Discrete Time Fourier Series (DTFS) and Fourier Transform

(DTFT)

Dr. J. Aravinth (Mentor)

Ms. Devi Vijayan
Recall: Four Classes of Fourier Representations
Recall: Frequency Response of DT Systems
Consider a discrete-time LTI system with impulse response ℎ[𝑛].

Suppose a complex sinusoid is applied as input. i.e., 𝑥[𝑛] = 𝑒

𝑗Ω𝑛
.
Then, the output 𝑦[𝑛] of the DT LTI system is given by
∞ ∞

𝑗Ω(𝑛−𝑘)
𝑦[𝑛] = ℎ[𝑘]𝑥[𝑛 − 𝑘] = ℎ[𝑘]𝑒
∑ ∑
𝑘=−∞ 𝑘=−∞

𝑗Ω𝑛 −𝑗Ω𝑘 𝑗Ω 𝑗Ω𝑛

= 𝑒 ℎ[𝑘]𝑒 = 𝐻(𝑒 )𝑒 , where
∑
𝑘=−∞

𝑗Ω −𝑗Ω𝑘
𝐻(𝑒 ) = ℎ[𝑘]𝑒 = Frequency Response.
∑
𝑘=−∞
Recall: Frequency Response of DT Systems
𝑗Ω

Substituting 𝐻(𝑒
𝑗Ω
) = |𝐻(𝑒
𝑗Ω
)|𝑒
𝑗 arg{𝐻(𝑒 )}
, we obtain the output of the DT LTI system
corresponding to a complex sinusoid 𝑒𝑗Ω𝑛 as

𝑗Ω
𝑗Ω 𝑗Ω𝑛 𝑗Ω 𝑗(Ω𝑛+arg{𝐻(𝑒 )})
𝑦[𝑛] = 𝐻(𝑒 )𝑒 = |𝐻(𝑒 )|𝑒 ,

which implies that in steady state the DT LTI system modi es the magnitude of the input by
a factor |𝐻(𝑒𝑗Ω )| and modi es the phase by a shift of arg{𝐻(𝑒𝑗Ω )} .

Therefore, |𝐻(𝑒𝑗Ω )| is called the magnitude response and arg{𝐻(𝑒𝑗Ω )} is called the
phase response of the DT LTI system.
Discrete-Time Fourier Series (DTFS) for Periodic
Signals
Harmonically-Related DT Complex Sinusoids
DTFS De nition

Notation: Here, Ω0 = (2𝜋/𝑁) denotes the fundamental (angular) frequency of

the signal 𝑥[𝑛], and Ω𝑘 = 𝑘Ω0 = (2𝑘𝜋/𝑁) denotes its k-th harmonic frequency.
DTFS De nition

Notation:

Here, 𝑎𝑘 = 𝑋[𝑘] ,𝑘 = 0, … , 𝑁 − 1 , denote the DTFS coef cients.

Since both 𝑥[𝑛] and 𝑒𝑗(2𝜋/𝑁)𝑘𝑛 = 𝑒𝑗Ω 𝑘𝑛 are periodic functions with period 𝑁 , the
0

DTFS coef cients 𝑎𝑘 or 𝑋[𝑘] , as a function of index 𝑘, is also periodic with period
𝑁.
DTFS De nition
To denote that the DT signal (or, sequence) 𝑥[𝑛] has the Fourier series coef cient
sequence 𝑋[𝑘] , we write
DTFS
𝑥[𝑛] ⟷ 𝑋[𝑘].

The DTFS is the only Fourier representation that can be numerically evaluated and
manipulated in a computer.

This is because both the time and the frequency domain representations of the
signal are exactly characterized by a nite set of 𝑁 numbers.

DTFS is used for approximating the other three Fourier representations for the
purpose of implementation on a computer.
DTFS Examples and Exercises
Example: (Method of Inspection) Determine the DTFS coef cients of
𝑥[𝑛] = cos(𝜋𝑛/3 + 𝜙) . Sketch the magnitude and phase spectrum.

