Prepared by Dr Harish M S

Digital Signal Processing

Semester 5, Course code BEC502, Credits 4, Total, marks 100(50+50)

Text Books:
1. Proakis & Manolakis, “Digital Signal Processing - Principles Algorithms & Applications”, 4th
Edition, Pearson education, New Delhi, 2007. ISBN: 81-317-1000-9.
2. Li Tan, Jean Jiang, “Digital Signal processing - Fundamentals and Applications”, Academic
Press, 2013, ISBN: 978- 0-12-415893.
3. Vinay K. Ingle, John G Proakis, “Digital Signal Processing Using MATLAB, A problem
Solving Companion”, Cengage Learning, 2018, ISBN: 93-86668-11-4

Module-1

Introduction

Most of the signals encountered in science and engineering are analog in nature (continuous time
signal). These analog signals are processed directly using appropriate analog processor. Digital
signal processing is an alternate method for processing the analog signal. Digital signal
processing is an area of science and engineering that has developed rapidly over the past 40
years. This rapid development is due to significant advances in digital computer technology and
integrated circuit fabrication. Because of the rapid development in the integrated-circuit
technology starting from MSI, LSI and now VLSI circuit leads to the development of powerful,
cheaper, smaller and faster digital computer and special purpose digital hardware. These highly
sophisticated digital systems capable of performing complex digital signal processing functions
and tasks, which are usually too difficult, performed by analog signal processor.

The digital signal processors are cheaper, more reliable and programmable using software
program. Through software program one can modify the signal processing functions. Thus, using
hardware and software higher order of precision achievable compared to the analog signal
processor.

Processing of analog signals by digital means has some drawbacks

1. We know that, analog signal to digital signal conversion consists of sampling and
quantization. During this process distortion prevents us to reconstructing the original
 Prepared by Dr Harish M S

analog signals from the quantized samples. This distortion can be avoided by proper
choice of sampling rate and precession in the quantization process.
2. There are finite precision effects that must be considered in the digital signal processing
of the quantized samples.

Signals, Systems and Signal Processing

Signals

The signal is defined as any physical quantity that varies with time, space, or any other
independent variable or variables. Mathematically signal is a function of one or more
independent variables. For example

𝑠1 (𝑡) = 5𝑡
𝑠2 (𝑡) = 20𝑡 2

The function 𝑠1 (𝑡) and 𝑠2 (𝑡) describe two signals, 𝑠1 (𝑡) varies linearly with time t, and 𝑠2 (𝑡)
that varies quadratically with time t. Let us consider another example 𝑠(𝑥, 𝑦) = 3𝑥 + 2𝑥𝑦 +
10𝑦 2. This function describes a signal of two independent variables x and y in the two
dimensional space. These signals are depends on the independent variables and can be
represented by mathematical expressions.

Some signals are there, they cannot be represented by mathematical expressions. For example
speech signal cannot be represented by mathematical expressions. But speech signal may be
represented by sum of the segment of speech of several sinusoids of different amplitudes and
frequencies, that is, as

∑ 𝐴𝑖 (𝑡)sin[2𝜋𝐹𝑖 (𝑡)𝑡 + 𝜃𝑖 (𝑡)]

𝑖=1

Where {𝐴𝑖 (𝑡), 𝐹𝑖 (𝑡)} and {𝜃𝑖 (𝑡)} are the sets of time varying amplitudes, frequencies and phases
respectively of sinusoids. The amplitude, phase and frequency of the any short time segment
convey the information in that short time segment of the speech signal. Therefore this is an
information bearing signal. In another example, an electrocardiogram (ECG) signal provides
information about the condition of the patient’s heart. Similarly an electroencephalogram (EEG)
signal provides information about the activity of brain. Speech, ECG and EEG signals are the
 Prepared by Dr Harish M S

example of information bearing signals. These signals are the varying with respect independent
variable time called one dimensional signal.

If the signal function depends on two or more independent variable, then the signal is known as
multidimensional signal. For example, image signal 𝐼(𝑥, 𝑦) is function of two independent
variable x and y in the two dimensional space. And the block and white TV picture signal
𝐼(𝑥, 𝑦, 𝑡)is three dimensional signal. Because, its brightness intensity is function of three
independent variable x, y and t. Here x and y represents the points in the two dimensional space
and t indicating time.
Systems
The system is defined as a physical device that performs the desired operation on the input signal
and produces a desired output signal. The block diagram of the system is as shown in figure 1.1.
For example, filter is a system; it takes the corrupt signal with noise and interference as an input
signal and performs the filter operation on the input signal and reducing the noise and
interference and producing the filtered output.

Signal Processing
If the signal is applied as an input to the system, at the output of the system we have a processed
signal. If the system is filter, then processing of the signal involves filtering the noise and
interference from the input signal and produces filtered signal. In general, system is
characterized by type of operation on the signal. Such an operation usually referred to as signal
processing. Thus signal processing is defined as changing the characteristic of the signal or
extracting the desired information from the signal.
Basic elements of a digital signal processing
Most of the signals encountered in science and engineering are analog in nature (continuous time
signal). In analog signals processing, analog signal may be processed directly by appropriate
analog processor (such as filter, amplifier, frequency analyzer, frequency multiplier etc) to
 Prepared by Dr Harish M S

change the characteristics of a signal or extract the information from the signal. Analog signal
processing is as shown in below Figure 1.2. Both the input signal and the output signal are in
analog form.

Digital signal processing provides an alternate method for processing the analog signal. The
basic elements of the digital signal processing as shown in Figure1.32. To perform the processing
digitally, there is a need for an interface between analog signal and the digital processor. The
interface is called ADC. The digital signal processor (DSP) may be a large programmable digital
computer or a small microprocessor programmed to perform the desired operations on the input
signal. The DSPs are developed for verity of applications such as hard disc controller, cellular
phones, speech processing, and image processing and so on. DSPs are from companies such as
analog devices, Motorola, Texas instruments etc. DSPs may be fixed point DSP or floating point
DSP, consists of hardware like Flip-flaps and logic gates etc. In applications, where the digital
output from the DSP is to be given to the user in an analog from, such as speech signal, we must
convert the digital signal into the analog signal. Such operation is performed by DAC.

Advantages of digital over analog signal processing

Following are the advantages of digital signal processing over analog signal processing.

Sl.No Digital signal processing Analog signal processing

1 Digital signal processor is a programmable Analog signal processor is not a
device. It can be reconfigured for any programmable device. Analog signal
 Prepared by Dr Harish M S

other processing operation by changing processor is designed specific processing

the software program. operation. It can be reconfigured any other
processing operation by changing the
hardware.

2 Digital signal processing provides much Analog processing having less tolerance to
better control of accuracy requirements, control the accuracy, thus have limited
thus accuracy is more. accuracy.
3 Digital signals can be easily stored on Stored analog signals deteriorate rapidly as
magnetic media. As a result, signals signal time progress and cannot be retrieve the
become transportable and can be original signal.
processed off-line in a remote laboratory
4 The lower cost may be due to the fact that Cost of the analog signal processing is
the digital hardware is cheaper. more
5 The digital signal processor has the limited Band width is large.
range of band width. The wider bandwidth
requires fast sampling rate, A/D converter
and fast digital signal processor.

Classification of Signals
Signals are classified into
1. Multichannel and multidimensional signals
2. Continuous-time and discrete-time signals
3. Continuous-valued and discrete-valued signals
4. Deterministic and random signals

1. Multichannel and multidimensional signals

We know that the signal is a function of one or more independent variables. The signal may be
real valued or complex valued or perhaps a vector. For example

𝑠1 (𝑡) = 𝐴𝑠𝑖𝑛3𝜋𝑡, this signal is a real valued signal

 Prepared by Dr Harish M S

𝑠2 (𝑡) = 𝐴𝑒 𝑗3𝜋𝑡 = 𝐴𝑐𝑜𝑠3𝜋𝑡 + 𝑗𝐴𝑠𝑖𝑛3𝜋𝑡 , is a complex valued signal.

In some applications, signals are generated from the multiple sources or multiple sensors. Such
signals are represented in vector form. Such signals we referred to as multichannel signal. In
electrocardiography (ECG), for example, 3-lead and 12-lead ECG are often used in practice,
which results in 3-channel and 12-channel signals. Mathematically the 3-channel signal vector
𝑠1 (𝑡)
represented as 𝑆3 (𝑡) = [𝑠2 (𝑡)]
𝑠3 (𝑡)

We know that image is two dimensional signal and block and white TV picture signal is a three
dimensional signal. It is possible to have multichannel and multidimensional signal
simultaneously. For example color TV picture is described by three intensity functions of the
form 𝐼𝑟 (𝑥, 𝑦, 𝑡), 𝐼𝑔 (𝑥, 𝑦, 𝑡) and 𝐼𝑏 (𝑥, 𝑦, 𝑡), corresponding to the brightness of the three principle
colors (red, green, blue) as a function of time. Hence the color TV picture signal is a 3-channel,
3-dimensional signal, which can be represented by the vector

𝐼𝑟 (𝑥, 𝑦, 𝑡)
𝐼 (𝑥, 𝑦 , 𝑡) = [𝐼𝑔 (𝑥, 𝑦, 𝑡) ]
𝐼𝑏 (𝑥, 𝑦, 𝑡)

Thus Speech is a one dimensional signal, image is a two dimensional signal, block and white
picture signal is three dimensional signal, ECG signal is a 3-channel or 12-channel signal and
color TV picture is a 3-channel, 3-dimensional signal.

2. Continuous-time and discrete-time signals

In continuous-time signal, time is continuous and amplitude is continuous, continuous time
signal is also called as analog signal. The magnitude of the signal is defined for all values of
independent variable t in the interval (a, b). Where a can be −∞ and 𝑏 can be ∞. For example
a. 𝑥1 (𝑡) = 𝑐𝑜𝑠𝜋𝑡 − ∞ ≤ 𝑡 ≤ ∞
b. 𝑥2 (𝑡) = 𝑒 |𝑡| −∞ ≤𝑡 ≤∞
 Prepared by Dr Harish M S

Similarly, in discrete-time signal, time is discrited and amplitude is continuous. The magnitude
of the discrete time signal is defined for discrete instant. All discrete time signals are derived
from continuous time signal by uniform sampling i.e., the discrete time signal 𝑥(𝑛) =
𝑥(𝑡)|𝑡=𝑛𝑇 , n is the time index; it takes the discrete values like 𝑛 = 0, ±1, ±2, …. etc and 𝑇 is the
sampling interval or sampling period in seconds. Example
0.8𝑛 𝑖𝑓 𝑛 > 0
a. 𝑥(𝑛) = {
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

3. Continuous-valued and discrete-valued signals

The values of the continuous time or discrete time signal can be continuous or discrete. If a
signal takes on all possible values on a finite or an infinite range, it is said to be a continuous-
valued signal. If the signal takes on values from a finite set of possible values, it is said to be a
discrete-valued signal. It is also called digital signal. In digital signal, time is discritized as well
as amplitude is also discrited. Figure 1.6 shows a digital signal that takes on one of four
possible values. In order to process the signal digitally, it must be discrete in time and discrete in
amplitude (i.e., it must be a digital signal). Thus to obtain the digital signal from the analog
signal, first analog signal is sampled by uniform sampling as result we a have discrete-time
signal or continued-valued signal. This continuous-valued signal is converted into discrete-
valued signal called quantization.
 Prepared by Dr Harish M S

4. Deterministic and random signals

For the mathematical analysis and processing of signal requires the mathematical descriptions.
This mathematical description is also known as signal modeling. If the signal is represented by
mathematical representation or mathematical expression, then the signal is known as
deterministic signal. The deterministic signal determines the past, present and feature values of
the signal precisely without any uncertainty. Therefore, in deterministic signal there is no
uncertainty with respect to its value at time and the signal is anticipatory. Example
𝑥1 (𝑡) = 𝑐𝑜𝑠𝛺𝑡
𝑥2 (𝑛) = sin(0.1𝜋𝑛)
In many real time applications, signals cannot be expressed by mathematical expressions are
known as random signals. Therefore, in random signals, there is an uncertainty with respect to
value before its actual occurrence. The random signals cannot be completely represented by
mathematical expression. The mathematical analysis of random is provided by theory of
probability and stochastic process. Example, Noise signal, speech signal, seismic signal etc.
The Concept of Frequency in Continuous Time and Discrete Time Sinusoidal Signals
The concept of frequency is a very familiar word for Science and engineers. This frequency is
basic in designing of radio receiver, a high fidelity system or a spectral filter for color
photography. We know that, frequency is closely related to periodic motion called harmonic
motion or harmonic oscillations. The periodic motion is represented by sinusoidal functions. The
frequency is closely related to time, frequency has the dimension of inverse time.
Continuous time sinusoidal signals
A simple harmonic oscillation is mathematically described by the following continuous time
sinusoidal signal:
𝑥𝑎 (𝑡) = 𝐴𝑐𝑜𝑠(𝛺𝑡 + 𝜃) − ∞ ≤ 𝑡 ≤ ∞, as shown in figure 1.7.
The signal 𝑥𝑎 (𝑡) is an analog signal. This signal characterized by three parameter 𝐴, 𝛺 and 𝜃.
Where 𝐴 the amplitude of the sinusoid, Ω is the frequency in radians per second (rad/sec),
and 𝜃 is the phase angle in radian. Instead of 𝛺, we often use the frequency F in cycles per
second or hertz, thus 𝛺 = 2𝜋𝐹. Therefore
𝑥𝑎 (𝑡) = 𝐴𝑐𝑜𝑠(2𝜋𝐹𝑡 + 𝜃) −∞ ≤𝑡 ≤∞
 Prepared by Dr Harish M S

The analog signal has the following properties

1. The analog signal 𝑥𝑎 (𝑡) is a periodic signal, it will satisfy the condition𝑥𝑎 (𝑡) =
𝑥𝑎 (𝑡 + 𝑇𝑝 ). Thus 𝑥𝑎 (𝑡) is a periodic signal with fundamental period of sinusoidal 𝑇𝑝 =
1/𝐹.
2. Continuous time sinusoidal signals with distinct frequencies are themselves distinct.
3. As the signal frequency F increases, results in an increase in rate of oscillations of the
signal
Using Euler’s identity 𝑒 𝑗𝜃 = 𝑐𝑜𝑠𝜃 + 𝑗𝑠𝑖𝑛𝜃, then
𝐴 𝑗(𝛺𝑡+𝜃) 𝐴 −(𝛺𝑡+𝜃)
𝑥𝑎 (𝑡) = 𝐴𝑐𝑜𝑠( 𝛺𝑡 + 𝜃) = 𝑒 + 𝑒
2 2
Therefore, the sinusoidal signal 𝑥𝑎 (𝑡) is obtained by adding two equal amplitude complex
conjugate exponential signals, sometimes called phasor as shown in figure 1.8. As the time
progress, the phasor rotates in the opposite directions with angular frequency ±𝛺 𝑟𝑎𝑑/𝑠𝑒𝑐. The
𝐴 𝐴
complex exponential 𝑒 𝑗(𝛺𝑡+𝜃) has amplitude with positive frequency 𝛺 𝑟𝑎𝑑/𝑠𝑒𝑐 rotates in
2 2
𝐴 𝐴
anticlockwise direction. And complex exponential 𝑒 −𝑗(𝛺𝑡+𝜃) has amplitude with negative
2 2

frequency 𝛺 𝑟𝑎𝑑/𝑠𝑒𝑐 rotates in clockwise directions.

