=4 and FS coefficients 𝑎𝑘 = 0 for k > 1;
3. A signal with FS coefficients 𝑏𝑘 = 𝑒 −𝑗𝜋𝑘/2 𝑎−𝑘 is odd;
1 2 1
4. ‫ = 𝑡𝑑 | 𝑡 𝑥| ׬‬2.
4 4

 Solution
𝜋 𝜋
𝑗( 2 )𝑡 −𝑗 𝑡
• From 2, 𝑥 𝑡 = 𝑎0 + 𝑎1 𝑒 + 𝑎−1 𝑒 2
• 𝑏𝑘 = 𝑒 −𝑗𝜋𝑘/2 𝑎−𝑘 corresponds to the signal 𝑥 −𝑡 + 1 , which is real and odd
1 1 1
• ‫׬‬4 |𝑥 𝑡 |2 𝑑𝑡 = ‫׬‬4 |𝑥 −𝑡 + 1 |2 𝑑𝑡 = σ∞ |𝑏
𝑘=−∞ 𝑘 |2
= |𝑏0 | 2
+ |𝑏1 |2
+|𝑏−1 | 2
=
4 4 2
j j
• 𝑥 −𝑡 + 1 is real and odd ⇒ 𝑏𝑘 = −𝑏−𝑘 ⇒ 𝑏0 = 0, 𝑏1 = −𝑏−1 = or -
2 2
• 𝑎0 = 0, 𝑎1 = −1/2, 𝑎−1 = 1/2 42
Fourier Series Representation of Periodic
signals (ch.3)
 The response of LTI systems to complex exponentials
 Fourier series representation of continuous periodic signals
 Convergence of the Fourier series
 Properties of continuous-time Fourier series
 Fourier series representation of discrete –time periodic signals
 Properties of discrete FS
 Fourier series and LTI systems
 Fourier series representation of D-T periodic signals
Linear combination of harmonically related complex exponentials
 Harmonically related complex exponentials
∅𝑘 [𝑛] = 𝑒 𝑗𝑘(2𝜋/𝑁)𝑛 , 𝑘 = 0, ±1, ±2, …
2𝜋
• Fundamental frequency |𝑘|( )
𝑁
• Only N distinct signals in ∅𝑘 [𝑛], since ∅𝑘 [𝑛] = ∅𝑘+𝑟𝑁 [𝑛]

 Linear combination of ∅𝑘 [𝑛] is also periodic

𝑥[𝑛] = ෍ 𝑎𝑘 ∅𝑘 [𝑛] = ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑛 = ෍ 𝑎𝑘 𝑒 𝑗𝑘 2𝜋/𝑁 𝑛

𝑘= 𝑁 𝑘= 𝑁 𝑘= 𝑁

 σ𝑘= 𝑁 𝑎𝑘 𝑒 𝑗𝑘 2𝜋/𝑁 𝑛 : Discrete-Time Fourier Series; 𝑎𝑘 : Fourier Series

coefficients 44
 Fourier series representation of D-T periodic signals
Determine the Fourier Series Representation

෍ 𝑥[𝑛]𝑒 −𝑗𝑟 2𝜋/𝑁 𝑛 = ෍ ෍ 𝑎𝑘 𝑒 𝑗𝑘 2𝜋/𝑁 𝑛 𝑒 −𝑗𝑟 2𝜋/𝑁 𝑛

𝑛= 𝑁 𝑛= 𝑁 𝑘= 𝑁 𝑁, 𝑘 = 𝑟
=ቊ = 𝑁𝛿[𝑘 − 𝑟]
0, 𝑘 ≠ 𝑛

= ෍ 𝑎𝑘 ෍ 𝑒 𝑗(𝑘−𝑟) 2𝜋/𝑁 𝑛
= 𝑁𝑎𝑟
𝑘= 𝑁 𝑛= 𝑁

1
∴ 𝑎𝑘 = ෍ 𝑥[𝑛]𝑒 −𝑗𝑘 2𝜋/𝑁 𝑛
𝑁
𝑛= 𝑁

45
 Fourier series representation of D-T periodic signals
Determine the Fourier Series Representation
 Discrete Fourier series pair

