Chapter 4: Discrete-Time Signals in Transform

Domain
• Introduction
• Frequency Analysis of Discrete-time Signals
• Discrete-time Fourier Transform (DTFT)
• Properties of DTFT
• DTFT Theorems
• Discrete Fourier Transform (DFT)
• Properties of DFT
• Fast Fourier Transform(FFT)
• z –Transform
• Properties of z-transform
• Inverse z-transforms ASTU | ECE 3205 |A.Y 2022 - 2023
Introduction
• Fourier series
 used to represent the frequency contents of a periodic and continuous-time signal.
 A continuous-time function x(t) is said to be periodic if there exists Tp>0 such that
x(t) = x(t+Tp), tє (-∞, ∞)
 The smallest Tp for which the above equation holds is the fundamental period.
 Every periodic signal can be expanded into a Fourier series
x(t) = ,tє(, )
Where
=

and = is called the fundamental frequency.

ASTU | ECE 3205 |A.Y 2022 - 2023

Introduction
• Frequency analysis of continuous-time periodic signals.

ASTU | ECE 3205 |A.Y 2022 - 2023

Introduction
• Fourier Transform:
 is used to represent the frequency content of an aperiodic and continuous-
time signals x(t)
 Frequency analysis of continuous-time aperiodic signals is given as:

Synthesis equation inverse transform

Analysis equation Direct transform X(j) = dt

ASTU | ECE 3205 |A.Y 2022 - 2023

Introduction
Note:
• Since the frequency range of continuous-time signals extend from it
is possible to have signals that contain an infinite number of
frequency components.
• In contrast, frequency range for discrete-time signals is unique over
the interval (-π, π ) or (0, 2π ).
• Consequently, the Fourier series representation of the discrete-time
periodic signal will contain at most N frequency components (if N; is
the fundamental period of the discrete-time signal)

ASTU | ECE 3205 |A.Y 2022 - 2023

Frequency Analysis of Discrete-Time Signals
• The Fourier series for discrete-time periodic signals(DTFS):
 Suppose that we are given a periodic sequence x[n], with period
N, i.e.,
x[n] = x[n + N] for all.
 The Fourier series representation for x[n] consists of N harmonically
related exponential functions:

, k = 0, 1, . . . ,N-1 and is expressed as

x[n] = . . . . . . . . .(*)
Where { } are the coefficients in the series representation.
ASTU | ECE 3205 |A.Y 2022 - 2023
Frequency Analysis of Discrete-Time Signals
• To obtain the expression for the Fourier coefficients:
= N, k =
0, otherwise
from geometric series summation formula:
= N , a=1
,

• Multiply both sides of equation (*) by and sum the product from
n =0 to n =N-1:

ASTU | ECE 3205 |A.Y 2022 - 2023

Frequency Analysis of Discrete-Time Signals
=

= N, k-l = 0,
0, otherwise

Hence = , l =0, 1, . . .,N-1

ASTU | ECE 3205 |A.Y 2022 - 2023

Frequency Analysis of Discrete-Time Signals
• Frequency Analysis of Discrete-time periodic signals
Synthesis equation X[n]=

Analysis equation =

• Fourier coefficients {Ck}, k=0,1, . . .,N-1:

 provides the description of x[n] in the frequency domain, representing the
amplitude and phase associated with the frequency component.
ASTU | ECE 3205 |A.Y 2022 - 2023
Frequency Analysis of Discrete-Time Domain
 When viewed beyond the range k=0,1, . ..,N-1, also satisfy a periodicity
condition.
+ N = , i.e. is a periodic sequence with fundamental period N.
Power spectral Density of Periodic Signals:
• The average power of a discrete-time periodic signal with period N:
=
• We shall now derive an expression of Px in terms of the Fourier coefficients {}
=
= ) {}
= ]
ASTU | ECE 3205 |A.Y 2022 - 2023
Power Spectral Density of Periodic Signals
= =
• The average power in the signal is the sum of the powers of the individual
frequency components.
• The sequence for k = 0, 1, 2, . . ., N-1 is the distribution of power as a function of
frequency
• Energy of the sequence x[n] over a single period,
=N
Exercise: Determine the spectral of the signal x[n] {1, 1, 0, 0} and with period N=4
and plot the spectral content of this periodic signal.