For the mathematical convenience, we use both negative and positive frequencies. Hence the
range of analog sinusoids is −∞ ≤ 𝐹 ≤ ∞.
 Prepared by Dr Harish M S

Discrete-time sinusoidal signals

A discrete-time sinusoidal may be expressed as
𝑥(𝑛) = 𝐴𝑐𝑜𝑠(𝜔𝑛 + 𝜃) −∞≤𝑛≤∞
Where n is an integer variable, called sample number, A is the amplitude of the sinusoid, 𝜔 is
digital frequency in radian per sample. Since sample is the unit less quantity, thus the digital
frequency ω is in radians. And 𝜃 is the phase in radians.
If instead of digital frequency 𝜔 we the use the digital frequency f defined by
𝜔 ≡ 2𝜋𝑓
Therefore
𝑥(𝑛) = 𝐴𝑐𝑜𝑠(𝜔𝑛 + 𝜃) = 𝐴𝑐𝑜𝑠(2𝜋𝑓𝑛 + 𝜃) −∞≤𝑛 ≤∞
The frequency 𝑓 has dimension of cycles per sample. Since sample is the unit less quantity, thus
the digital frequency 𝑓 is in cycles.
𝜋 𝜋
Let us consider 𝑥 (𝑛) = 𝐴𝑐𝑜𝑠 ( 6 𝑛 + 3 ). Figure 1.9 shows a discrete-time sinusoidal with
𝜋 𝜋
frequency 𝜔 = 6 𝑟𝑎𝑑/𝑠𝑎𝑚𝑝𝑙𝑒, phase 𝜃 = 3 𝑟𝑎𝑑 .
𝜔 𝜋 1
𝜔 = 2𝜋𝑓, 𝑓 = 2𝜋 = 6×2𝜋 = 12 𝑐𝑦𝑐𝑙𝑒/𝑠𝑎𝑚𝑝𝑙𝑒.

Let us calculate the 𝑥(𝑛) for various value of 𝑛

𝑥(𝑛) 𝜋 𝜋 𝑥(𝑛) 𝜋 𝜋
𝑥(𝑛) = 𝐴𝑐𝑜𝑠 ( 6 𝑛 + 3 ). 𝑥(𝑛) = 𝐴𝑐𝑜𝑠 ( 6 𝑛 + 3 ).

𝑥(−6) −0.500𝐴 𝑥(6) −0.500𝐴

𝑥(−5) 0.000𝐴 𝑥(7) 0. 000𝐴
𝑥(−4) 0.500𝐴 𝑥(8) 0.500𝐴
𝑥(−3) 0.866𝐴 𝑥(9) 0.866𝐴
𝑥(−2) 1.000𝐴 𝑥(10) 1.000𝐴
𝑥(−1) 0.866𝐴 𝑥(11) 0.866𝐴
𝑥(0) 0.500𝐴 𝑥(12) 0.500𝐴
𝑥(1) 0.00𝐴 𝑥(13) 0.000𝐴
𝑥(2) −0.500𝐴 𝑥(14) −0.500𝐴
𝑥(3) −0.866𝐴 𝑥(15)
𝑥(4) −1.000𝐴
𝑥(5) −0.866𝐴
 Prepared by Dr Harish M S

Similar to the continuous-time sinusoidal, the discrete-time sinusoids are having the following
characteristics.
2) If discrete time signal is periodic only if its frequency f is a rational number. We know
that if 𝑥(𝑛) is periodic, it has to satisfy
𝑥 (𝑛 ) = 𝑥 (𝑛 + 𝑁 ) 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑛 𝑜𝑟 −∞≤𝑛≤∞ .
The smallest value of 𝑁 is called fundamental period of 𝑥(𝑛). Let us consider the signal
𝑥(𝑛) = 𝐴𝑐𝑜𝑠(𝜔𝑛 + 𝜃) = 𝐴𝑐𝑜𝑠(2𝜋𝑓𝑛 + 𝜃) −∞≤𝑛 ≤∞
If 𝑥(𝑛) is periodic, then 𝑥(𝑛) = 𝑥(𝑛 + 𝑁)
𝐴𝑐𝑜𝑠(𝜔𝑛 + 𝜃) = 𝐴𝑐𝑜𝑠(𝜔[𝑛 + 𝑁] + 𝜃)
𝐴𝑐𝑜𝑠(𝜔𝑛 + 𝜃) = 𝐴𝑐𝑜𝑠(𝜔𝑛 + 𝜔𝑁] + 𝜃)
This relation is true, if 𝜔𝑁 is integer multiple of 2𝜋.i.e 𝜔𝑁 = 2𝜋𝑘 or
2𝜋𝑓𝑁 = 2𝜋𝑘
𝑘
𝑓=
𝑁
For periodic sinusoids, the frequency 𝑓 should be a rational number and it is a ratio of two
integers. If the frequency is rational number, then integer 𝑘 and 𝑁 are relatively prime number.
Observe that a small change in frequency can result large change in period N. For example 𝑓1 =
𝑘 31 𝑘 30
= 60 implies that fundamental period 𝑁1 = 60 samples in one period, whereas 𝑓2 = 𝑁 = 60 =
𝑁1 2

1
result in 𝑁2 = 2 samples in one period.
2

3) Discrete-time sinusoids whose frequencies are separated by an integer multiple of 2𝜋 are

identical.
To prove this assertion, let us consider the sinusoidal signal 𝑥1 (𝑛) = 𝑐𝑜𝑠(𝜔𝑛 + 𝜃)
 Prepared by Dr Harish M S

Let 𝑥2 (𝑛) = 𝑐𝑜𝑠([𝜔 + 2𝜋]𝑛 + 𝜃) = 𝑐𝑜𝑠( 𝜔𝑛 + 2𝜋𝑛 + 𝜃) = 𝑐𝑜𝑠(𝜔𝑛 + 𝜃) = 𝑥1 (𝑛),

In general 𝑥𝑘 (𝑛) = 𝑐𝑜𝑠(𝜔𝑘 𝑛 + 𝜃), 𝑘 = 0, 1, 2 … ..
Where
𝜔𝑘 = 𝜔 + 2𝜋𝑘 −𝜋 ≤𝜔≤𝜋
1
Any sequence resulting from a sinusoid with a frequency |𝜔| > 𝜋, or |𝑓| > 2, is identical to a
1
sequence obtained from a sinusoidal signal with frequency |𝜔| < 𝜋, or |𝑓| < 2. Because of this
1
similarity, we call the sinusoid having the frequency|𝜔| > 𝜋, or |𝑓| > 2 an alias of a
1
corresponding sinusoid with frequency|𝜔| < 𝜋, or |𝑓| < 2. Thus we regard frequencies in the
1 1 1
range −𝜋 ≤ 𝜔 ≤ 𝜋 or − ≤ 𝑓 ≤ , as unique and all frequencies|𝜔| > 𝜋, or|𝑓| > , as aliases.
2 2 2
1 1
Thus, discrete time sinusoidal signal frequencies range is −𝜋 ≤ 𝜔 ≤ 𝜋 𝑜𝑟 − 2 ≤ 𝑓 ≤ 2 and

continuous time sinusoidal signal frequencies range is −∞ ≤ 𝛺 ≤ ∞ 𝑜𝑟 − ∞ ≤ 𝐹 ≤ ∞.

𝜋 𝜋 𝜋 𝜋
Problem: Sketch the signal 𝑥 (𝑛) = cos(𝜔𝑛), if 𝜔 = 0, 𝜔 = , 𝜔= , 𝜔= and 𝜔 = 2 .
8 4 2

Solution: 𝑥(𝑛) = cos(𝜔𝑛)

1. When angular frequency 𝜔 = 0, then frequency𝑓 = 0, 𝑁 = ∞ signal is a DC signal
𝑥1 (𝑛) = 𝑐𝑜𝑠𝜔𝑛 = 1 −∞≤𝑛≤∞
𝜋 1
2. When angular frequency 𝜔 = 8 , then frequency 𝑓 = 16 , 𝑁 = 16 𝑠𝑎𝑚𝑝𝑙𝑒 in one period
𝜋
𝑥2 (𝑛) = 𝑐𝑜𝑠𝜔𝑛 = cos( 𝑛) −∞ ≤𝑛 ≤∞
8
𝜋 1
3. When angular frequency 𝜔 = 4 , then frequency 𝑓 = 8 , 𝑁 = 8 𝑠𝑎𝑚𝑝𝑙𝑒 in one period
𝜋
𝑥3 (𝑛) = 𝑐𝑜𝑠𝜔𝑛 = cos( 𝑛) −∞ ≤𝑛 ≤∞
4
𝜋 1
4. When frequency 𝜔 = 2 , then frequency 𝑓 = 4 , 𝑁 = 4 𝑠𝑎𝑚𝑝𝑙𝑒 in one period
𝜋
𝑥4 (𝑛) = 𝑐𝑜𝑠𝜔𝑛 = cos( 𝑛) −∞ ≤𝑛 ≤∞
2
1
5. When angular frequency 𝜔 = 𝜋, then frequency 𝑓 = , 𝑁 = 2𝑠𝑎𝑚𝑝𝑙𝑒𝑠 in one period
2

𝑥5 (𝑛) = 𝑐𝑜𝑠𝜔𝑛 = cos(𝜋𝑛) −∞≤𝑛≤∞

Let us calculate the signal value for various value of 𝑛
 Prepared by Dr Harish M S

n 𝑥1 (𝑛) 𝑥 2 (𝑛 ) 𝑥 3 (𝑛 ) 𝑥 4 (𝑛 ) 𝑥 5 (𝑛 )
𝑥(−15) 1.000 0.924 0.707 0.000 −1.000
𝑥(−14) 1.000 0.707 0.000 −1.000 1.000
𝑥(−13) 1.000 0.383 −0.707 0.000 −1.000
𝑥(−12) 1.000 0.000 −1.000 1.000 1.000
𝑥(−11) 1.000 −0.383 −0.707 0.000 −1.000
𝑥(−10) 1.000 −0.707 0.000 −1.000 1.000
𝑥(−9) 1.000 −0.924 0.707 0.000 −1.000
𝑥(−8) 1.000 −1.000 1.000 1.000 1.000
𝑥(−7) 1.000 −0.924 0.707 0.000 −1.000
𝑥(−6) 1.000 −0.707 0.000 −1.000 1.000
𝑥(−5) 1.000 −0.383 −0.707 0.000 −1.000
𝑥(−4) 1.000 0.000 −1.000 1.000 1.000
𝑥(−3) 1.000 0.383 −0.707 0.000 −1.000
𝑥(−2) 1.000 0.707 0.000 −1.000 1.000
𝑥(−1) 1.000 0.924 0.707 0.000 −1.000
𝑥(0) 1.000 1.000 1.000 1.000 1.000
𝑥(1) 1.000 0.924 0.707 0.000 −1.000
𝑥(2) 1.000 0.707 0.000 −1.000 1.000
𝑥(3) 1.000 0.383 −0.707 0.000 −1.000
𝑥(4) 1.000 0.000 −1.000 1.000 1.000
𝑥(5) 1.000 −0.383 −0.707 0.000 −1.000
𝑥(6) 1.000 −0.707 0.000 −1.000 1.000
𝑥(7) 1.000 −0.924 0.707 0.000 −1.000
𝑥(8) 1.000 −1.000 1.000 1.000 1.000
𝑥(9) 1.000 0.924 0.707 0.000 −1.000
𝑥(10) 1.000 −0.707 0.000 −1.000 1.000
𝑥(11) 1.000 −0.383 −0.707 0.000 −1.000
𝑥(12) 1.000 0.000 −1.000 1.000 1.000
𝑥(13) 1.000 0.383 −0.707 0.000 −1.000
 Prepared by Dr Harish M S

𝑥(14) 1.000 0.707 0.000 −1.000 1.000

𝑥(15) 1.000 0.924 0.707 0.000 −1.000

𝜋 𝜋 𝜋
Figure 1.10 shows a discrete-time sinusoidal with frequency 𝜔 = 0, 𝜔 = , 𝜔= ,𝜔 =
8 4 2

and 𝜔 = 𝜋 .