1 Analysis equation; 𝑎𝑘 : Fourier Series

𝑎𝑘 = ෍ 𝑥[𝑛]𝑒 −𝑗𝑘 2𝜋/𝑁 𝑛
𝑁 coefficients
𝑛= 𝑁

𝑥[𝑛] = ෍ 𝑎𝑘 𝑒 𝑗𝑘 2𝜋/𝑁 𝑛
Synthesis equation; Fourier Series (Finite)
𝑘= 𝑁

 𝑎𝑘 is periodic 𝑥 𝑛 = ෍ 𝑎𝑘 ∅𝑘 𝑛 = 𝑎0 ∅0 𝑛 + 𝑎1 ∅1 𝑛 + ⋯ + 𝑎𝑁−1 ∅𝑁−1 [𝑛]

𝑘= 𝑁
= 𝑎1 ∅1 𝑛 + 𝑎2 ∅2 𝑛 + ⋯ + 𝑎𝑁 ∅𝑁 [𝑛]
= 𝑎2 ∅2 𝑛 + 𝑎3 ∅3 𝑛 + ⋯ + 𝑎𝑁+1 ∅𝑁+1 [𝑛]
∴ 𝑎𝑘 = 𝑎𝑘+𝑟𝑁 46
 Fourier series representation of D-T periodic signals
Determine the Fourier Series Representation
 Examples 𝑥[𝑛] = sin 𝜔0 𝑛
2𝜋
If 𝜔0 = 𝑥[𝑛] is periodic with fundamental period of N.
,
𝑁
1 𝑗(2𝜋/𝑁)𝑛 1 𝑗(2𝜋/𝑁)𝑛
𝑥[𝑛] = sin 𝜔0 𝑛 = 𝑒 − 𝑒
2𝑗 2𝑗
1 1
∴ 𝑎1 = 𝑎−1 = − 𝑎𝑘 = 0, for 𝑘 ≠ ±1 in one period
2𝑗 2𝑗
𝑁=5
 𝑎𝑘 is periodic and only one
period is utilized in the synthesis
equation
47
 Fourier series representation of C-T periodic signals
Determine the Fourier Series Representation
2𝜋 2𝜋 4𝜋 𝜋
 Examples: 𝑥[𝑛] = 1 + sin( )𝑛 + 3 cos( )𝑛 + cos 𝑛+
𝑁 𝑁 𝑁 2
1 𝑗 2𝜋/𝑁 𝑛
3 𝑗 2𝜋/𝑁 𝑛
𝑥[𝑛] = 1 + 𝑒 − 𝑒 −𝑗 2𝜋/𝑁 𝑛 + 𝑒 + 𝑒 −𝑗 2𝜋/𝑁 𝑛
2𝑗 2
1 𝑗 4𝜋𝑛/𝑁+𝜋/2
+ 𝑒 + 𝑒 −𝑗 4𝜋𝑛/𝑁+𝜋/2
2
𝑎1 𝑎−1
𝑎0 3 1 𝑗 2𝜋/𝑁 𝑛 3 1 −𝑗 2𝜋/𝑁 𝑛
∴ 𝑥 [𝑛] = 1 + + 𝑒 + − 𝑒
2 2𝑗 2 2𝑗
𝑎2 𝑎−2
1 𝑗𝜋/2 𝑗2 2𝜋/𝑁 𝑛 1 −𝑗𝜋/2 −𝑗2 2𝜋/𝑁 𝑛
+ 𝑒 𝑒 + 𝑒 𝑒
2 2 48
 Fourier series representation of D-T periodic signals
Linear combination of harmonically related complex exponentials
 Real signal ∗
𝑎𝑘 = 𝑎−𝑘 , or 𝑎𝑘∗ = 𝑎−𝑘