ASTU | ECE 3205 |A.Y 2022 - 2023

Discrete-Time Fourier Transform
• The frequency analysis tool for aperiodic and discrete-time signals is discrete-
time Fourier Transform (DTFT).
• The Discrete-time Fourier transform (DTFT) of sequence x[n]:
X() =
• Physically, X() represents the frequency content of the signal x[n]. In other words,
X() is a decomposition of x[n] into its frequency components
• X() is periodic with period 2π, i.e.,
X( + 2πk) =
=
= = X()

ASTU | ECE 3205 |A.Y 2022 - 2023

Discrete-Time Fourier Transform
• This property is just a consequence of the fact that frequency range for
any discrete-time signal is limited to (- , ) or (0,), and any frequency
outside this interval is equivalent to a frequency within the interval.
• The inverse DTFT, is a synthesis formula:
x[n] = d
• In general, DTFT is a complex-valued function of ,
X() = () + j()
or in polar form:
X() = | X() |

ASTU | ECE 3205 |A.Y 2022 - 2023

Continued
• | X() | and arg( are the magnitude and phase, respectively of the DTFT.
• If x[n] is absolutely summable, then X() exists. i.e.,
if then
|X(ω)| =||≤
 Absolute summability is the sufficient for the existence of DTFT
Example:
a. Let x[n] = u[n]
X() == =
if || or |a|

ASTU | ECE 3205 |A.Y 2022 - 2023

 Symmetry properties of the DTFT
• A discrete-time Fourier transform X() can be decomposed into sum of
conjugate-symmetric and conjugate-antisymmetric functions as:
X() = () where
) = [)] and
() =
• Note that ) is conjugate symmetric and () is conjugate antisymmetric,
i.e.,
) = ) and
() = - X()

ASTU | ECE 3205 |A.Y 2022 - 2023

 Symmetry properties of the DTFT
Sequence x[n] DTFT X)
)
2. )
3. {x[n]} )
4. j{x[n]} )
5. ={X()}
6. j=j{X()}

ASTU | ECE 3205 |A.Y 2022 - 2023

 Symmetry properties of the DTFT
• The following properties apply only when x[n] is real:
7. Any real x[n] X() =()
8. Any real x[n] = ,real part is even
9. Any real x[n] = ,imaginary part is odd

10. Any real x[n] |X() | = |X() |, magnitude is even

11. Any real x[n] arg(X()) = -X(), phase is odd
12.
13. (odd part of x[n]) j)
ASTU | ECE 3205 |A.Y 2022 - 2023
Discrete-time Fourier Transform Theorems
Sequence x[n], y[n] DTFT x(), y()
1. ax[n] + by[n] aX() + bY()
2. x[n-] ( is an integer) X()
3. x[n] X()
4. x[-n] X(),(), if x[n] is real
5. nx[n] j
6. x[n]*y[n] X()Y()

ASTU | ECE 3205 |A.Y 2022 - 2023

Discrete-Time Fourier Transform Theorems
7. x[n]y[n] Y() d

Parserval’s theorem
8. = d
9. =d
The above theorems are:
1. Linearity of DTFT
2 & 3. Time shifting and Frequency shifting

ASTU | ECE 3205 |A.Y 2022 - 2023

Fundamental DTFT Pairs
4. Time reversal
5. Differentiation in Frequency
6. The Convolution theorem
7. The modulation or windowing theorem
Sequence DTFT
1. δ[n] 1
2. δ[n -]
3. 1 (-∞ )
4. u[n](|a| 1) )

ASTU | ECE 3205 |A.Y 2022 - 2023

Fundamental DTFT pairs
5 u[n] +)
6. (n +1)u[n](|a| )
7. u[n](|r| ) .

8. X()= 1, ||
0, |
9. x[n] = 1, 0 .
0, otherwise
10. )

ASTU | ECE 3205 |A.Y 2022 - 2023

Continued
11. (+k)
+(+k)]

• Exercise: Find the DTFT of the following discrete-time signal; and plot X() in
MATLAB for =
x[n] = 1,|n|
0,|n|

ASTU | ECE 3205 |A.Y 2022 - 2023

The Discrete Fourier Transform (DFT)
• To perform frequency analysis of a discrete-time signal {x[n]}, we convert the
time-domain sequence to an equivalent frequency-domain representation.
 DTFT of the sequence x[n]
• How ever, x() is a continuous function of frequency and therefore its not
computationally convenient representation of the sequence {x[n]}.
• Consider the representation of sequence {x[n]} by samples of its spectrum x()
 Discrete Fourier Transform (DFT)
• DFT is a powerful computational tool for performing frequency analysis of
discrete-time signals.
• Suppose that aperiodic discrete-time signal x[n];
X() = . . . . . . . . . .equation(**)
ASTU | ECE 3205 |A.Y 2022 - 2023
Discrete Fourier Transform (DFT)
• And suppose that we sample X() periodically in frequency at spacing of ∆ω
radians between successive samples.
• Since X() is periodic with period 2π, only samples in the fundamental frequency
range are necessary
 We take N equidistant samples in 0 2π with spacing ∆ω =.
 N is the number of samples in the frequency domain
• If we evaluate equation (**) at ω =
X() = ,k=0, 1, . . . ., N-1