Discrete-Time Signals and Systems

 Prepared by Dr Harish M S

Discrete Time Signals

The graphical representation of a discrete-time signal 𝑥(𝑛), is a function of an independent
integer variable n as shown in figure 1.11. The independent variable n is also called as time index
of discrete-time signal. In discrete-time signal, signal is defined at discrete instant of time only
and the signal is not defined between the successive samples. The discrete-time signal 𝑥(𝑛)is
derived from the analog signal 𝑥𝑎 (𝑡) obtained from uniform sampling. Mathematically the
sampling processing is represented as
𝑥(𝑛) = 𝑥𝑎 (𝑡)|𝑡=𝑛𝑇
Where, T is the sampling period or sampling interval or distance between the successive
samples.

There are alternate method to represent the signal are

1, 𝑓𝑜𝑟 𝑛 = 1, 3
1. Functional representation 𝑥(𝑛) = { 4, 𝑓𝑜𝑟 𝑛 = 2
0, 𝑒𝑙𝑠𝑒𝑤ℎ𝑒𝑟𝑒
𝑛 ……. −2 −1 0 1 2 3 4 5…….
2. Tabular representation, such as |
𝑥(𝑛) ……. 0 0 0 1 4 1 0 0 ……..

3. Sequence representation
An infinite-duration signal with origin (𝑛 = 0) indicated by the symbol ↑ is represented
as
… … … … … … .0, 0, 1, 4, 1, 0, 0, … . .
𝑥 (𝑛 ) = ( )
↑
A sequence 𝑥(𝑛), which is zero for 𝑛 < 0, can be represented as
0, 1, 1, 4, 1,0, 0, … ..
𝑥 (𝑛 ) = ( )
↑
A finite duration sequence can be represented as
3, −1, −2, 5, 0, 4, −1
𝑥 (𝑛 ) = ( )
↑
A finite duration sequence 𝑥(𝑛) = 0 for 𝑛 < 0 can be represented as
 Prepared by Dr Harish M S

0, 1, 4, 1
𝑥 (𝑛 ) = ( )
↑
Some Elementary Discrete-Time Signals
In our study of discrete-time signals and systems, there are a number of basic signals that appear
often and play an important role. These signals are defined as
1. The unit sample sequence or unit impulse is denoted as 𝛿 (𝑛) and is defined as
1, 𝑓𝑜𝑟 𝑛 = 0
𝛿 (𝑛 ) = { , The graphical representation as shown in figure 1.12(a)
0, 𝑓𝑜𝑟 𝑛 ≠ 0
2. The unit step signal is denoted as 𝑢(𝑛) and is defined as
1, 𝑓𝑜𝑟 𝑛 ≥ 0
𝑢 (𝑛 ) = { , The graphical representation as shown in figure 1.12(b)
0, 𝑓𝑜𝑟 𝑛 < 0
3. The unit ramp signal is denoted as 𝑢𝑟 (𝑛) and is defined as
𝑛, 𝑓𝑜𝑟 𝑛 ≥ 0
𝑢𝑟 (𝑛 ) = { . The graphical representation as shown in figure 1.12(c)
0, 𝑓𝑜𝑟 𝑛 < 0

4) The exponential signal is defined as

𝑥(𝑛) = 𝑎𝑛 , for all n
If the parameter 𝑎 is real, then 𝑥(𝑛) is real exponential signal as shown in figure 1.13.

When the parameter 𝑎 is complex, it can be expressed as 𝑎 ≡ 𝑟𝑒 𝑗𝜃 , then 𝑥(𝑛) is complex

exponential, now 𝑟 and 𝜃 are the parameters. Hence 𝑥(𝑛) can be expressed as
 Prepared by Dr Harish M S

𝑥(𝑛) = (𝑟𝑒 𝑗𝜃 )𝑛 = 𝑟 𝑛 𝑒 𝑗𝜃𝑛 = 𝑟 𝑛 (𝑐𝑜𝑠𝜃𝑛 + 𝑗𝑠𝑖𝑛𝜃𝑛) = 𝑟 𝑛 𝑐𝑜𝑠𝜃𝑛 + 𝑟 𝑛 𝑗𝑠𝑖𝑛𝜃𝑛

Now real part of 𝑥(𝑛) is 𝑥𝑅 (𝑛) = 𝑟 𝑛 𝑐𝑜𝑠𝜃𝑛 and imaginary part of 𝑥(𝑛) is 𝑥𝐼 (𝑛) =
𝑟 𝑛 𝑠𝑖𝑛𝜃𝑛. Thus graphs of 𝑥(𝑛) has two plots, Figure 1.14 shows the graphs of 𝑥𝑅 (𝑛) and
𝜋
𝑥𝐼 (𝑛) for 𝑟 = 0.9 and 𝜃 = 10. We observed that both 𝑥𝑅 (𝑛) and 𝑥𝐼 (𝑛)decaying

exponentially
Alternatively, the complex exponential signal 𝑥(𝑛) can be expressed both in magnitude
and phase as
𝑥 (𝑛) = 𝑟 𝑛 𝑒 𝑗𝜃𝑛 = |𝑥(𝑛)|∠𝑥(𝑛) = 𝑟 𝑛 ∠𝜃𝑛
Magnitude |𝑥(𝑛)| = 𝐴(𝑛) = 𝑟 𝑛 and phase ∠𝑥(𝑛) = 𝜑(𝑛) = 𝜃𝑛
𝜋
Figure 1.15 illustrates magnitude and phase plot of 𝑥(𝑛) for 𝑟 = 0.9 and 𝜃 = 10.

Classification of Discrete-Time Signals

Based on the characteristic of the signal, signals are classified into
 Prepared by Dr Harish M S

1. Energy and power signals

2. Periodic and aperiodic signals
3. Symmetric(even) and antisymmetric (odd ) signals
Energy and power signals
The total energy E of the signal 𝑥(𝑛) is defined as
∞

𝐸 ≡ ∑ |𝑥(𝑛)|2
𝑛=−∞

The average power P of the signal 𝑥(𝑛) is defined as

𝑁
1
𝑃≡ ∑ |𝑥(𝑛)|2
lim 2𝑁 + 1
𝑁→∞ 𝑛=−𝑁

If energy is finite and power is zero, then the signal is known as energy signal. If energy is
infinite and power is finite then the signal is known as power signal. In brief
0 < 𝐸 < ∞ 𝑎𝑛𝑑 𝑃 = 0 𝑖𝑠 𝑡ℎ𝑒 𝑒𝑛𝑒𝑟𝑔𝑦 𝑠𝑖𝑔𝑛𝑎𝑙
0 < 𝑃 < ∞ 𝑎𝑛𝑑 𝐸 = ∞ 𝑖𝑠 𝑡ℎ𝑒 𝑝𝑜𝑤𝑒𝑟 𝑠𝑖𝑔𝑛𝑎𝑙
Formulas to be remembered
1−𝑎 𝑁
1. ∑𝑁−1 𝑛
𝑛=0 𝑎 = if 𝑎 ≠ 1
1−𝑎

1−𝑎 𝑁+1
2. ∑𝑁 𝑛
𝑛=0 𝑎 = if 𝑎 ≠ 1
1−𝑎

3. ∑𝑁
𝑛=0 1 = 𝑁 + 1

4. ∑𝑁−1
𝑛=0 1 = 𝑁

5. ∑𝑁2
𝑛=𝑁1 1 = 𝑁2 − 𝑁1 + 1

𝑁(𝑁+1)(2𝑁+1)
6. ∑𝑁 2
𝑛=0 𝑛 = 6
2𝑁(𝑁+1)
7. ∑𝑁 3
𝑛=0 𝑛 = 4
1
8. ∑∞ 𝑛
𝑛=0 𝑎 = 1−𝑎 if 𝑎 < 1
𝑎
9. ∑∞ 𝑛
𝑛=0 𝑛𝑎 = (1−𝑎)2 if 𝑎 < 1

Problem: Determine the power and energy of the unit step sequence.
Solution: Let the unit step signal 𝑥(𝑛) = 𝑢(𝑛)
 Prepared by Dr Harish M S

1 0≤𝑛≤∞
𝑥 (𝑛 ) = {
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
The average power P of the signal 𝑥(𝑛) is defined as
𝑁
1
𝑃≡ ∑ |𝑥(𝑛)|2
lim 2𝑁 + 1
𝑁→∞ 𝑛=−𝑁
𝑁 𝑁
1 1 𝑁+1
𝑃= ∑ 𝑢(𝑛)2 = ∑1 =
lim 2𝑁 + 1 lim 2𝑁 + 1 lim 2𝑁 + 1
𝑁→∞ 𝑛=0 𝑁→∞ 𝑛=0 𝑁→∞
1 1
(1+ )𝑁 (1+ ) 1
𝑁 𝑁
𝑃= 1 = 1 = 2 Watt
lim 𝑁(2+ ) lim (2+ )
𝑁→∞ 𝑁 𝑁→∞ 𝑁

The total energy E of the signal 𝑥(𝑛) is defined as

∞

𝐸 ≡ ∑ |𝑥(𝑛)|2
𝑛=−∞
∞ ∞

𝐸 ≡ ∑ 𝑢(𝑛)2 = ∑ 12 = 1 + 1 + 1 + 1 + ⋯ + 1 = ∞
𝑛=−∞ 𝑛=0

Since power is finite, energy is infinity, and the given signal is power signal.
2, 3, 4, 5 6
Problem: Determine the power and energy of the signal 𝑥(𝑛) = ( ).
↑
2, 3, 4, 5 6
Solution: Let signal 𝑥(𝑛) = ( )
↑
The average power P of the signal 𝑥(𝑛) is defined as
𝑁
1
𝑃≡ ∑ |𝑥(𝑛)|2
lim 2𝑁 + 1
𝑁→∞ 𝑛=−𝑁
3
1
𝑃 == ∑ |𝑥(𝑛)|2
lim 2𝑁 + 1
𝑁→∞ 𝑛=−1

1
= {|𝑥(−1)|2 + |𝑥(0)|2 + |𝑥(1)|2 + |𝑥(2)|2 + |𝑥(3)|2 }
lim 2𝑁 + 1
𝑁→∞

1 90
= {|2|2 + |3|2 + |4|2 + |5|2 + |6|2 } = =0
lim 2𝑁 + 1 lim 2𝑁 + 1
𝑁→∞ 𝑁→∞

The total energy E of the signal 𝑥(𝑛) is defined as

 Prepared by Dr Harish M S

∞ 3

𝐸 ≡ ∑ |𝑥(𝑛)|2 = ∑ |𝑥(𝑛)|2 = {|𝑥(−1)|2 + |𝑥(0)|2 + |𝑥(1)|2 + |𝑥(2)|2 + |𝑥(3)|2 }

𝑛=−∞ 𝑛=−1

= 90 Joules
Since power is zero, energy is finite and the given signal is energy signal.
Problem: Determine the power and energy of the signal 𝑥(𝑛) = 𝑛𝑢(𝑛)
Solution: Let 𝑥(𝑛) = 𝑛𝑢(𝑛)
𝑛 0≤𝑛≤∞
𝑥 (𝑛 ) = {
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
The total energy of the signal 𝐸 ≡ ∑∞
𝑛=−∞|𝑥(𝑛)|
2

∞ ∞

𝐸 ≡ ∑ [𝑛𝑢(𝑛)] = ∑ 𝑛2 = 0 + 12 + 22 + 32 + ⋯ = ∞
2

𝑛=−∞ 𝑛=0
1
The average power of the signal is 𝑃 ≡ ∑𝑁
𝑛=−𝑁 |𝑥(𝑛)|
2
lim 2𝑁+1
𝑁→∞

𝑁 𝑁
1 1 𝑁(𝑁 + 1)(2𝑁 + 1) 𝑁(𝑁 + 1)
𝑃= ∑ |𝑛 |2 = ∑ 𝑛2 = = =∞
lim 2𝑁 + 1 lim 2𝑁 + 1 6 lim 2𝑁 + 1 6 lim
𝑁→∞ 𝑛=0 𝑁→∞ 𝑛=0 𝑁→∞ 𝑁→∞

Since energy is infinite, power is infinite; the given signal is neither energy nor a power signal.
Problem: Determine the power and energy of the signal 𝑥(𝑛) = 𝐴𝑒 𝑗𝜔0 𝑛
Solution: Let 𝑥(𝑛) = 𝐴𝑒 𝑗𝜔0 𝑛
𝑥(𝑛) = 𝐴𝑒 𝑗𝜔0 𝑛 −∞≤𝑛≤∞
The total energy of the signal 𝐸 ≡ ∑∞
𝑛=−∞|𝑥(𝑛)|
2

∞ ∞ ∞ ∞

𝐸 = ∑ |𝑥(𝑛) |2 = ∑ 𝑥(𝑛)𝑥 𝑛) = ∑ 𝐴𝑒∗( 𝑗𝜔0 𝑛

× 𝐴𝑒 −𝑗𝜔0 𝑛
= ∑ 𝐴2 = ∞
𝑛=−∞ 𝑛=−∞ 𝑛=−∞ 𝑛=−∞

The average power of the periodic signal is

𝑁−1 𝑁−1 𝑁−1
1 1 1 1
𝑃 ≡ ∑ |𝑥(𝑛)|2 = ∑ 𝑥(𝑛)𝑥 ∗ (𝑛) = ∑ 𝐴2 = 𝐴2 𝑁 = 𝐴2 𝑊𝑎𝑡𝑡𝑠
𝑁 𝑁 𝑁 𝑁
𝑛=0 𝑛=0 𝑛=0