𝑥[𝑛] = ෍ 𝑎𝑘 𝑒 𝑗𝑘 2𝜋/𝑁 𝑛

𝑘= 𝑁

𝑥 ∗ [𝑛] = ෍ 𝑎𝑘∗ 𝑒 −𝑗𝑘 2𝜋/𝑁 𝑛 ∗

= ෍ 𝑎−𝑘 𝑒 𝑗𝑘 2𝜋/𝑁 𝑛

𝑘= 𝑁 𝑘= 𝑁

∗
𝑥 𝑛 = 𝑥∗ 𝑛 ⟹ 𝑎𝑘 = 𝑎−𝑘
49
 Fourier series representation of D-T periodic signals
𝑥[𝑛]
Determine the Fourier Series Representation
 Examples: 𝑥[𝑛] discrete square
1 𝑁1 1 𝑁1
𝑎𝑘 = ෍ 𝑥[𝑛]𝑒 −𝑗𝑘 2𝜋/𝑁 𝑛
= ෍ 𝑒 −𝑗𝑘 2𝜋/𝑁 𝑛
𝑁 𝑛=−𝑁1 𝑁 𝑛=−𝑁1
𝑚 = 𝑛 + 𝑁1 2𝑁1 2𝑁1
1 1 𝑗𝑘
= ෍ 𝑒 −𝑗𝑘 2𝜋/𝑁 (𝑚−𝑁1 )
= 𝑒 2𝜋/𝑁 𝑁1
෍ 𝑒 −𝑗𝑘 2𝜋/𝑁 𝑚
𝑁 𝑚=0 𝑁 𝑚=0
2𝑁1 + 1
, 𝑘 = 0, ±𝑁, ±2𝑁, …
𝑁
= 1 sin 2𝑘𝜋(𝑁 + 1/2)/𝑁
1
, 𝑘 ≠ 0, ±𝑁, ±2𝑁, …
𝑁 sin 𝑘𝜋/𝑁 𝑎𝑘 (2𝑁1 + 1 = 5, 𝑁 = 20)

50
 2𝑁1 + 1 = 5, 𝑁 = 9
Fourier series representation of D-T periodic signals
Linear combination of harmonically related complex exponentials
 Approximate a discrete square by 𝑥[𝑛]
ො
𝑀
𝑥[𝑛]
ො =෍ 𝑎𝑘 𝑒 𝑗𝑘 2𝜋/𝑁 𝑛
𝑘=−𝑀

2𝑁1 +1
, 𝑘 = 0, ±𝑁, ±2𝑁, …
𝑁
With 𝑎𝑘 = ൞ 1 sin 2𝑘𝜋(𝑁1 +1/2)/𝑁
, else
𝑁 sin 𝑘𝜋/𝑁

 For M=4, 𝑥[𝑛]

ො =𝑥 𝑛

 No convergence issues for the

discrete−time Fourier series!
51
Fourier Series Representation of Periodic
signals (ch.3)
 The response of LTI systems to complex exponentials
 Fourier series representation of continuous periodic signals
 Convergence of the Fourier series
 Properties of continuous-time Fourier series
 Fourier series representation of discrete –time periodic signals
 Properties of discrete FS
Fourier series and LTI systems
 Properties of discrete-time FS
ℱ𝒮 ℱ𝒮
𝑥[𝑛] 𝑎𝑘 𝑦[𝑛] 𝑏𝑘
 Multiplication
ℱ𝒮 ෍ 𝑎𝑙 𝑏𝑘−𝑙
𝑥[𝑛]𝑦[𝑛]
𝑙= 𝑁

 First difference
ℱ𝒮
𝑥 𝑛 − 𝑥[𝑛 − 1] 1 − 𝑒 −𝑗𝑘 2𝜋/𝑁 𝑎𝑘

 Parseval’s relation

1
෍ |𝑥[𝑛]|2 = ෍ |𝑎𝑘 |2
𝑁
𝑙= 𝑁 𝑙= 𝑁 53
 Properties of discrete-time FS
 Examples 𝑥 𝑛 = 𝑥1 𝑛 + 𝑥2 [𝑛]

 𝑥1 𝑛 is a square wave with 𝑁 = 5 and

𝑁1 = 1
2𝑁1 + 1 3
, 𝑘 = ±𝑁, ±2𝑁, … , 𝑘 = ±5, ±10, …
𝑁 5
𝑏𝑘 = 1 sin 2𝑘𝜋(𝑁 + 1/2)/𝑁 = 1 sin 3𝑘𝜋/5
1
, else , else
𝑁 sin 𝑘𝜋/𝑁 5 sin 𝑘𝜋/5