ASTU | ECE 3205 |A.Y 2022 - 2023

Discrete Fourier Transform
• Consequently, the frequency samples X() k=0, 1, . . . ., N-1, uniquely represent the
finite duration x[n].
• Hence, a finite-duration sequence x[n] of length L(i.e., x[n] =0 for n and
n) has a discrete-time Fourier transform:
X() =, 0 2π
where the upper and lower indices in the summation reflect the fact that x[n] = 0
outside the range 0 L-1
• When we sample X() at equally spaced frequencies , k=0,1,2,...,N, where
N≥L, the resultant samples are :

ASTU | ECE 3205 |A.Y 2022 - 2023

Discrete Fourier Transform
X[k] = X() =
= where k = 0,1,2, . . . ,N-1

• The upper index in the sum has been increased from L-1 to N-1 for convenience
since x[n] = 0 for n.
• In turn, the formula to recover the sequence x[n] from the frequency samples.
x[n] =, n = 0,1, . . .,N-1

 Inverse DFT (IDFT)

ASTU | ECE 3205 |A.Y 2022 - 2023

Discrete Fourier Transform
Example: A finite-duration sequence of length L is given as
x[n] = 1, 0 L-1
0, otherwise
Determine the N-point DFT of this sequence for N
Solution: The DTFT of this sequence is
X() =
= = =
• The N-point DFT of x[n] is simply X() evaluated at the set of N
equally spaced frequencies = , k=0, 1, . . ., N-1

ASTU | ECE 3205 |A.Y 2022 - 2023

Discrete Fourier Transform

X[k] = , k=0, 1, . . ., N-1

=
• If N is selected such that N =L, then the DFT becomes:
X[k] = L, k =0
0, k=1,2, . . .,L-1
There is only one non zero value in the DFT.
Exercise: compute the DFT of the four-point sequence x[n] = {0, 1, 2, 3}

ASTU | ECE 3205 |A.Y 2022 - 2023

Properties of DFT
1. Symmetry properties of the DFT

N-Point Sequence x[n] ,0 N- Point DFT

x[n] X[k]
x*[n] X*[N-k]
x*[N-n] X*[k]
[n] [n] =[X[k] + X*(N-k)]

j[n] [n] =[X[k] - X*(N-k)]

[n] =[x[n] + x*(N-n)] [n]

[n] =[x[n] - x*(N-n)] *[n]

ASTU | ECE 3205 |A.Y 2022 - 2023
Properties of DFT

• Real signals
Any real signal x[n] X[k] = X*[N-k]
[k] = [N-k]
[k] = -[N-k]

ASTU | ECE 3205 |A.Y 2022 - 2023

2 Properties of DFT

Property Time-Domain Frequency Domain

Notation x[n], y[n] X[k], Y[k]

Periodicity x[n+N] X[k] = X[k+N]
Time reversal x[N-n] X[N-k]
Circular Time shift X[k]

Circular frequency shift x[n]

ASTU | ECE 3205 |A.Y 2022 - 2023

Continued

Complex conjugate x*[n] X*[N-k]

Circular convolution
N
Multiplication of two
sequences N
Parserval’s theorem

ASTU | ECE 3205 |A.Y 2022 - 2023

 The z-Transform
• The z-transform for discrete-time signals is the counter part of the Laplace
transform for continuous-time signals.
• The discrete-time Fourier transform of a sequence x[n]:
X() = . . . . . . . .(*)
• The z-transform of a sequence of a sequence x[n] is defined as:
X(z) =
This equation is, in general, an infinite sum with z being a complex variable.
• z-transform operator Z{ . }, defined as:
z{x[n]} = = X(z) . . . . . . . . . . . . .(**)

ASTU | ECE 3205 |A.Y 2022 - 2023

 z-Transform
• Transform the sequence x[n] into the function X(Z), where z is a continuous
complex variable.
• Equation (**) is referred to as the two-sided for bilateral z-transform
X(z) =
 is one-sided or unilateral z-transform

• The bilateral and unilateral z-transforms are equivalent only if x[n] = 0 for n
• The relationship between the z-transform and DTFT is that if we replace the
complex variable z with the complex variable , then the z-transform reduces to
the DTFT.
• We can express the complex variable z in polar form:
z=r
ASTU | ECE 3205 |A.Y 2022 - 2023
 z-Transform
• Equation (**) becomes:
X(r) =
= . . . . . . .(***)
(***) can be interpreted as the DTFT of the product of the original sequence x[n]
and the exponential sequence .