Since energy is infinite, power is finite; the given signal is a power signal.
Problem: Determine the power and energy of the signal 𝑥(𝑛) = 2𝑒 𝑗3𝑛
Solution: Let 𝑥(𝑛) = 2𝑒 𝑗3𝑛
𝑥(𝑛) = 2𝑒 𝑗3𝑛 −∞≤𝑛 ≤∞
The signal 𝑥(𝑛) is a non periodic signal
 Prepared by Dr Harish M S

The total energy of the signal 𝐸 ≡ ∑∞

𝑛=−∞|𝑥(𝑛)|
2

∞ ∞ ∞ ∞

𝐸 = ∑ |𝑥(𝑛)|2 = ∑ 𝑥(𝑛)𝑥 ∗ (𝑛) = ∑ 2𝑒 𝑗3𝑛 × 2𝑒 −𝑗3𝑛 = ∑ 4 = ∞

𝑛=−∞ 𝑛=−∞ 𝑛=−∞ 𝑛=−∞

The average power of the nonperiodic signal is

𝑁 𝑁 𝑁
1 1 1
𝑃≡ ∑ |𝑥(𝑛)|2 = ∑ 𝑥 (𝑛 )𝑥 ∗ (𝑛 ) = ∑ 4
lim 2𝑁 + 1 lim 2𝑁 + 1 lim 2𝑁 + 1
𝑁→∞ 𝑛=−𝑁 𝑁→∞ 𝑛=−𝑁 𝑁→∞ 𝑛=−𝑁

4(2𝑁 + 1)
𝑃= = 4𝑤𝑎𝑡𝑡𝑠
lim 2𝑁 + 1
𝑁→∞

Since energy is infinite, power is finite; the given signal is a power signal.
1
Problem: Determine the power and energy of the signal 𝑥(𝑛) = (3)𝑛 𝑢(𝑛)
1
Solution: Let 𝑥(𝑛) = (3)𝑛 𝑢(𝑛)
1 𝑛
𝑥 𝑛 = { 3)
( ) ( 0≤𝑛≤∞
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
The total energy of the signal 𝐸 ≡ ∑∞
𝑛=−∞|𝑥(𝑛)|
2

∞ ∞
1 1 1 9
𝐸 ≡ ∑( )2𝑛 = ∑( )𝑛 = 1 = Joules
3 9 1−9 8
𝑛=0 𝑛=0
1
The average power of the signal is 𝑃 ≡ ∑𝑁
𝑛=−𝑁 |𝑥(𝑛)|
2
lim 2𝑁+1
𝑁→∞

𝑁 𝑁 1
1 1 2𝑛 1 1 𝑛 1 1 − (9)𝑁+1
𝑃= ∑( ) = ∑ ( ) = × 1 =0
lim 2𝑁 + 1 3 lim 2𝑁 + 1 9 lim 2𝑁 + 1 1−
𝑁→∞ 𝑛=0 𝑁→∞ 𝑛=0 𝑁→∞ 9

Periodic and aperiodic signals

The discrete time signal 𝑥(𝑛) is said to be periodic if and only if
𝑥 (𝑛 + 𝑁 ) = 𝑥 (𝑛 ) for all n
If this condition is not satisfied, then the signal is known as aperiodic or non-periodic signal.
The smallest value of N is called fundamental period of 𝑥(𝑛).
For example, let us consider 𝑥(𝑛) = 𝑠𝑖𝑛2𝜋𝑓0 𝑛
The signal 𝑥(𝑛) is periodic when discrete frequency 𝑓0 is a rational number, that is, if 𝑓0 can be
k
expressed as f0 = N, where k and N are integers. We have to select the smallest integer k such

that N should be an integer.

 Prepared by Dr Harish M S

k 2πk 2πk
∴N= = =
f0 2πf0 ω0
Problem: Find period of the signal 𝑥(𝑛) = 5𝑠𝑖𝑛(0.2𝜋𝑛)
Solution: 𝑥(𝑛) = 5𝑠𝑖𝑛(0.2𝜋𝑛) = 5𝑠𝑖𝑛𝜔0 𝑛
2πk 2𝜋𝑘
Period of the signal N = = 0.2𝜋 = 10𝑘
ω0

If we select 𝑘 = 1, then N is an integer, 𝑁 = 10 samples in one period.

Problem: Find period of the signal 𝑥(𝑛) = 5𝑠𝑖𝑛(0.2𝑛)
Solution: 𝑥(𝑛) = 5𝑠𝑖𝑛(0.2𝑛) = 5𝑠𝑖𝑛𝜔0 𝑛
2πk 2𝜋𝑘
Period of the signal N = = = 10𝜋𝑘
ω0 0.2

If we select any integer value for 𝑘, N is not an integer; the given signal is non-periodic signal.
Symmetric (even) and antisymmetric (odd) signals
A real valued signal 𝑥(𝑛) is called symmetric (even) if
𝑥(−𝑛) = 𝑥(𝑛)
On the other hand, a signal 𝑥(𝑛) is called antisymmetric (odd) if
𝑥(−𝑛) = −𝑥(𝑛)
Note that if 𝑥(𝑛) is add, then 𝑥(0) = 0
Examples of even and odd signal as shown in figure 1.15.

In any given signal 𝑥 (𝑛), can be expressed as sum of even part and add part of (𝑛) . That is
𝑥(𝑛) = 𝑥𝑒 (𝑛) + 𝑥𝑜 (𝑛)
Where, 𝑥𝑒 (𝑛) is the even part of 𝑥(𝑛) and 𝑥𝑜 (𝑛) is the odd part of 𝑥(𝑛). The even part of 𝑥(𝑛)
is computed using the expression
1
𝑥𝑒 (𝑛) = [𝑥(𝑛) + 𝑥(−𝑛)]
2
 Prepared by Dr Harish M S

And odd part of 𝑥(𝑛) is computed using the expression

1
𝑥𝑜 (𝑛) = [𝑥(𝑛) − 𝑥(−𝑛)]
2
Problem: Sketch the even and odd part of the signal 𝑥(𝑛) as shown in below figure1.17.

2, 3, 4, 5, 6
Solution: 𝑥(𝑛) = ( ) −2 ≤𝑛 ≤ 2
↑
6, 5, 4, 3, 2
𝑥(−𝑛) = 𝑥 ′ (𝑛) = ( ) −2 ≤𝑛 ≤2
↑
The even part of 𝑥(𝑛)
1
𝑥 𝑒 (𝑛 ) = [𝑥(𝑛) + 𝑥(−𝑛)] − 2 ≤ 𝑛 ≤ 2
2
1
𝑥𝑒 (𝑛) = [𝑥(𝑛) + 𝑥′(𝑛)] − 2 ≤ 𝑛 ≤ 2
2
1 1
𝑥𝑒 (−2) = [𝑥(−2) + 𝑥′(−2)] = [2 + 6] = 4
2 2
1 1
𝑥𝑒 (−1) = [𝑥(−1) + 𝑥′(−1)] = [3 + 5] = 4
2 2
1 1
𝑥𝑒 (0) = [𝑥(0) + 𝑥′(0)] = [4 + 4] = 4
2 2
1 1
𝑥𝑒 (1) = [𝑥(1) + 𝑥′(1)] = [5 + 3] = 4
2 2
1 1
𝑥𝑒 (2) = [𝑥(2) + 𝑥′(2)] = [6 + 2] = 4
2 2
4, 4, 4, 4, 4
𝑥 𝑒 (𝑛) = ( )
↑
The odd part of 𝑥(𝑛)
1
𝑥0 (𝑛) = [𝑥(𝑛) − 𝑥(−𝑛)] −2≤𝑛≤2
2
1
𝑥0 (𝑛) = [𝑥(𝑛) − 𝑥′(𝑛)] −2≤𝑛≤2
2
 Prepared by Dr Harish M S

1 1
𝑥0 (−2) = [𝑥(−2) − 𝑥′(−2)] = [2 − 6] = −2
2 2
1 1
𝑥0 (−1) = [𝑥(−1) − 𝑥′(−1)] = [3 − 5] = −1
2 2
1 1
𝑥0 (0) = [𝑥(0) − 𝑥′(0)] = [4 − 4] = 0
2 2
1 1
𝑥0 (1) = [𝑥(1) − 𝑥′(1)] = [5 − 3] = 1
2 2
1 1
𝑥0 (𝑛) = [𝑥(2) − 𝑥′(2)] = [6 − 2] = 2
2 2
−2, −1, 0, 1, 2
𝑥 0 (𝑛 ) = ( )
↑

Problem: Sketch the even and odd part of the signal 𝑥(𝑛) as shown in figure 1.20

2, 2, 2, 2, −2, −2, −2, −2

Solution: 𝑥(𝑛) = ( ) 0≤𝑛≤7
↑
−2, −2, −2, −2, 2, 2, 2, 2
𝑥(−𝑛) = 𝑥 ′ (𝑛) = ( ) −7≤𝑛≤0
↑
The even part of 𝑥(𝑛)
1
𝑥 𝑒 (𝑛 ) = [𝑥(𝑛) + 𝑥(−𝑛)] −7≤𝑛≤7
2
1
𝑥𝑒 (𝑛) = [𝑥(𝑛) + 𝑥′(𝑛)] −7 ≤𝑛 ≤ 7
2
 Prepared by Dr Harish M S

1 1
𝑥𝑒 (−7) = [𝑥(−7) + 𝑥′(−7)] = [0 + (−2)] = −1
2 2
1 1
𝑥𝑒 (−6) = [𝑥(−6) + 𝑥′(−6)] = [0 + (−2)] = −1
2 2
1 1
𝑥𝑒 (−5) = [𝑥(−5) + 𝑥′(−5)] = [0 + (−2)] = −1
2 2
1 1
𝑥𝑒 (−4) = [𝑥(−4) + 𝑥′(−4)] = [0 + (−2)] = −1
2 2
1 1
𝑥𝑒 (−3) = [𝑥(−3) + 𝑥′(−3)] = [0 + (2)] = 1
2 2
1 1
𝑥𝑒 (−2) = [𝑥(−2) + 𝑥′(−2)] = [0 + (2)] = 1
2 2
1 1
𝑥𝑒 (−1) = [𝑥(−1) + 𝑥′(−1)] = [0 + (2)] = 1
2 2
1 1
𝑥𝑒 (0) = [𝑥(0) + 𝑥′(0)] = [2 + 2)] = 2
2 2
1 1
𝑥𝑒 (1) = [𝑥(1) + 𝑥′(1)] = [2 + 0] = 1
2 2
1 1
𝑥𝑒 (2) = [𝑥(2) + 𝑥′(2)] = [2 + 0] = 1
2 2
1 1
𝑥𝑒 (3) = [𝑥(3) + 𝑥′(3)] = [2 + 0] = 1
2 2
1 1
𝑥𝑒 (4) = [𝑥(4) + 𝑥′(4)] = [−2 + 0] = −1
2 2
1 1
𝑥𝑒 (5) = [𝑥(5) + 𝑥′(5)] = [−2 + 0)] = −1
2 2
1 1
𝑥𝑒 (6) = [𝑥(6) + 𝑥′(6)] = [−2 + 0] = −1
2 2
1 1
𝑥𝑒 (7) = [𝑥(7) + 𝑥′(7)] = [−2 + 0] = −1
2 2
−,1 − 1, −1, −1,1,1,1, 2, 1, 1,1, −1, −1, −1, −1
𝑥 𝑒 (𝑛 ) = ( )
↑
 Prepared by Dr Harish M S

The odd part of 𝑥(𝑛)

1
𝑥0 (𝑛) = [𝑥(𝑛) − 𝑥(−𝑛)] − 7 ≤ 𝑛 ≤ 7
2
1
𝑥0 (𝑛) = [𝑥(𝑛) − 𝑥′(𝑛)] − 7 ≤ 𝑛 ≤ 7
2
1 1
𝑥0 (−7) = [𝑥(−7) − 𝑥′(−7)] = [0 − (−2)] = 1
2 2
1 1
𝑥0 (−6) = [𝑥(−6) − 𝑥′(−6)] = [0 − (−2)] = 1
2 2
1 1
𝑥0 (−5) = [𝑥(−5) − 𝑥′(−5)] = [0 − (−2)] = 1
2 2
1 1
𝑥0 (−4) = [𝑥(−4) − 𝑥′(−4)] = [0 − (−2)] = 1
2 2
1 1
𝑥0 (−3) = [𝑥(−3) − 𝑥′(−3)] = [0 − (2)] = −1
2 2
1 1
𝑥0 (−2) = [𝑥(−2) − 𝑥′(−2)] = [0 − (2)] = −1
2 2
1 1
𝑥0 (−1) = [𝑥(−1) − 𝑥′(−1)] = [0 − (2)] = −1
2 2
1 1
𝑥0 (0) = [𝑥(0) − 𝑥′(0)] = [2 − 2)] = 0
2 2
1 1
𝑥0 (1) = [𝑥(1) − 𝑥′(1)] = [2 − 0] = 1
2 2
1 1
𝑥0 (2) = [𝑥(2) − 𝑥′(2)] = [2 − 0] = 1
2 2
1 1
𝑥0 (3) = [𝑥(3) − 𝑥′(3)] = [2 − 0] = 1
2 2
1 1
𝑥0 (4) = [𝑥(4) − 𝑥′(4)] = [−2 − 0] = −1
2 2
 Prepared by Dr Harish M S

1 1
𝑥 0 (5) = [𝑥(5) − 𝑥′(5)] = [−2 − 0)] = −1
2 2
1 1
𝑥0 (6) = [𝑥(6) − 𝑥′(6)] = [−2 − 0] = −1
2 2
1 1
𝑥0 (7) = [𝑥(7) − 𝑥′(7)] = [−2 − 0] = −1
2 2
1,1,1,1, −1, −1, −1, 0, 1, 1,1, −1, −1, −1, −1
𝑥 0 (𝑛 ) = ( )
↑