 For 𝑥2 𝑛
8
, 𝑘 = ±5, ±10, …
1, 𝑘 = ±𝑁, ±2𝑁, … 5
𝑐𝑘 = ቊ ∴ 𝑎𝑘 = 𝑏𝑘 + 𝑐𝑘 = 1 sin 3𝑘𝜋/5
0, else
, else
5 sin 𝑘𝜋/5 54
 Properties of discrete-time FS
 Examples

 Solution
1
• σ5𝑛=0 𝑥[𝑛] = 2 ⟹ 𝑎0 = σ 𝑥 𝑛 𝑒 −𝑗0 2𝜋/𝑁 𝑛 = 1/3.
𝑁 𝑛= 𝑁
• σ7𝑛=2 −1 𝑛 𝑥[𝑛] = 1 ⟹ σ𝑛= 𝑁 𝑥 𝑛 𝑒 −𝑗3 2𝜋/𝑁 𝑛 = 1 ⟹ 𝑎3 = 1/6

• from 4, 𝑎1 = 𝑎2 = 𝑎4 = 𝑎5 = 0
1 1 1 1
• ∴ 𝑥 𝑛 = 𝑎0 𝑒 −𝑗0 2𝜋/𝑁 𝑛 + 𝑎3 𝑒 −𝑗3 2𝜋/𝑁 𝑛 = + 𝑒 −𝑗𝜋𝑛 = + (−1)𝑛
3 6 3 6 55
Fourier Series Representation of Periodic
signals (ch.3)
 The response of LTI systems to complex exponentials
 Fourier series representation of continuous periodic signals
 Convergence of the Fourier series
 Properties of continuous-time Fourier series
 Fourier series representation of discrete –time periodic signals
 Properties of discrete FS
Fourier series and LTI systems
 Fourier series and LTI systems
 Recall ∞

𝑒 𝑠𝑡 LTI 𝐻(𝑠)𝑒 𝑠𝑡 𝐻 𝑠 = න ℎ 𝜏 𝑒 −𝑠𝜏 𝑑𝜏

−∞
∞

𝑧𝑛 LTI 𝐻[𝑧]𝑧 𝑛 𝐻 𝑧 = ෍ ℎ[𝑘]𝑧 −𝑘

𝑘=−∞

 System functions: 𝐻 𝑠 and 𝐻 𝑧

For periodic signal, CT Fourier Series (Ch3)
𝑠 pure imagery 𝑒 𝑠𝑡 → 𝑒 𝑗𝜔𝑡 For aperiodic signal, CT Fourier Transform (Ch4)
𝑒 𝑠𝑡 𝑠 complex number 𝑒 𝑠𝑡 Laplase Transform (Ch9)
For periodic signal, DT Fourier Series (Ch3)
𝑛 𝑗𝜔𝑛
𝑧 pure imagery 𝑧 → 𝑒 For aperiodic signal, DT Fourier Transform (Ch5)
𝑧𝑛
𝑧 complex number 𝑧 𝑛 Z-Transform (Ch10) 57
 Fourier series and LTI systems
 Frequency response for CT system: 𝐻 𝑗𝜔
∞ 𝑠=𝑗𝜔 ∞
𝐻 𝑠 = න ℎ 𝜏 𝑒 −𝑠𝜏 𝑑𝜏 𝐻 𝑗𝜔 = න ℎ 𝜏 𝑒 −𝑗𝜔𝜏 𝑑𝜏
−∞ −∞