Fig: the unit circle in the complex Z-plane

ASTU | ECE 3205 |A.Y 2022 - 2023

 z-Transform
• In the z-plane, the contour corresponding to |Z| = 1 is a circle of unit radius.
• Since the z-transform is an infinite power series, it only exists for those values of z
for which the series converges.
• The region of convergence (ROC) of X(z) is the set of values of z for which X(z)
attains a finite value.

 Condition for convergence of z-transform.

• It’s possible for z-transform to converge even if the DTFT doesn’t, since the
multiplication of the sequence by the real-exponential .
Example 1: Determine the z-transform of the following finite-duration sequences

ASTU | ECE 3205 |A.Y 2022 - 2023

 z-Transform
i. x1[n] = {1,2,5,7,0,1} ii . x2[n] = {1,2,5,7,0,1}

Solutions: X1(z) = 1+2+5+7+ ;ROC: entire z-plane except z=0

X2(z) = + 2+5+7+ ;ROC: entire z-pane except z=0&z=∞
• From these examples it is easily seen that the ROC of a finite-duration signal is
the entire z-plane, except possibly the point z=0 and/or z=∞.
Example 2: Determine the z-transform of the signal x[n] =(u[n]
Solution: The signal x[n] consists of an infinite number of nonzero samples.
X(z)= =
From geometric series, we recall that: 1+a+++ … = ; if |a|<1

ASTU | ECE 3205 |A.Y 2022 - 2023

 z-Transforms
• Consequently; for || < 1 or |z|> ;
X(z) = ; ROC |z|>
• x[n] = (u[n] is not absolutely summable hence DTFT doesn’t converge. But z-
transform exists if |z|>
• The ROC consists of all values of z such that the below inequality holds:

• If ROC doesn’t include the unit circle, the DTFT does not converge absolutely.
• The z-transform is useful when the infinite sum can be expressed in closed form;
i.e., X(z) is a rational function.

ASTU | ECE 3205 |A.Y 2022 - 2023

 z-Transform
X(z) =
where P(z) and Q(z) are polynomials of z.
• The values of z for which X(z) = 0 are called zeros of X(z) and the values of z for
which X(z) is infinite is referred to as the poles of X(z).
 The poles of X(z) for finite values of z are roots of the denominator polynomial.
 For rational z-transform, a number of relationship exist between the location of
poles of X(z) and the ROC of the z-transform.
 We can represent X(z) by a pole-zero plot(patterns) in the complex-plane, which
shows the location of poles by crossed (x) and the location of zeros by circles (o).
 By definition, the ROC of a z-transform should not contain any poles

ASTU | ECE 3205 |A.Y 2022 - 2023

 z-Transform
Example 3:
1. x[n] = u[n], a right-sided exponential sequence
X(z) = =
 For convergence of X(z), we require that

X(z) = = = , |z|
• The ROC is the range of values for which, |z|
• For a =1, x[n] is the unit step sequence and X(z) = , |z|

ASTU | ECE 3205 |A.Y 2022 - 2023

 z-Transform
• For this example, there is one zero, at z =0,and one pole, at z=a.

Fig: Pole-zero plot and region of convergence

ASTU | ECE 3205 |A.Y 2022 - 2023
 z-Transform
Example 4:
Left sided exponential sequence let x[n]= u[-n-1], the sequence is non-zero only for
n
X(z) = -
=-= 1
= -=1-
If || or |z| |a|
X(z) = 1- = = = , |z| |a|

ASTU | ECE 3205 |A.Y 2022 - 2023

 z-Transform

Fig: pole-zero plot and region of convergence of the left-sided exponential

ASTU | ECE 3205 |A.Y 2022 - 2023
 z-Transform
• Note that the causal signal u[n] and the anticausal signal u[-n-1] have identical
closed-form expressions for the z-transform, that is;
Z{u[n]}=Z{u[-n-1]} =
• This implies that a closed-form expression for the z-transform doesn’t uniquely
specify the signal in the time-domain. The ambiguity can be resolved only if in
addition to the closed –form z-transform expression, the ROC is specified.
• In summary, a discrete-time signal x[n] is uniquely determined by its z-transform
X(z) and the ROC of X(z).
• Another point is that the ROC of a causal signal is the exterior of a circle of some
radius r2 while the ROC of an anticausal signal is the interior of a circle of radius
r1. Generally, the ROC of a signal depends on both the duration (finite or infinite)
and on whether it’s causal, anticausal or two-sided.
ASTU | ECE 3205 |A.Y 2022 - 2023
 z-Transform
Example 5:
x[n] = u[n] + u[n]
Sum of two exponential sequences
X(z) =
= +
= + = +
= || 1 and 1 or |z| and |z|