Simple manipulations of Discrete-Time Signals

In this section we consider some simple manipulations or modifications on signal independent
variable and signal amplitude (dependent variable). The modifications on the independent
variables are
1. Time shifting
2. Folding or reflection
3. Time scaling
The modifications on the amplitude are
1. Addition
2. Multiplication
3. Scaling of sequences
Time shifting: A signal 𝑥(𝑛) may be shifted in time by replacing the independent variable 𝑛
by 𝑛 − 𝑘, where 𝑘 is an integer. If 𝑘 is positive, then the signal is delayed by 𝑘 samples. If 𝑘 is
negative, then the signal is advanced by k samples.
Problem: A signal 𝑥(𝑛) as shown in figure 1.23. Sketch the time shifted signal 𝑥(𝑛 − 3)
and 𝑥(𝑛 + 2).
 Prepared by Dr Harish M S

−1, 0, 1, 2, 3, 4, 4, 4, 4, 4
Solution: Let the sequences of the signal is 𝑥(𝑛) = ( ), −5 ≤𝑛 ≤4
↑
Let 𝑦(𝑛) = 𝑥(𝑛 − 3) , − 2 ≤ 𝑛 ≤ 7
𝑦(−2) = 𝑥( −2 − 3) = 𝑥(−5) = −1
𝑦(−1) = 𝑥(−1 − 3) = 𝑥(−4) = 0
𝑦(0) = 𝑥(0 − 3) = 𝑥(−3) = 1
𝑦(1) = 𝑥(1 − 3) = 𝑥(−2) = 2
𝑦(2) = 𝑥(2 − 3) = 𝑥(−1) = 3
𝑦(3) = 𝑥(3 − 3) = 𝑥(0) = 4
𝑦(4) = 𝑥(4 − 3) = 𝑥(1) = 4
𝑦(5) = 𝑥(5 − 3) = 𝑥(2) = 4
𝑦(6) = 𝑥(6 − 3) = 𝑥(3) = 4
𝑦(7) = 𝑥(7 − 3) = 𝑥(4) = 4
𝑦(8) = 𝑥(8 − 3) = 𝑥(5) = 0
−1, 0, 1, 2, 3, 4, 4, 4, 4, 4
∴ 𝑦( 𝑛 ) = 𝑥 ( 𝑛 − 3 ) = ( ), − 2 ≤ 𝑛 ≤ 7
↑
Let 𝑧(𝑛) = 𝑥(𝑛 + 2) , −7≤𝑛≤2
𝑧(−7) = 𝑥(−7 + 2) = 𝑥 (−5) = −1
𝑧(−6) = 𝑥(−6 + 2) = 𝑥(−4) = 0
𝑧(−5) = 𝑥(−5 + 2) = 𝑥(−3) = 1
𝑧(−4) = 𝑥(−4 + 2) = 𝑥(−2) = 2
𝑧(−3) = 𝑥(−3 + 2) = 𝑥(−1) = 3
𝑧(−2) = 𝑥(−2 + 2) = 𝑥(0) = 4
𝑧(−1) = 𝑥(−1 + 2) = 𝑥(1) = 4
𝑧 (0) = 𝑥 (0 + 2) = 𝑥 (2) = 4
𝑧 (1) = 𝑥 (1 + 2) = 𝑥 (3) = 4
 Prepared by Dr Harish M S

𝑧 (2) = 𝑥 (2 + 2) = 𝑥 (4) = 4
𝑧 (3) = 𝑥 (3 + 2) = 𝑥 ( 5) = 0
−1, 0, 1, 2, 3, 4, 4, 4, 4, 4
∴ 𝑧 (𝑛 ) = 𝑥 (𝑛 + 2) = ( ), − 7 ≤ 𝑛 ≤ 2
↑
The sketch of 𝑦(𝑛) and 𝑧(𝑛) as shown in figure 1.24

Folding or reflection: The second modification on the independent variable of the signal is the
folded version of the signal 𝑥(𝑛) obtained by replacing the independent variable 𝑛 by −𝑛 .
Problem: A signal 𝑥(𝑛) as shown in figure 1.25. Sketch the time shifted signal 𝑥(−𝑛)
and 𝑥(−𝑛 + 2).

2, 2, 2, 0, 1, 2, 3, 4
Solution: Let the sequences of the signal is 𝑥 (𝑛) = ( ), −3 ≤𝑛 ≤4
↑
Let the folded signal 𝑦(𝑛) = 𝑥(−𝑛) , −4 ≤𝑛 ≤3
𝑦(−4) = 𝑥(−(−4)) = 𝑥(4) = 4
𝑦(−3) = 𝑥(−(−3)) = 𝑥(3) = 3
𝑦(−2) = 𝑥(−(−2)) = 2
𝑦(−1) = 𝑥(−(−1)) = 𝑥(1) = 1
𝑦(0) = 𝑥(−0) = 𝑥 (0) = 0
𝑦(1) = 𝑥(−1) = 2
𝑦(2) = 𝑥(−2) = 2
𝑦(3) = 𝑥(−3) = 2
4, 3, 2, 1, 0, 2, 2, 2
∴ 𝑦(𝑛) = 𝑥(−𝑛) = ( ), −4 ≤𝑛 ≤3
↑
 Prepared by Dr Harish M S

Let the folded and time shifted signal 𝑧(𝑛) = 𝑥(−𝑛 + 2) , − 2 ≤ 𝑛 ≤ 5

𝑧(−2) = 𝑥 (−(−2) + 2) = 𝑥 (2 + 2) = 𝑥(4) = 4
𝑧(−1) = 𝑥 (−(−1) + 2) = 𝑥 (1 + 2) = 𝑥(3) = 3
𝑧(0) = 𝑥(−0 + 2) = 𝑥(2) = 2
𝑧(1) = 𝑥(−1 + 2) = 𝑥(1) = 1
𝑧(2) = 𝑥(−2 + 2) = 𝑥(0) = 0
𝑧(3) = 𝑥 (−3 + 2) = 𝑥 (−1) = 2
𝑧(4) = 𝑥 (−4 + 2) = 𝑥 (−2) = 2
𝑧(5) = 𝑥 (−5 + 2) = 𝑥 (−3) = 2
4,3,2,1,0,2,2,2
∴ 𝑧(𝑛) = 𝑥(−𝑛 + 2) = ( )
↑
The sketch of 𝑦(𝑛) and 𝑧(𝑛 ) as shown in figure 1.26

It is important to note that the operation of folding and time delaying (or advance) signals are not
commutative. If we denote the time-delay operation by TD and the signal folding operation by
FD, we can write
𝑇𝐷𝑘 [𝑥(𝑛)] = 𝑥(𝑛 − 𝑘) 𝑘>0
𝐹𝐷[𝑥(𝑛)] = 𝑥(−𝑛)
Now
𝑇𝐷𝑘 {𝐹𝐷[𝑥(𝑛)]} = 𝑇𝐷𝑘 {𝑥 (−𝑛)] = 𝑥(−𝑛 + 𝑘) … … … . (𝐴)
Where as
𝐹𝐷{𝑇𝐷𝑘 [𝑥(𝑛)]} = 𝐹𝐷{𝑥(𝑛 − 𝑘)] = 𝑥(−𝑛 − 𝑘) … … … . (𝐵)
Since equation A and B are not same, the time-delaying and folding operations are not
commutative.
Time scaling: Third modification on independent variable of the signal is time scaling obtained
by replacing the independent variable 𝑛 by 𝜇𝑛. Where 𝜇 is the scaling factor, if 𝜇 is greater than
 Prepared by Dr Harish M S

1 signal is compressed called down sampling, if 𝜇 is less than 1 signal is expanded called up-
sampling.
Problem: The signal 𝑥(𝑛) as shown in figure 1.27. Sketch the signal 𝑥(2𝑛).

Solution: Let the sequences of 𝑥(𝑛) are

−3, −2, −1, 0,1,2, 3,4, 4, 4, 4, 4, 4, 4
𝑥 (𝑛 ) = ( ), − 7 ≤ 𝑛 ≤ 6
↑
Let 𝑦(𝑛) = 𝑥(2𝑛), −3 ≤ 𝑛 ≤ 3
𝑦(−3) = 𝑥(2(−3)) = x(−6) = −2
𝑦(−2) = 𝑥(2(−2)) = x(−4) = 0
𝑦(−1) = 𝑥(2(−1)) = x(−2) = 2
𝑦(0) = 𝑥(2(0)) = x(0) = 4
𝑦(1) = 𝑥(2(1)) = x(2) = 4
𝑦(2) = 𝑥(2(2)) = x(4) = 4
𝑦(3) = 𝑥(2(3)) = x(6) = 4
Since the scaling factor is greater than 1, signal is compressed, it is downsampling process, and
during the down sampling some samples are lost. The sketch of 𝑦(𝑛) = 𝑥(2𝑛) is as shown in
figure 1.28.
 Prepared by Dr Harish M S

Problem: The signal 𝑥(𝑛) as shown in figure 1.27. Sketch the signal 𝑥(0.5𝑛).

Solution: Let the sequences of 𝑥(𝑛) are

−3, −2, −1, 0,1,2, 3,4, 4, 4, 4, 4, 4, 4
𝑥 (𝑛 ) = ( ), − 7 ≤ 𝑛 ≤ 6
↑
Let 𝑦(𝑛) = 𝑥(0.5𝑛), −14 ≤ 𝑛 ≤ 12
𝑦(−14) = 𝑥(0.5(−14)) = x(−7) = −3
𝑦(−13) = 𝑥(0.5(−13)) = x(−6.5) = 0
𝑦(−12) = 𝑥(0.5(−12)) = x(−6) = −2
𝑦(−11) = 𝑥(0.5(−11)) = x(−5.5) = 0
𝑦(−10) = 𝑥(0.5(−10)) = x(−5) = −1
𝑦(−9) = 𝑥(0.5(−9)) = x(−4.5) = 0
𝑦(−8) = 𝑥(0.5(−8)) = x(−4) = 0
𝑦(−7) = 𝑥(0.5(−7)) = x(−3.5) = 0
𝑦(−6) = 𝑥(0.5(−6)) = x(−3) = 1
𝑦(−5) = 𝑥(0.5(−5)) = x(−2.5) = 0
𝑦(−4) = 𝑥(0.5(−4)) = x(−2) = 2
𝑦(−3) = 𝑥(0.5(−3)) = x(−1.5) = 0
𝑦(−2) = 𝑥(0.5(−2)) = x(−1) = 3
𝑦(−1) = 𝑥(0.5(−1)) = x(−0.5) = 0
𝑦(0) = 𝑥(0.5(0)) = x(0) = 4
𝑦(1) = 𝑥(0.5(1)) = x(0.5) = 0
𝑦(2) = 𝑥(0.5(2)) = x(1) = 4
𝑦(3) = 𝑥(0.5(3)) = x(1.5) = 0
 Prepared by Dr Harish M S

𝑦(4) = 𝑥(0.5(4)) = x(2) = 4

𝑦(5) = 𝑥(0.5(2)) = x(2.5) = 0
𝑦(6) = 𝑥(0.5(6)) = x(3) = 4
𝑦(7) = 𝑥(0.5(7)) = x(3.5) = 0
𝑦(8) = 𝑥(0.5(8)) = x(4) = 4
𝑦(9) = 𝑥(0.5(9)) = x(4.5) = 0
𝑦(10) = 𝑥(0.5(10)) = x(5) = 4
𝑦(11) = 𝑥(0.5(11)) = x(5.5) = 0
𝑦(12) = 𝑥(0.5(12)) = x(6) = 4
𝑦(13) = 𝑥(0.5(13)) = x(6.5) = 0
𝑦(14) = 𝑥(0.5(14)) = x(7) = 4
−3, 0, −2,0, −1 ,0,0,0,1,0,2,0,3, 0,4, 0, 4, 0, 4, 0, 4,0,4,0,4,0,4
𝑦 (𝑛 ) = ( )
↑

Since the scaling factor is less than 1, signal is expanded, it is upsampling process. The sketch of
𝑦(𝑛) = 𝑥(0.5𝑛) is as shown in figure 1.29.
Addition, Multiplication and scaling of sequences: The amplitude modification includes
addition, multiplication, and scaling of discreet-time signals. Let 𝑥1 (𝑛) and 𝑥2 (𝑛)be a two
discrete time-signals. If we add these two signals, then the resulted signal is
𝑦(𝑛) = 𝑥1 (𝑛) + 𝑥2 (𝑛)
And, if we multiply these two signals resulted signal is
𝑦(𝑛) = 𝑥1 (𝑛)𝑥2 (𝑛)
Let 𝑥(𝑛) be a discrete time signal, if this signal is scaled by scaling factor A, then the resulted
signal is
𝑦(𝑛) = 𝐴𝑥(𝑛)
 Prepared by Dr Harish M S

