EE 220 : Signals and Systems

Department of Electronics and Electrical Engineerng

Indian Institute of Technology, Guwahati
Monsoon 2023

Topic 15 : Properties of discrete time Fourier series

Instruction and notes by : Manish

1 Properties of discrete-time Fourier series

Just like we have seen for CTFS and CTFT in previous topics, the discrete time Fourier series
(DTFS) also follow similar properties. These are:

• Linearity • Periodic convolution

• Time shifting • Multiplication

• Frequency shifting • Even and odd symmetry

• Reflection • Conjugate symmetry

• Conjugation • Parseval’s theorem

Revision: A periodic sequence x[n] with a fundamental period N and fundamental fre-
quency Ωo = 2π/N can be represented as
N
X −1
x[n] = X[k]ejkΩo n ; (1)
k=0

where
N −1
1 X
X[k] = x[n]e−jkΩo n (2)
N n=0
are the DTFS coefficients of the signal x[n] and n, k ∈ Z.

Let’s assume following Fourier pairs before we learn these properties one by one:
DTFS
x[n] ←−−→ X[k]
DTFS
y[n] ←−−→ Y [k]

1.1 Linearity
A linear combination of sequences produces the same linear combination of their Fourier series
coefficients; i.e.
DTFS
αx[n] + βy[n] ←−−→ αX[k] + βY [k]

1
1.2 Time shifting (Translation)

DTFS
x[n − n0 ] ←−−→ e−jkΩ0 n0 X[k]

1.3 Frequency shifting (Modulation)

DTFS
X[k − k0 ] ←−−→ ejΩ0 k0 n x[n]

1.4 Time reversal (Reflection)

Time reversal of a sequence results in a time reversal of its Fourier series coefficients.
DTFS
x[−n] ←−−→ X[−k]

1.5 Conjugation
Conjugating a sequence has the effect of time reversing and conjugating the Fourier series
coefficient sequence.
DTFS
x∗ [n] ←−−→ X ∗ [−k]

1.6 Periodic convolution

Let x and y be two periodic sequences with the same period N.
Periodic convolution of two sequences corresponds to the multiplication (up to a scale factor)
of their Fourier-series coefficient sequences.
DTFS
x ⊛ y[n] ←−−→ N · X[k] · Y [k]

1.7 Multiplication
Let x and y be two periodic sequences with the same period N. The multiplication of two
periodic sequences corresponds to the DT convolution of their corresponding Fourier-series
coefficient sequences.

∞
DTFS
X
x[n] · y[n] ←−−→ X[n]Y [k − n] = X ⊛ Y [k]
n=−∞

The above summation is the DT convolution of X[k] and Y[k].

2
1.8 Even and odd symmetry
The even/odd symmetry properties of x[n] and X[k] always match.
DTFS
If x[n] is even ←−−→ X[k] is even
DTFS
If x[n] is odd ←−−→ X[k] is odd

1.9 Conjugate symmetry

A sequence x[n] is real if and only if its Fourier series coefficient sequence X[k] satisfies

X[k] = X ∗ [−k] ∀k ∈ Z

i.e., X[k] is conjugate symmetric. Note that x[n] being real does not necessarily imply that
X[k] is real.

From properties of complex numbers, one can show that X[k] = X ∗ [−k] is equivalent to

|X[k]| = |X ∗ [−k]| and arg X[k] = − arg X[−k]

i.e. magnitude part is even and argument is odd symmetric.

Thus, for a real-valued sequence, the negative-indexed Fourier series coefficients are redun-
dant, as they can be completely determined by the nonnegative-indexed coefficients.

1.10 Duality
DTFS
If x[n] ←−−→ X[k] then

DTFS 1
X[n] ←−−→ x[−k]
N

1.11 Parseval’s theorem

DTFS
If x[n] ←−−→ X[k] then
N −1 N −1
1 X 2
X
|x[n]| = |X[k]|2
N n=0 k=0

1.12 Other properties of a Fourier series

1. X[0] is the average value of x[n] over a single period T ;
DTFS
2. x[n] is real and even ←−−→ X[k] is real and even; and
DTFS
3. x[n] is real and odd ←−−→ X[k] is purely imaginary and odd.

3
1.13 Examples
Example 1: Find the DTFS representation of the signal
3 1
z[n] = x[n] + y[n] (3)
2 2
given that
DTFS
x[n] ←−−→ X[k] = e(jkπ/4)
DTFS
y[n] ←−−→ Y [k] = e(jkπ/2)
Solution: This is based on linearity property. The answer is:
DTFS 3 1
z[n] ←−−→ Z[k] = e(jkπ/4) + e(jkπ/2)
2 2

Example 2: Find the DTFS representation of the signal x[n] = sin[2πn/3]. What will be
DTFS representation of sin[2π(n − 2)/3]?
Solution:
ej2πn/3 − e−j2πn/3
sin[2πn/3] =
2j
Now, let’s compare with the DTFS form (Ω0 = 2π/3 i.e. N = 6 here) in the interval [-2,3]:
N
X −1
x[n] = X[k]ejkΩo n
k=0

We can notice that


−1
 2j , k = −1

X[k] = 2j1 , k = 1

0 elsewhere in the interval k= [-2,3]


So FS coefficients of sin[2π(n − 2)/3] should be

 j4π/3
−e
 2j , k = −1

−j4π/3
= e 2j , k = 1

0 elsewhere in the interval k= [-2,3]


Homework: Find the DTFS representation of the signal x[n] = 1. What will be DTFS
representation of the signal x[n] = ejpΩo n ?

2 Fourier series and frequency spectra

2.1 The frequency domain
The Fourier series coefficients of a sequence x provide a means to quantify how much informa-
tion x has at different frequencies. The distribution of information in a sequence over different
frequencies is referred to as the frequency spectrum of the sequence.

Note that:

4
 1. The Fourier series coefficients X[k] are referred to as the frequency spectrum of x[n].

2. The magnitudes |X[k]| of the Fourier series coefficients are referred to as the magnitude
spectrum of x[n].

3. The arguments arg[X[k]] of the Fourier series coefficients are referred to as the phase
spectrum of x[n].

4. Normally, the spectrum of a sequence is plotted against frequency kΩ0 instead of k.

5. Since the Fourier series only has frequency components at integer multiples of the fun-
damental frequency, the frequency spectrum is discrete

6. Due to the general appearance of frequency-spectrum plot (i.e., a number of vertical lines
at various frequencies), we refer to such spectra as line spectra.

3 Frequency response
For a LTI system with impulse response h, and an input ejΩn

ejΩn ∗ h[n] = H(Ω)ejΩn (4)

where Ω is a real constant and ∞

X
H(Ω) = h[n]e−jΩn (5)
−∞

i.e., ejΩn is an eigensequence of a LTI system and H(Ω) is the corresponding eigenvalue. We
refer to H(Ω) as the frequency response of the system and the above equation is the relationship
between frequency response and impulse response.

5
Appendix : Basic DTFS pairs

6
Appendix : List of DTFT and DTFS properties

Reference material
1. Textbook: Signals and Systems by Simon Haykin
2. Lecture notes are inspired from the course materials of JHU 520.214; MIT 6.003; Purdue
ECE-301; UVic ECE-260; and Imperial College E2.5