ASTU | ECE 3205 |A.Y 2022 - 2023

 z-Transforms
• Thus, the ROC is the region overlap, |z|

• Example 5 shows that if there is a ROC for an infinite duration ,two-

sided signal, it’s a ring (annular region) in the z-plane.
ASTU | ECE 3205 |A.Y 2022 - 2023
ASTU | ECE 3205 |A.Y 2022 - 2023
Some Common z-Transforms

ASTU | ECE 3205 |A.Y 2022 - 2023

Some Common z-transforms

ASTU | ECE 3205 |A.Y 2022 - 2023

Properties of the ROC for the Z-transform
Property 1: ROC is a ring or disk in the z-plane centered at the origin.
Property 2: The DTFT of x[n] converges absolutely if and only if the ROC
of the z-transform of x[n] includes the unit circle.
Property 3: The ROC cant contain any poles
Property 4: If x[n] is a finite-duration sequence, then the ROC is the entire
z-plane, except possibly z =0 or z =
Property 5: If x[n] is right-sided sequence, the ROC extends outward from
the outermost (largest magnitude) finite pole in X(z) to z = .
Property 6: If x[n] is left-sided sequence, the ROC extends inwards from the
innermost (smallest magnitude) nonzero pole in X(z) to possibly including z
= 0.
ASTU | ECE 3205 |A.Y 2022 - 2023
Properties of The ROC for the z-Transform
Property 7: If x[n] two-sided sequence; the ROC will consist of a ring in the z-
plane, bounded on the interior and exterior by the pole.
Property 8: The ROC must be a connected region stability, causality and the ROC.

ASTU | ECE 3205 |A.Y 2022 - 2023

 z-Transform Properties
• In this section, X(z) denotes the z-transform of x[n] and the ROC of X(z) is
indicated by ;i.e.,
x[n] X(z), ROC =
I. Linearity:
The linearity property states that
a + b a + b∩
II. Time shifting:
According to time shifting property,
x[n-] X(z), ROC = (except for the possible addition or deletion of z =0 or z
= ∞).

ASTU | ECE 3205 |A.Y 2022 - 2023

 z-Transform Properties
III. Multiplication by an exponential sequence
The exponential multiplication property is expressed mathematically as
x[n] X(), ROC = ||
IV. Differentiation of X(z)
The differentiation property states that
nx[n] -z, ROC =
V. Conjugation of a complex sequence
The conjugation property is expressed as
x*[n] x*(z*), ROC =

ASTU | ECE 3205 |A.Y 2022 - 2023

 z-Transform Properties
VI. Time reversal
x*[-n] X*(), ROC =
x[-n] X(), ROC =
VII.Convolution of sequences
,∩
Exercise: Determine the z-transform the below signals
a. x[n] = (cosωon)u[n]
b. X[n] = (sin ωon)u[n]

ASTU | ECE 3205 |A.Y 2022 - 2023

Causality and Stability
• If an LTI system is causal (with a right sided impulse response function h[n]=0 for
n<0), then the ROC of its z-transform X(z) is the exterior of a circle including
infinity.
• An LTI system is stable if and only if its impulse response is absolutely summable,
i.e., the frequency response function arg(H()) exits, i.e. the ROC of its z-transform
X(z) includes the unit circle |z|=1
• Therefore, a LTI system is BIBO stable if and only if ROC of the system function
includes the unit circle.
• A causal and stable system must have a system function that converges for |z|>r<1.
• Since the ROC can’t contain any poles of H(z), it follows that a causal and stable LTI
system is BIBO stable if and only if all the poles of H(z) are inside the unit circle

ASTU | ECE 3205 |A.Y 2022 - 2023

Causality and Stability
Example: A LTI system is characterized by the system function

ASTU | ECE 3205 |A.Y 2022 - 2023

The Inverse Z-Transform
A. Inspection Method
 Simply becoming familiar with, or recognizing “by inspection” of certain
transform pairs.
 Tables of Z-transforms are invaluable in applying the inspection method.
B. Partial fraction Expansion
 Obtain an alternative expression for X(z) as a sum of simpler terms, each of
which can be recognized and employ table lookup for each terms.
 In our treatment we deal with signals and systems in z-domain which have a
rational z-transform (i.e., a z-transforms that are a ratio of two polynomials).

ASTU | ECE 3205 |A.Y 2022 - 2023