Problem: A discrete-time signal 𝑥(𝑛) is defined as

𝑛
1 + , − 3 ≤ 𝑛 ≤ −1
3
𝑥 (𝑛 ) = {
1, 0≤𝑛≤3
0, 𝑒𝑙𝑠𝑒𝑤ℎ𝑒𝑟𝑒
a) Determine its value and sketch the signal 𝑥(𝑛)
b) Sketch the signals that result if we:
1. First fold 𝑥(𝑛) and then delay the resulting signal by four samples
2. First delay 𝑥(𝑛) by four samples and then fold the resulting signal
c) Sketch the signal 𝑥(−𝑛 + 4).
1 2
0, 3 , 3 , 1,1,1,1
Solution: a) 𝑥(𝑛) = ( ) , −3≤𝑛≤3
↑
b. 1. Folded signal 𝑥(𝑛), resultent signal
2 1
1,1 ,1,1, 3 , 3 , 0
𝑦(𝑛) = 𝑥(−𝑛) = ( ), −3 ≤𝑛 ≤ 3
↑
2 1
0,1,1,1,1, , , 0
Delayed by four samples 𝑧(𝑛) = 𝑦(𝑛 − 4) = ( 3 3 ), 1≤𝑛≤7
↑
2. Delay the signal 𝑥(𝑛) by four sample, resultant signal
1 2
0,0, 3 , 3 , 1,1,1,1
𝑦(𝑛) = 𝑥(𝑛 − 4) = ( ), 1≤𝑛≤7
↑
Folding the signal 𝑦(𝑛), resultant
1 2
1,1,1,1, 3 , 3 , 0,0
𝑧(𝑛) = 𝑦(−𝑛) = ( ) , − 7 ≤ 𝑛 ≤ −1
↑
c. Sketch the signal 𝑥(−𝑛 + 4)
1 2
𝑥 (𝑛) = (0, 3 , 3 , 1,1,1,1) , − 3 ≤ 𝑛 ≤ 3
↑
2 1
0,1,1,1,1, 3 , 3 , 0
Let 𝑦(𝑛) = 𝑥(−𝑛 + 4) = ( ), 1 ≤ 𝑛 ≤ 7
↑
1 1
1,1,1,1, 2 , 2
Problem: Signal 𝑥(𝑛) is given by 𝑥(𝑛) = ( ) , − 1 ≤ 𝑛 ≤ 4. Sketch the
↑
following signals.
(a) 𝑥(𝑛 − 2), (b)𝑥(4 − 𝑛), (c) 𝑥(𝑛 + 2), (d) 𝑥(𝑛)𝑢(2 − 𝑛) (e) 𝑥 (𝑛 − 1)𝛿(𝑛 − 3),
 Prepared by Dr Harish M S

(f) 𝑥 (𝑛2 ), (g) even part of 𝑥(𝑛) (h) odd part of 𝑥(𝑛)
1 1
1,1,1,1, 2 , 2
Solution: 𝑥(𝑛) = ( ) , −1≤𝑛≤4
↑

1 1
0,1,1,1,1, 2 , 2
(a) 𝑦1 (𝑛) = 𝑥(𝑛 − 2) = ( ), 1 ≤ 𝑛 ≤ 6
↑

1 1
, , 1,1,1,1
(b) 𝑦2 (𝑛) = 𝑥(4 − 𝑛) = (2 2 ), 0≤𝑛≤5
↑

1 1
1,1,1,1, 2 , 2
(c) 𝑦3 (𝑛) = 𝑥(𝑛 + 2) = ( ), − 3 ≤ 𝑛 ≤ 2
↑

1,1,1,1
(d) 𝑦4 (𝑛) = 𝑥(𝑛)𝑢(2 − 𝑛) = ( ), − 1 ≤ 𝑛 ≤ 2
↑
 Prepared by Dr Harish M S

(e) 𝑦5 (𝑛) = 𝑥(𝑛 − 1)𝛿(𝑛 − 3)

𝑦5 (3) = 𝑥(𝑛 − 1)𝛿(𝑛 − 3) = 𝑥(3 − 1)𝛿 (3 − 3) = 𝑥(2)𝛿(0) = 𝑥(2) = 1
0,0,0,1
𝑦5 (3) = ( )
↑
1 1
, 1, 1,1, 2
(f) 𝑦6 (𝑛) = 𝑥(𝑛2 ) = (… . 𝑥 (4), 𝑥(1), 𝑥(0), 𝑥(1), 𝑥(4)) = (2 ), − 2 ≤ 𝑛 ≤ 2
↑
1
(g) The even part of 𝑥(𝑛) is 𝑦7 (𝑛) = 2 [𝑥(𝑛) + 𝑥(−𝑛)] , − 4 ≤ 𝑛 ≤ 4
1
𝑦7 (𝑛) = [𝑥(𝑛) + 𝑥′(𝑛)]
2
1 1 1 1
𝑦7 (−4) = [𝑥(−4) + 𝑥′(−4)] = [0 + ] =
2 2 2 4
1 1 1 1
𝑦7 (−3) = [𝑥(−3) + 𝑥′(−3)] = [0 + ] =
2 2 2 4
1 1 1
𝑦7 (−2) = [𝑥(−2) + 𝑥′(−2)] = [0 + 1] =
2 2 2
1 1
𝑦7 (−1) = [𝑥(−1) + 𝑥′(−1)] = [1 + 1] = 1
2 2
1 1
𝑦7 (0) = [𝑥(0) + 𝑥′(0)] = [1 + 1] = 1
2 2
1 1
𝑦7 (1) = [𝑥(1) + 𝑥′(1)] = [1 + 1] = 1
2 2
1 1 1
𝑦7 (2) = [𝑥(2) + 𝑥′(2)] = [1 + 0] =
2 2 2
1 1 1 1
𝑦7 (3) = [𝑥(3) + 𝑥′(3)] = [ + 0] =
2 2 2 4
1 1 1 1
𝑦7 (4) = [𝑥(4) + 𝑥′(4)] = [ + 0] =
2 2 2 4
1 1 1 1 1 1
𝑦7 (𝑛) = (4 , 4 , 2 , 1,1,1, 2 , 4 , 4) , − 4 ≤ 𝑛 ≤ 4
↑
1
(h) The odd part of 𝑥(𝑛) is 𝑦8 (𝑛) = 2 [𝑥(𝑛) − 𝑥(−𝑛)] , − 4 ≤ 𝑛 ≤ 4
 Prepared by Dr Harish M S

1
𝑦8 (𝑛) = [𝑥(𝑛) − 𝑥′(𝑛)]
2
1 1 1 1
𝑦8 (−4) = [𝑥(−4) − 𝑥′(−4)] = [0 − ] = −
2 2 2 4
1 1 1 1
𝑦8 (−3) = [𝑥(−3) − 𝑥′(−3)] = [0 − ] = −
2 2 2 4
1 1 1
𝑦8 (−2) = [𝑥(−2) − 𝑥′(−2)] = [0 − 1] = −
2 2 2
1 1
𝑦8 (−1) = [𝑥(−1) − 𝑥′(−1)] = [1 − 1] = 0
2 2
1 1
𝑦8 (0) = [𝑥(0) − 𝑥′(0)] = [1 − 1] = 0
2 2
1 1
𝑦8 (1) = [𝑥(1) − 𝑥′(1)] = [1 − 1] = 0
2 2
1 1 1
𝑦8 (2) = [𝑥(2) − 𝑥′(2)] = [1 − 0] =
2 2 2
1 1 1 1
𝑦8 (3) = [𝑥(3) − 𝑥′(3)] = [ − 0] =
2 2 2 4
1 1 1 1
𝑦8 (4) = [𝑥(4) − 𝑥′(4)] = [ − 0] =
2 2 2 4
1 1 1 1 1 1
−
𝑦8 (𝑛) = ( 4 4 2, − , − , 0,0,0, , ,
2 4 4), − 4 ≤ 𝑛 ≤ 4
↑
Problem: Show that 𝛿(𝑛) = 𝑢(𝑛) − 𝑢(𝑛 − 1)
Solution: The unit step signal is defined as 𝑢(𝑛) = 1 , 0≤𝑛≤∞
If the step signal is delayed by one sample then 𝑢(𝑛 − 1) = 1 , 1 ≤ 𝑛 ≤ ∞
Now 𝑢(𝑛) − 𝑢(𝑛 − 1) = 1, 𝑎𝑡 𝑛 = 0
0, 𝑛<0
( ) ( ) ( )
∴ 𝑢 𝑛 − 𝑢 𝑛 − 1 = 𝛿 𝑛 = { 1, 𝑛 = 0
0 𝑛>0

Problem: Show that

 Prepared by Dr Harish M S

𝑛 ∞

𝑢(𝑛) = ∑ 𝛿 (𝑘) = ∑ 𝛿(𝑛 − 𝑘)

𝑘=−∞ 𝑘=0

Solution: The unit step signal 𝑢(𝑛) = 1 , 0≤𝑛≤∞

𝑛

𝑢(𝑛) = ∑ 𝛿 (𝑘) = ⋯ … … 𝛿(𝑛 − 3) + 𝛿(𝑛 − 2) + 𝛿(𝑛 − 1) + 𝛿(𝑛)

𝑘=−∞

𝑢 (0) = ⋯ … … 𝛿 (0 − 3) + 𝛿 (0 − 2) + 𝛿 (0 − 1) + 𝛿 ( 0) = 1
𝑢 (1) = ⋯ … … 𝛿 (1 − 3) + 𝛿 (1 − 2) + 𝛿 (1 − 1) + 𝛿 (1) = 𝛿 (1 − 1) = 𝛿 (0) = 1
𝑢 (2) = ⋯ … … 𝛿 (2 − 3) + 𝛿 (2 − 2) + 𝛿 (2 − 1) + 𝛿 (2) = 𝛿 (2 − 2) = 𝛿 (0) = 1
𝑛

∴ 𝑢 (𝑛 ) = ∑ 𝛿 (𝑘 )
𝑘=−∞

Next 𝑢(𝑛) = ∑∞
𝑘=0 𝛿 (𝑛 − 𝑘 ) = 𝛿 (𝑛 ) + 𝛿( 𝑛 − 1) + 𝛿 (𝑛 − 2) + 𝛿 (𝑛 − 3) + ⋯ ….

𝑢(𝑛) = 𝛿 (𝑛) + 𝛿(𝑛 − 1) + 𝛿 (𝑛 − 2) + 𝛿(𝑛 − 3) + ⋯ ….

𝑢 (0) = 𝛿 (0) + 𝛿 (0 − 1) + 𝛿 (0 − 2) + 𝛿 (0 − 3) + ⋯ … . = 1
𝑢 (1) = 𝛿 (1) + 𝛿 ( 1 − 1) + 𝛿 (1 − 2) + 𝛿 (1 − 3) + ⋯ … . = 𝛿 (0) = 1
𝑢 (2) = 𝛿 (2) + 𝛿 ( 2 − 1) + 𝛿 (2 − 2) + 𝛿 (2 − 3) + ⋯ … . = 𝛿 (0) = 1
∞

∴ 𝑢(𝑛) = ∑ 𝛿(𝑛 − 𝑘)
𝑘=0

Discrete-Time Systems
We know that, system is an interconnection of components (device) or algorithm; it takes signal
as an input, and performs the prescribed operation on the input signal, produces the desired
output signal. The input signal is called as excitation and output signal is called as response. If
the system having discrete-time signal 𝑥(𝑛) as an input and producing another discrete-time
output 𝑦(𝑛), then the system is known as discrete time system. Mathematically, the input output
relation of the discrete-time system can be expressed as
𝑦(𝑛) ≡ 𝑇[𝑥(𝑛)]
Where, the symbol T denotes the operation on the input signal by the system. This mathematical
relationship is as shown in the figure 1.30.
 Prepared by Dr Harish M S

Problem: Determine the response of the following system to the input signal
|𝑛 | , −3≤ 𝑛≤3
𝑥 (𝑛 ) = {
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
a) 𝑦(𝑛) = 𝑥(𝑛) (identity system)
b) 𝑦(𝑛) = 𝑥(𝑛 − 1) (unit delay system)
c) 𝑦(𝑛) = 𝑥(𝑛 + 1) (unit advance system)
1
d) 𝑦(𝑛) = 3 [𝑥(𝑛 + 1) + 𝑥(𝑛) + 𝑥(𝑛 − 1)] (moving average filter)

e) 𝑦(𝑛) = 𝑚𝑒𝑑𝑖𝑎𝑛{𝑥(𝑛 + 1), 𝑥(𝑛), 𝑥(𝑛 − 1)} (median filter)

f) 𝑦(𝑛) = ∑𝑛𝑘=−∞ 𝑥 (𝑘) = 𝑥(𝑛) + 𝑥(𝑛 − 1) + 𝑥(𝑛 − 2) + ⋯ . .(accumulator)
Solution:
a) 𝑦(𝑛) = 𝑥(𝑛) (identity system)
3,2 1,0,1,2,3
𝑥 (𝑛 ) = ( ), − 3 ≤ 𝑛 ≤ 3
↑

3,2 1,0,1,2,3
𝑦 (𝑛 ) = ( ), −3≤𝑛≤3
↑
b) 𝑦(𝑛) = 𝑥(𝑛 − 1) (unit delay system)
3,2 1,0,1,2,3
𝑦 (𝑛 ) = ( ), − 2 ≤ 𝑛 ≤ 4
↑
c) 𝑦(𝑛) = 𝑥(𝑛 + 1) (unit advance system)
3,2 1,0,1,2,3
( ), − 4 ≤ 𝑛 ≤ 2
↑
1
d) 𝑦(𝑛) = 3 [𝑥(𝑛 + 1) + 𝑥(𝑛) + 𝑥(𝑛 − 1)] (moving average filter)
1
𝑦(𝑛) = [𝑥(𝑛 + 1) + 𝑥(𝑛) + 𝑥(𝑛 − 1)] , − 4 ≤ 𝑛 ≤ 4
3
1 1 2
𝑦(0) = [𝑥(0 + 1) + 𝑥(0) + 𝑥(0 − 1)] = [1 + 0 + 1] =
3 3 3
 Prepared by Dr Harish M S