𝑒 𝑗𝜔𝑡 LTI 𝐻(𝑗𝜔)𝑒 𝑗𝜔𝑡

∞ ∞

𝑥 𝑡 = ෍ 𝑎𝑘 𝑒 𝑗𝑘𝜔0 𝑡 LTI 𝑦 𝑡 = ෍ 𝑎𝑘 𝐻(𝑗𝑘𝜔0 )𝑒 𝑗𝑘𝜔0 𝑡

ℱ𝒮 𝑘−∞ ℱ𝒮 𝑘−∞
𝑎𝑘 𝑏𝑘
𝑏𝑘 = 𝑎𝑘 𝐻(𝑗𝑘𝜔0 )
58
 Fourier series and LTI systems
 Frequency response for CT system: example
1 1 1
𝑥 𝑡 = σ3𝑘=−3 𝑎𝑘 𝑒 𝑗𝑘2𝜋𝑡 (𝑎0 = 1, 𝑎1 = 𝑎−1 = , 𝑎2 = 𝑎−2 = , 𝑎3 = 𝑎−3 = ) is
4 2 3
the input of a LTI system with ℎ 𝑡 = 𝑒 −𝑡 𝑢(𝑡), determine 𝑦 𝑡
 Solution ∞ ∞
1
𝑦 𝑡 = ෍ 𝑎𝑘 𝐻(𝑗𝑘𝜔0 )𝑒 𝑗𝑘2𝜋𝑡 𝐻 𝑗𝜔 = න 𝑒 −𝜏 𝑒 −𝑗𝜔𝜏 𝑑𝜏 =
0 1 + 𝑗𝜔
𝑘=−∞
1 1 1 1 1
𝑏𝑘 = 𝑎𝑘 𝐻 𝑗𝑘𝜔0 = 𝑎𝑘 𝑏1 = 𝑏−1 =
1 + 𝑗𝑘2𝜋 𝑏0 = 1 ∙ 1 = 1 4 1 + 𝑗2𝜋 4 1 − 𝑗2𝜋
1 1 1 1 1 1 1 1
𝑏2 = 𝑏−2 = 𝑏3 = 𝑏−3 =
2 1 + 𝑗4𝜋 2 1 − 𝑗4𝜋 3 1 + 𝑗6𝜋 3 1 − 𝑗6𝜋
59
 Fourier series and LTI systems

 Frequency response DT system: 𝐻 𝑒 𝑗𝜔

∞ ∞
𝑧=𝑒 𝑗𝜔
𝐻 𝑧 = ෍ ℎ[𝑘]𝑧 −𝑛 𝐻 𝑒 𝑗𝜔 = ෍ ℎ[𝑛]𝑒 −𝑗𝜔𝑛
𝑛=−∞ 𝑛=−∞

𝑒 𝑗𝜔𝑛 LTI 𝐻 𝑒 𝑗𝜔 𝑒 𝑗𝜔𝑛

𝑥[𝑛] = ෍ 𝑎𝑘 𝑒 𝑗𝑘 2𝜋/𝑁 𝑛 𝑦[𝑛] = ෍ 𝑎𝑘 𝐻 𝑒 𝑗𝑘 2𝜋/𝑁 𝑒 𝑗𝑘 2𝜋/𝑁 𝑛

𝑘= 𝑁
LTI 𝑘= 𝑁

𝑏𝑘 = 𝑎𝑘 𝐻(𝑒 𝑗𝑘 2𝜋/𝑁 ) 60
 Fourier series and LTI systems
 Frequency response DT system: example
ℎ 𝑛 = 𝛼𝑛𝑢 𝑛 , 𝛼 < 1
2𝜋𝑛
𝑥[𝑛] = cos LTI 𝑦𝑛?
𝑁
 Solution
1 𝑗 2𝜋/𝑁 𝑛
1 −𝑗 2𝜋/𝑁 𝑛
𝑥𝑛 = 𝑒 + 𝑒
2 2
∞ ∞
𝑗𝜔 −𝑗𝜔𝑛 𝑛 −𝑗𝜔𝑛
1
𝐻 𝑒 = ෍ ℎ[𝑛]𝑒 = ෍𝛼 𝑒 =
1 − 𝛼𝑒 −𝑗𝜔
𝑛=−∞ 𝑛=0

1 1 𝑗 2𝜋/𝑁 𝑛
1 1 −𝑗 2𝜋/𝑁 𝑛
𝑥𝑛 = 𝑒 + 𝑒
2 1 − 𝛼𝑒 −𝑗2𝜋/𝑁 2 1 − 𝛼𝑒 𝑗2𝜋/𝑁
61