Solution: The period is 𝑁 = 6. The fundamental frequency is

Ω0 = 2𝜋/𝑁 = 𝜋/3 . The DTFS coef cients are:

𝑗𝜙 −𝑗𝜙
𝑋[1] = 𝑒 /2, 𝑋[−1] = 𝑒 /2.

Can you draw the magnitude and phase plot?

Exercise: Determine the DTFS coef cients of 𝑥[𝑛] = 1 + sin(𝜋𝑛/12 + 3𝜋/8) . Sketch the
magnitude and phase spectrum.
DTFS Examples and Exercises
Example: Determine the DTFS coef cients of the 𝑁 -periodic impulse train given by
∞

𝑥[𝑛] = 𝛿[𝑛 − ℓ𝑁].

∑
ℓ=−∞

Sketch the magnitude and phase spectrum.

Solution: We have

𝑁−1 𝑁−1
1 −𝑗𝑘𝑛2𝜋/𝑁
1 −𝑗𝑘𝑛2𝜋/𝑁
1
𝑋[𝑘] = 𝑥[𝑛]𝑒 = 𝛿[𝑛]𝑒 = .
𝑁 ∑ 𝑁 ∑ 𝑁
𝑛=0 𝑛=0
DTFS Examples and Exercises
Example: Determine the DTFS coef cients of the signal shown below. Sketch the
magnitude and phase spectrum.

Solution: We have 𝑁 = 10 and Ω0 = 2𝜋/𝑁 = 𝜋/5 . Therefore,

the DTFS coef cients are given by
DTFS Examples and Exercises

Solution: (continued..)

𝑁−1 4
1 −𝑗𝑘𝑛2𝜋/𝑁
1 −𝑗𝑘𝑛𝜋/5
𝑋[𝑘] = 𝑥[𝑛]𝑒 = 𝑒
𝑁 ∑ 10 ∑
𝑛=0 𝑛=0

−𝑗𝑘𝜋 −𝑗𝑘(𝜋/2) 𝑗𝑘(𝜋/2) −𝑗𝑘(𝜋/2)

1 1 − 𝑒 1 𝑒 𝑒 − 𝑒 .
= =
−𝑗𝑘(𝜋/5) −𝑗𝑘(𝜋/10) 𝑗𝑘(𝜋/10) −𝑗𝑘(𝜋/10)
10 1 − 𝑒 10 𝑒 𝑒 − 𝑒

1 −𝑗𝑘(2𝜋/5)
sin(𝑘(𝜋/2)
= 𝑒 , 𝑘 = 0, 1, . . . , 9.
10 sin(𝑘(𝜋/10)
DTFS Trigonometric Form
Suppose 𝑁 is even; then, we can rewrite the DTFS by separating the terms corresponding
to 𝑘 = 0 and 𝑘 = 𝑁/2 from the remaining terms as
𝑁/2

𝑗𝑘Ω0 𝑛
𝑥[𝑛] = 𝑋[𝑘]𝑒
∑
𝑘=−(𝑁/2)+1

(𝑁/2)−1

𝑗𝜋𝑛 𝑗𝑘Ω0 𝑛 −𝑗𝑘Ω0 𝑛

= 𝑋[0] + 𝑋[𝑁/2]𝑒 + 𝑋[𝑘]𝑒 + 𝑋[−𝑘]𝑒
∑
𝑘=1

(𝑁/2)−1

= 𝑋[0] + 𝑋[𝑁/2] cos(𝜋𝑛) + {𝐵[𝑘] cos(𝑘Ω0 𝑛) + 𝐴[𝑘] sin(𝑘Ω0 𝑛)},

∑
𝑘=1

where 𝐵[𝑘] = 𝑋[𝑘] + 𝑋[−𝑘] and 𝐴[𝑘] = 𝑗(𝑋[𝑘] − 𝑋[−𝑘]) .