5 2 5
( )
𝑦 𝑛 =( 31, , 2,1, , 1,2, ,1
3 3 ), − 4 ≤ 𝑛 ≤ 4
↑
e) 𝑦(𝑛) = 𝑚𝑒𝑑𝑖𝑎𝑛{𝑥(𝑛 + 1), 𝑥(𝑛), 𝑥(𝑛 − 1)} (median filter)
𝑦(𝑛) = 𝑚𝑒𝑑𝑖𝑎𝑛{𝑥(𝑛 + 1), 𝑥(𝑛), 𝑥(𝑛 − 1)}, −4 ≤ 𝑛 ≤ 4
Let us find the median at 𝑛 = 0,
At 𝑛 = 0, 𝑥(0 + 1) = 𝑥(1) = 1
At 𝑛 = 0, 𝑥(0) = 𝑥(0) = 0
At 𝑛 = 0, 𝑥(0 − 1) = 𝑥(−1) = 1
Now consider the set is (1,0,1), write in the ascending order we get(0,1,1). Here numbers of
𝑛+1 3+1
digits are 𝑛 = 3. The median is = = 2. Thus second digit is the median i, e.1
2 2

0,2,2,1,1,1,2,2,0, … . .
∴ 𝑦(𝑛) = 𝑚𝑒𝑑𝑖𝑎𝑛{𝑥(𝑛 + 1), 𝑥(𝑛), 𝑥(𝑛 − 1)} = ( )
↑
f) 𝑦(𝑛) = ∑𝑛𝑘=−∞ 𝑥 (𝑘) = 𝑥(𝑛) + 𝑥(𝑛 − 1) + 𝑥(𝑛 − 2) + ⋯ . .(accumulator)
𝑛

𝑦(𝑛) = ∑ 𝑥 (𝑘) = ⋯ … 𝑥 (𝑛 − 2) + 𝑥 (𝑛 − 1) + 𝑥(𝑛)

𝑘=−∞
𝑛−1

𝑦(𝑛 − 1) = ∑ 𝑥(𝑘) = ⋯ . 𝑥(𝑛 − 2) + 𝑥(𝑛 − 1)

𝑘=−∞

∴ 𝑦(𝑛) = 𝑦(𝑛 − 1) + 𝑥(𝑛)

𝑦(−3) = y(−4) + x(−3) = 0 + 3 = 3
𝑦(−2) = y(−3) + x(−2) = 3 + 2 = 5
𝑦(−1) = y(−2) + x(−1) = 5 + 1 = 6
𝑦(0) = y(−1) + x(0) = 6 + 0 = 6
𝑦(1) = y(0) + x(1) = 6 + 1 = 7
𝑦(2) = y(1) + x(2) = 7 + 2 = 9
𝑦(3) = y(2) + x(3) = 9 + 3 = 12
… … ,3,5,6,6,7,9,12, … . .
𝑦 (𝑛 ) = ( )
↑
Thus system computes the current value of the output by adding the current input value to the
previous output value.
Block Diagram representation of Discrete-Time System
 Prepared by Dr Harish M S

To introduce a block diagram representation of the system, we need some building blocks such
as adder, a constant multiplier, a signal multiplier, a unit delay and unit advance, are
interconnected to form complex systems.
An Adder: Figure 1.31 illustrates adder that perform the addition of two signal sequences to
form another signal sequences.

A constant multiplier: Figure 1.32 illustrates the multiplication scaling factor a with the input
signal sequences.

A signal multiplier: Figure 1.33 illustrates the multiplication of two input signal sequences to
form another signal sequences.

A unit delay element: The unit delay is a system that simply delays the signal passing through it
by one sample. Figure 1.34 illustrates such system. If the input signal is 𝑥(𝑛) , the output
is 𝑥(𝑛 − 1). In fact, the sample 𝑥(𝑛 − 1) is stored in memory at time 𝑛 − 1 and it is recalled
from the memory at time n to form
𝑦(𝑛) = 𝑥(𝑛 − 1)
Thus it requires memory, the use of symbol 𝑧 −1 to denote the unit delay.
 Prepared by Dr Harish M S

A unit advance element: In contrast to the unit delay element, a unit advance moves the input
𝑥(𝑛) ahead by one sample in time 𝑥(𝑛 + 1), i.e. it predicting the future input sample. Figure
1.35 illustrates this operation, with the operator z being used to denote the unit advance. In the
real time application it is impossible. On the other hand, for non real time application, if we store
the signal in the computer memory, we can recall any sample at any time.

Problem: Using the basic building blocks, sketch the block diagram representation of the
discrete-time system described by the input-output relation
1 1 1
𝑦(𝑛) = 𝑦(𝑛 − 1) + 𝑥(𝑛) + 𝑥(𝑛 − 1)
4 2 2
Where, 𝑥 (𝑛) is the input and 𝑦 (𝑛) is the output of the system.
Solution:

Classification of Discrete-time systems

We know that, system is an interconnection of components, it takes signal as an input, and
performs the prescribed operation on the input signal, produces the desired output signal. If the
input signal to the system is 𝑥(𝑛) and the corresponding is output is 𝑦(𝑛). Mathematically,
the input output relation of the discrete-time system can be expressed as
𝑦(𝑛) ≡ 𝑇[𝑥(𝑛)]
Where, the symbol T denotes the operation on the input signal by the system. Based on the input-
output relation system can be classified into
1. Static versus dynamic systems.
2. Time-invariant versus time-variant systems.
 Prepared by Dr Harish M S

3. Linear versus nonlinear systems.

4. Casual versus noncausal systems.
5. Stable versus unstable systems.
Static versus dynamic systems: A discrete time system is said to static or memoryless if its
present output sample depends on the present input sample, but not on the past or future input
samples. If the output of the system at time n is completely determined by the input samples in
the interval from 𝑛 − 𝑁 to 𝑛 (𝑁 ≥ 0), the system is said to have memory of duration N. If 𝑁 =
0 the system is static. If 0 < 𝑁 < ∞, the system is said to have finite memory, whereas if 𝑁 =
∞, the system is said to have infinite memory.
Consider the two different systems represented by input relations
1. 𝑦(𝑛) = 𝑎𝑥(𝑛)
2. 𝑦(𝑛) = 𝑛𝑥(𝑛) + 𝑏𝑥 2 (𝑛)
Are both static and memoryless systems, note that there is no need to store any of the past inputs
or past outputs in order to compute the present output.
Let us consider the another three systems represented by input output relations
1. 𝑦(𝑛) = 𝑥(𝑛) + 3𝑥(𝑛 − 1)
2. 𝑦(𝑛) = ∑𝑛𝑘=0 𝑥(𝑛 − 𝑘)
3. 𝑦(𝑛) = ∑∞
𝑘=0 𝑥(𝑛 − 𝑘)

All the above three systems are dynamic systems, first two systems are finite memory, third
system is dynamic system with infinite memory.
Problem: Determine, whether the following systems represented by input output relations are
static or dynamic.
a) 𝑦(𝑛) = cos[𝑥 (𝑛)]
b) 𝑦(𝑛) = ∑𝑛+1
𝑘=−∞ 𝑥(𝑘)

c) 𝑦(𝑛) = 𝑥(𝑛)cos(𝜔0 𝑛)
d) 𝑦(𝑛) = 𝑥(−𝑛 + 2)
e) 𝑦(𝑛) = |𝑥(𝑛)|
f) 𝑦(𝑛) = 𝑥(𝑛)𝑢(𝑛)
g) 𝑦(𝑛) = 𝑥(𝑛) + 𝑛𝑥(𝑛 + 1)
h) 𝑦(𝑛) = 𝑥(2𝑛)
i) 𝑦(𝑛) = 𝑥(−𝑛)
 Prepared by Dr Harish M S

Solution:
a) 𝑦(𝑛) = cos[𝑥 (𝑛)]
Present output depends on the present input, not on the past or future input values, system is
static.
b) 𝑦(𝑛) = ∑𝑛+1
𝑘=−∞ 𝑥(𝑘)
𝑛+1

𝑦(𝑛) = ∑ 𝑥(𝑘) = ⋯ … . 𝑥(𝑛 − 3) + 𝑥(𝑛 − 2) + 𝑥(𝑛 − 1) + 𝑥(𝑛) + 𝑥(𝑛 + 1)

𝑘=−∞

Present output depends on the past or future input values, system has a memory.
c) 𝑦(𝑛) = 𝑥(𝑛)cos(𝜔0 𝑛)
Present output depends on the present input, not on the past or future input values, system is
static.
d) 𝑦(𝑛) = 𝑥(−𝑛 + 2)
Present output value depends on the future input values, system has a memory.
e) 𝑦(𝑛) = |𝑥(𝑛)|
Present output values depends on the magnitude of the present input value, not on the magnitude
of the past or future input values, system is static.
f) 𝑦(𝑛) = 𝑥(𝑛)𝑢(𝑛)
Present output depends on the present input, not on the past or future input values, system is
static.
g) 𝑦(𝑛) = 𝑥(𝑛) + 𝑛𝑥(𝑛 + 1)
Present output depends on the present input, past input values, system is dynamic.
h) 𝑦(𝑛) = 𝑥(2𝑛)
Present output depends on the present input, future input values, system is dynamic.
i) 𝑦(𝑛) = 𝑥(−𝑛)
Present output depends on the present input, future input values, system is dynamic.
Time-invariant versus time-variant systems: The time- invariant system is also known as shift
invariant system. It defined as, if the input samples are shifted by k samples, there is an identical
shift in the output samples then the system is known as time-invariant system otherwise time-
variant system.
 Prepared by Dr Harish M S

To identify the time-invariant system, let us consider an arbitrary input 𝑥(𝑛) applied to the
system, and the corresponding output is 𝑦(𝑛) represented as 𝑦(𝑛) = 𝑇{𝑥(𝑛)}. If this output is
time shifted by k samples represented as 𝑦(𝑛 − 𝑘) as shown in figure 1.36.
If the same input 𝑥(𝑛) is time shifted by k samples denoted as 𝑥(𝑛 − 𝑘) is applied to the same
system, and the corresponding output is represented as 𝑦(𝑛, 𝑘). If 𝑦(𝑛, 𝑘) and 𝑦(𝑛 − 𝑘) are
identical, then the system is known as time –invariant system.
Problem: Determine, whether the following systems represented by input output relations are
time invariant or time varying.
a) 𝑦(𝑛) = cos[𝑥 (𝑛)]
b) 𝑦(𝑛) = ∑𝑛+1
𝑘=−∞ 𝑥(𝑘)

c) 𝑦(𝑛) = 𝑥(𝑛)cos(𝜔0 𝑛)
d) 𝑦(𝑛) = 𝑥(−𝑛 + 2)
e) 𝑦(𝑛) = |𝑥(𝑛)|
f) 𝑦(𝑛) = 𝑥(𝑛)𝑢(𝑛)
g) 𝑦(𝑛) = 𝑥(𝑛) + 𝑛𝑥(𝑛 + 1)
h) 𝑦(𝑛) = 𝑥(2𝑛)
i) 𝑦(𝑛) = 𝑥(−𝑛)
j) 𝑦(𝑛) = 𝑥(𝑛) − 𝑥(𝑛 − 1)
k) 𝑦(𝑛) = 𝑛𝑥(𝑛)
Solution:
a) 𝑦(𝑛) = cos[𝑥 (𝑛)]
 Prepared by Dr Harish M S

Since 𝑦(𝑛, 𝑘) and 𝑦(𝑛 − 𝑘) are identical, the system is time invariant system.
b) 𝑦(𝑛) = ∑𝑛+1 𝑛+1
𝑘=−∞ 𝑥 (𝑘 ) = ∑𝑘𝑜 =−∞ 𝑥(𝑘𝑜 )
𝑛+1

𝑦(𝑛, 𝑘) = ∑ 𝑥(𝑘0 − 𝑘) … . (1)

𝑘0 =−∞

𝑦(𝑛, 𝑘) = ⋯ … + 𝑥(𝑛 − 𝑘) + 𝑥(𝑛 + 1 − 𝑘)

𝑛−𝑘+1

𝑦(𝑛 − 𝑘) = ∑ 𝑥 (𝑘) … . . (2)

𝑘=−∞

𝑦(𝑛 − 𝑘) = ⋯ + 𝑥(𝑛 − 𝑘) + 𝑥(𝑛 − 𝑘 + 1)

Equation 1 and 2 are identical, system is time-invariant system.
c) 𝑦(𝑛) = 𝑥(𝑛)cos(𝜔0 𝑛)
𝑦(𝑛, 𝑘) = 𝑥(𝑛 − 𝑘)𝑐𝑜𝑠(𝜔0 𝑛) … . (1)
𝑦(𝑛 − 𝑘) = 𝑥 (𝑛 − 𝑘) cos(𝜔0 [𝑛 − 𝑘]) … . (2)
Equation 1 and 2 are not identical, system is time variant system.
d) 𝑦(𝑛) = 𝑥(−𝑛 + 2)
𝑦(𝑛, 𝑘) = 𝑥(−𝑛 + 𝑘 + 2) … … . (1)
𝑦(𝑛 − 𝑘) = 𝑥 (−𝑛 + 𝑘 + 2) … … (2)
Equation 1 and 2 are identical, system is time-invariant system.
e) 𝑦(𝑛) = |𝑥(𝑛)|
𝑦(𝑛, 𝑘) = |𝑥(𝑛 − 𝑘)| … … . (1)
𝑦(𝑛 − 𝑘) = |𝑥(𝑛 − 𝑘)| … … (2)
Equation 1 and 2 are identical, system is time-invariant system.
f) 𝑦(𝑛) = 𝑥(𝑛)𝑢(𝑛)
𝑦(𝑛, 𝑘) = 𝑥 (𝑛 − 𝑘)𝑢(𝑛) … … . (1)
𝑦(𝑛 − 𝑘) = 𝑥 (𝑛 − 𝑘)𝑢(𝑛 − 𝑘) … … (2)
Equation 1 and 2 are not identical, system is time variant system.
 Prepared by Dr Harish M S

g) 𝑦(𝑛) = 𝑥(𝑛) + 𝑛𝑥(𝑛 + 1)