DTFS Trigonometric Form
Suppose 𝑁 is odd; then, we can rewrite the DTFS by separating the term corresponding to
𝑘 = 0 from the remaining terms as

(𝑁−1)/2

𝑗𝑘Ω0 𝑛
𝑥[𝑛] = 𝑋[𝑘]𝑒
∑
𝑘=−(𝑁−1)/2

(𝑁−1)/2

𝑗𝑘Ω0 𝑛 −𝑗𝑘Ω0 𝑛
= 𝑋[0] + 𝑋[𝑘]𝑒 + 𝑋[−𝑘]𝑒
∑
𝑘=1

(𝑁−1)/2

= 𝑋[0] + {𝐵[𝑘] cos(𝑘Ω0 𝑛) + 𝐴[𝑘] sin(𝑘Ω0 𝑛)},

∑
𝑘=1

where 𝐵[𝑘] = 𝑋[𝑘] + 𝑋[−𝑘] and 𝐴[𝑘] = 𝑗(𝑋[𝑘] − 𝑋[−𝑘]) .

DTFS Properties
Linearity
Translation (Time Shifting)
Modulation (Frequency Shifting)
Re ection (Time Reversal)
Conjugation
Duality
Periodic Convolution

Notation: Here, the circular convolution of 𝑥[𝑛] and 𝑦[𝑛] is given by

𝑁−1

𝑥[𝑛] ⊛ ℎ[𝑛] = 𝑥[𝑚]𝑦[𝑛 − 𝑚].

∑
𝑚=0
Multiplication

Notation: Here, the circular convolution of 𝑎𝑘 = 𝑋[𝑘] and 𝑏𝑘 = 𝑌 [𝑘] is given by

𝑁−1

𝑋[𝑘] ⊛ 𝑌 [𝑘] = 𝑋[𝑚]𝑌 [𝑘 − 𝑚].

∑
𝑚=0
Parseval's relation
Other Properties
Other Properties
Other Properties
Discrete Time Fourier Transform (DTFT)
DTFT from DTFS
FT and Inverse FT Equations
DTFT Convergence
DTFT Examples and Exercises
Example: Find the DTFT of the signal 𝑥[𝑛] 𝑛
= 𝛼 𝑢[𝑛] . Sketch the magnitude and phase
spectrum assuming 𝛼 to be real-valued.

Solution: The DTFT is given by

∞ ∞

𝑗Ω 𝑛 −𝑗Ω𝑛 𝑛 −𝑗Ω𝑛
𝑋(𝑒 ) = 𝛼 𝑢[𝑛]𝑒 = 𝛼 𝑒
∑ ∑
𝑛=−∞ 𝑛=0

1
= , |𝛼| < 1.
−𝑗Ω
1 − 𝛼𝑒

If 𝛼 is real-valued and |𝛼| < 1 , then we have

𝑗Ω
1
𝑋(𝑒 ) =
1 − 𝛼 cos(Ω) + 𝑗𝛼 sin(Ω)
DTFT Examples and Exercises
𝑗Ω
1
⇒ |𝑋(𝑒 )| = , and
1/2
2 2 2
((1 − 𝛼 cos(Ω)) + 𝛼 sin (Ω))

𝑗Ω −1
𝛼 sin(Ω)
arg{𝑋(𝑒 )} = − tan
( 1 − 𝛼 cos(Ω) )

Can you draw the magnitude and phase plot? (see next slide)

Exercise: Determine the DTFT of the following signals:

(𝑖) 𝑥[𝑛] = 𝛿[𝑛], (𝑖𝑖) 𝑥[𝑛] = 1, (𝑖𝑖𝑖) cos(Ω0 𝑛), (𝑖𝑣) sin(Ω0 𝑛).

Exercise: Determine the inverse DTFT of the spectrum

𝑗Ω
1, |Ω| < 𝑊
𝑋(𝑒 ) = .
{ 0, 𝑊 < |Ω| < 𝜋
Periodic Magnitude and Phase Spectra
DTFT Properties
Periodicity
Linearity
Translation or Time Shifting
Modulation or Frequency Shifting
Conjugation and Time Reversal
Convolution
Parseval's Relation
Other Properties
Other Properties
DT LTI Systems Given By Di erence Equations
DTFT Properties: Examples and Exercises
Example: Find the output of a DT LTI system whose impulse response is given by
ℎ[𝑛] = (1/(𝜋𝑛)) sin(𝜋𝑛/2) if the input is 𝑥[𝑛] = (1/2) 𝑢[𝑛] .
𝑛

Solution: The DTFT of the input is given by

𝑗Ω
1
𝑋(𝑒 ) = .
−𝑗Ω
1 − (1/2)𝑒

We have

sin(𝜋𝑛) DTFT
= sinc(𝑛) ⟷ rect(Ω/(2𝜋))
𝜋𝑛

sin(𝜋𝑛/2) DTFT
𝑗Ω
⇒ ℎ[𝑛] = = (1/2)sinc(𝑛/2) ⟷ rect(Ω/𝜋) = 𝐻(𝑒 )
𝜋𝑛
DTFT Properties: Examples and Exercises

Therefore, the DTFT of the output is given by

𝑗Ω 𝑗Ω 𝑗Ω
rect(Ω/𝜋)
𝑌 (𝑒 ) = 𝑋(𝑒 )𝐻(𝑒 ) =
−𝑗Ω
1 − (1/2)𝑒

⎧ 1
⎪ |Ω| ≤ 𝜋/2
−𝑗Ω
= ⎨ 1 − (1/2)𝑒
⎪
⎩ 0, 𝜋/2 < |Ω| ≤ 𝜋

Example: Consider the two-path communication channel given by the input-output

relationship
𝑦[𝑛] = 𝑥[𝑛] + 𝑎𝑥[𝑛 − 1], |𝑎| < 1.

Find the impulse response of the inverse system by the convolution property of DTFT.
DTFT Properties: Examples and Exercises

Solution: The impulse response of the system is

ℎ[𝑛] = 𝛿[𝑛] + 𝑎𝛿[𝑛 − 1].

The frequency response of the system is

𝑗Ω −𝑗Ω
𝐻(𝑒 ) = 1 + 𝑎𝑒 ,

which implies that

𝑖𝑛𝑣 𝑗Ω
1 𝑖𝑛𝑣 𝑛
𝐻 (𝑒 ) = ⇒ ℎ [𝑛] = (−𝑎) 𝑢[𝑛].
−𝑗Ω
1 + 𝑎𝑒
DTFT Properties: Examples and Exercises
Example: Use the frequency-differentiation property and nd out the DTFT of the signal
𝑛
𝑥[𝑛] = (𝑛 + 1)𝛼 𝑢[𝑛], |𝛼| < 1.

Solution: We have
𝑛
DTFT 1
𝛼 𝑢[𝑛] ⟷
−𝑗Ω
1 − 𝛼𝑒

−𝑗Ω
𝑛
DTFT 𝑑 1 𝛼𝑒
⇒ 𝑛𝛼 𝑢[𝑛] ⟷ 𝑗 =
𝑑Ω ( 1 − 𝛼𝑒
−𝑗Ω ) −𝑗Ω 2
(1 − 𝛼𝑒 )

−𝑗Ω
𝑛
DTFT 1 𝛼𝑒 1
⇒ (𝑛 + 1)𝛼 𝑢[𝑛] ⟷ + =
−𝑗Ω −𝑗Ω 2 −𝑗Ω 2
1 − 𝛼𝑒 (1 − 𝛼𝑒 ) (1 − 𝛼𝑒 )
DTFT Table of Properties
DTFT Table of Properties
DTFT of Common DT Signals
References:
[1] Simon Haykin and Barry Van Veen, Signals and Systems, Second
Edition, John Wiley and Sons, 2003.

[2] Lecture Notes by Michael D. Adams.

https://www.ece.uvic.ca/~frodo/sigsysbook/downloads/lecture_slides_for_signals_and_systems
2.0.pdf
(https://www.ece.uvic.ca/~frodo/sigsysbook/downloads/lecture_slides_for_signals_and_system
2.0.pdf)

[3] Lecture Notes by Richard Baraniuk.

https://www.di.univr.it/documenti/OccorrenzaIns/matdid/matdid018094.pdf
(https://www.di.univr.it/documenti/OccorrenzaIns/matdid/matdid018094.pdf)