𝑦(𝑛, 𝑘) = 𝑥(𝑛 − 𝑘) + 𝑛𝑥(𝑛 − 𝑘 + 1) … … . (1)
𝑦(𝑛 − 𝑘) = 𝑥 (𝑛 − 𝑘) + (𝑛 − 𝑘)𝑥(𝑛 − 𝑘 + 1)) … … (2)
Equation 1 and 2 are not identical, system is time variant system.
h) 𝑦(𝑛) = 𝑥(2𝑛)
𝑦(𝑛, 𝑘) = 𝑥(2𝑛 − 2𝑘) … … . (1)
𝑦(𝑛 − 𝑘) = 𝑥 (2𝑛 − 2𝑘) … … (2)
Equation 1 and 2 are identical, system is time-invariant system.
i) 𝑦(𝑛) = 𝑥(−𝑛)
𝑦(𝑛, 𝑘) = 𝑥(−𝑛 − 𝑘) … … . (1)
𝑦(𝑛 − 𝑘) = 𝑥 (−𝑛 + 𝑘) … … (2)
Equation 1 and 2 are not identical, system is time variant system.
j) 𝑦(𝑛) = 𝑥(𝑛) − 𝑥(𝑛 − 1)

𝑦(𝑛, 𝑘) = 𝑥(𝑛 − 𝑘) − 𝑥(𝑛 − 𝑘 − 1) … . (1)

𝑦(𝑛 − 𝑘) = 𝑥 (𝑛 − 𝑘) − 𝑥(𝑛 − 𝑘 − 1) … … (2)
Equation 1 and 2 are identical, system is time-invariant system.
k) 𝑦(𝑛) = 𝑛𝑥(𝑛)
𝑦(𝑛, 𝑘) = 𝑛𝑥(𝑛 − 𝑘) … . (1)
𝑦(𝑛 − 𝑘) = (𝑛 − 𝑘)𝑥(𝑛 − 𝑘) … … (2)
Equation 1 and 2 are not identical, system is time variant system.
Linear versus nonlinear systems: A linear system should satisfy the superposition principal.
The superposition principal stated that “response of the system due to weight sum of input signal
is equal to weighted sum of response of the system” i.e.
⏟[𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)]
𝑇 =𝑎
⏟1 T[x1 (n)] + a2 T[x2 (n)]
𝑅𝑒𝑠𝑝𝑜𝑛𝑠𝑒 𝑑𝑢𝑒 𝑡𝑜 𝑤𝑒𝑖𝑔ℎ𝑡𝑒𝑑 𝑠𝑢𝑚 𝑜𝑓 𝑖𝑛𝑝𝑢𝑡 Weighted sum of response

Where 𝑎1 and 𝑎2 are any arbitrary constants, 𝑥1 (𝑛) and 𝑥2 (𝑛) are any arbitrary input to the same
system. Figure 1.37 gives a pictorial representation of superposition principal.
 Prepared by Dr Harish M S

Problem: Determine, whether the following systems represented by input output relations are
linear or nonlinear.
a) 𝑦(𝑛) = cos[𝑥 (𝑛)]
b) 𝑦(𝑛) = ∑𝑛+1
𝑘=−∞ 𝑥(𝑘)

c) 𝑦(𝑛) = 𝑥(𝑛)cos(𝜔0 𝑛)
d) 𝑦(𝑛) = 𝑥(−𝑛 + 2)
e) 𝑦(𝑛) = |𝑥(𝑛)|
f) 𝑦(𝑛) = 𝑥(𝑛)𝑢(𝑛)
g) 𝑦(𝑛) = 𝑥(𝑛) + 𝑛𝑥(𝑛 + 1)
h) 𝑦(𝑛) = 𝑥(2𝑛)
i) 𝑦(𝑛) = 𝑥(−𝑛)
j) 𝑦(𝑛) = 𝑥(𝑛) − 𝑥(𝑛 − 1)
k) 𝑦(𝑛) = 𝑛𝑥(𝑛)
l) 𝑦(𝑛) = 𝑥(𝑛2 )
m) 𝑦(𝑛) = 𝑥 2 (n)
n) 𝑦(𝑛) = 𝐴𝑥(𝑛) + 𝐵
o) 𝑦(𝑛) = 𝑒 𝑥(𝑛)
Solution:
a) 𝑦(𝑛) = cos[𝑥 (𝑛)]
𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} = cos[a1 𝑥1 (𝑛) + a2 𝑥2 (𝑛)] … (1)
𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)} = 𝑎1 cos[𝑥1 (𝑛)] + 𝑎2 cos[𝑥2 (𝑛)] … … (2)
Equation 1 and 2 are not same, system is non-linear.
b) 𝑦(𝑛) = ∑𝑛+1
𝑘=−∞ 𝑥(𝑘)
 Prepared by Dr Harish M S

n+1

𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} = ∑ {𝑎1 𝑥1 (𝑘) + 𝑎2 𝑥2 (𝑘)} … (1)

k=−∞
𝑛+1 𝑛+1

𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)} = ∑ 𝑎1 𝑥1 (𝑘) + ∑ 𝑎2 𝑥2 (𝑘) … … . (2)

𝑘=−∞ 𝑘=−∞

Equation 1 and 2 are same, system is linear.

c) 𝑦(𝑛) = 𝑥(𝑛)cos(𝜔0 𝑛)
𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} = [𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)]𝑐𝑜𝑠𝜔𝑜 𝑛 … (1)
𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)} = 𝑎1 𝑥1 (𝑛)𝑐𝑜𝑠𝜔𝑜 𝑛 + 𝑎2 𝑥2 (𝑛)𝑐𝑜𝑠𝜔𝑜 𝑛 … … . . (2)
Equation 1 and 2 are same, system is linear.
d) 𝑦(𝑛) = 𝑥(−𝑛 + 2)
𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} = 𝑎1 𝑥1 (−𝑛 + 2) + 𝑎2 𝑥2 (−𝑛 + 2) … . . (1)
𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)} = 𝑎1 𝑥1 (−𝑛 + 2) + 𝑎2 𝑥2 (−𝑛 + 2) … . . (2)
Equation 1 and 2 are same, system is linear.
e) 𝑦(𝑛) = |𝑥(𝑛)|
𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} = |𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)| … . . (1)
𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)} = |𝑎1 𝑥1 (𝑛)| + |𝑎2 𝑥2 (𝑛)| … . . (2)
Equation 1 and 2 are not same, system is non-linear
f) 𝑦(𝑛) = 𝑥(𝑛)𝑢(𝑛)
𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} = {𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)}𝑢(𝑛) … … (1)
𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)} = 𝑎1 𝑥1 (𝑛)𝑢(𝑛) + 𝑎2 𝑥2 (𝑛)𝑢(𝑛) … . . (2)
Equation 1 and 2 are same, system is linear
g) 𝑦(𝑛) = 𝑥(𝑛) + 𝑛𝑥(𝑛 + 1)
𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)}
= 𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛) + 𝑛[𝑎1 𝑥1 (𝑛 + 1)
+ 𝑎2 𝑥2 (𝑛 + 1)] … … (1)
𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)}
= 𝑎1 𝑥1 (𝑛) + 𝑛𝑎1 𝑥1 (𝑛 + 1) + 𝑎2 𝑥2 (𝑛)
+ 𝑛𝑎2 𝑥2 (𝑛 + 1) … . . (2)
Equation 1 and 2 are same, system is linear
h) 𝑦(𝑛) = 𝑥(2𝑛)
 Prepared by Dr Harish M S

𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} = 𝑎1 𝑥1 (2𝑛) + 𝑎2 𝑥2 (2𝑛) … (1)

𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)} = 𝑎1 𝑥1 (2𝑛) + 𝑎2 𝑥2 (2𝑛) … . . (2)
Equation 1 and 2 are same, system is linear
i) 𝑦(𝑛) = 𝑥(−𝑛)
𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} = 𝑎1 𝑥1 (−𝑛) + 𝑎2 𝑥2 (−𝑛) … (1)
𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)} = 𝑎1 𝑥1 (−𝑛) + 𝑎2 𝑥2 (−𝑛) … . . (2)
Equation 1 and 2 are same, system is linear
j) 𝑦(𝑛) = 𝑥(𝑛) − 𝑥(𝑛 − 1)
𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)}
= 𝑎1 𝑥1 (𝑛)−𝑎1 𝑥1 (𝑛 − 1) + 𝑎2 𝑥2 (𝑛) − 𝑎2 𝑥2 (𝑛 − 1) … (1)
𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)}
= 𝑎1 𝑥1 (𝑛)−𝑎1 𝑥1 (𝑛 − 1) + 𝑎2 𝑥2 (𝑛) − 𝑎2 𝑥2 (𝑛 − 1) … . . (2)
Equation 1 and 2 are same, system is linear
k) 𝑦(𝑛) = 𝑛𝑥(𝑛)
𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} = 𝑛{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} … . . (1)
𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)} = 𝑛𝑎1 𝑥1 (𝑛) + 𝑛𝑎2 𝑥2 (𝑛) … … (2)
Equation 1 and 2 are same, system is linear
l) 𝑦(𝑛) = 𝑥(𝑛2 )
𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} = 𝑎1 𝑥1 (𝑛2 ) + 𝑎2 𝑥2 (𝑛2 ) … . . (1)
𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)} = 𝑎1 𝑥1 (𝑛2 ) + 𝑎2 𝑥2 (𝑛2 ) … … (2)
Equation 1 and 2 are same, system is linear
m) 𝑦(𝑛) = 𝑥 2 (n)
𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} = [𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)]2 … . . (1)
𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)} = 𝑎1 𝑥1 2 (𝑛) + 𝑎2 𝑥2 2 (𝑛) … … (2)
Equation 1 and 2 are not same, system is nonlinear
n) 𝑦(𝑛) = 𝐴𝑥(𝑛) + 𝐵
𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} = 𝐴[𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)] + 𝐵 … . . (1)
𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)} = 𝐴𝑎1 𝑥1 (𝑛) + 𝑎1 𝐵 + 𝑎2 𝑥2 (𝑛) + 𝑎2 𝐵 … … (2)
Equation 1 and 2 are not same, system is nonlinear
o) 𝑦(𝑛) = 𝑒 𝑥(𝑛)
𝑇{𝑎1 𝑥1 (𝑛) + 𝑎2 𝑥2 (𝑛)} = 𝑒 𝑎1𝑥1 (𝑛)+𝑎2𝑥2 (𝑛) … . . (1)
 Prepared by Dr Harish M S

𝑇{𝑎1 𝑥1 (𝑛)} + 𝑇{𝑎2 𝑥2 (𝑛)} = 𝑎1 𝑒 𝑥1(𝑛) + 𝑎2 𝑒 𝑥2 (𝑛) … … (2)

Equation 1 and 2 are not same, system is nonlinear
Casual versus noncausal systems: A system is said to be causal, if the output of the system at
any time n depends only on present and past inputs, but not on the future inputs. Mathematically
𝑦(𝑛) = 𝑇[𝑥(𝑛), 𝑥(𝑛 − 1), 𝑥(𝑛 − 2), … . . ]
If the system does not satisfy this condition, then the system is known as noncausal system.
Problem: Determine, whether the following systems represented by input output relations are
causal or noncausal.
a) 𝑦(𝑛) = cos[𝑥 (𝑛)]
b) 𝑦(𝑛) = ∑𝑛+1
𝑘=−∞ 𝑥(𝑘)

c) 𝑦(𝑛) = 𝑥(𝑛)cos(𝜔0 𝑛)
d) 𝑦(𝑛) = 𝑥(−𝑛 + 2)
e) 𝑦(𝑛) = |𝑥(𝑛)|
f) 𝑦(𝑛) = 𝑥(𝑛)𝑢(𝑛)
g) 𝑦(𝑛) = 𝑥(𝑛) + 𝑛𝑥(𝑛 + 1)
h) 𝑦(𝑛) = 𝑥(2𝑛)
i) 𝑦(𝑛) = 𝑥(−𝑛)
j) 𝑦(𝑛) = 𝑥(𝑛) − 𝑥(𝑛 − 1)
k) 𝑦(𝑛) = 𝑛𝑥(𝑛)
l) 𝑦(𝑛) = 𝑥(𝑛2 )
m) 𝑦(𝑛) = 𝑥 2 (n)
n) 𝑦(𝑛) = 𝐴𝑥(𝑛) + 𝐵
o) 𝑦(𝑛) = 𝑒 𝑥(𝑛)
Solution:
a) 𝑦(𝑛) = cos[𝑥 (𝑛)]
Since the output depends on current input, not on the future input, system is causal.
b) 𝑦(𝑛) = ∑𝑛+1
𝑘=−∞ 𝑥(𝑘)
𝑛+1

𝑦(𝑛) = ∑ 𝑥(𝑘) = ⋯ … . 𝑥(𝑛 − 2) + 𝑥(𝑛 − 1) + 𝑥(𝑛) + 𝑥(𝑛 + 1)

𝑘=−∞
 Prepared by Dr Harish M S

Since the output depends on current input, past input, as well as future inputs, system is
noncausal.
c) 𝑦(𝑛) = 𝑥(𝑛)cos(𝜔0 𝑛)
Since the output depends on current input, not on the future input, system is causal.
d) 𝑦(𝑛) = 𝑥(−𝑛 + 2)
Since the output depends on future inputs, system is noncausal.
e) 𝑦(𝑛) = |𝑥(𝑛)|
Since the output depends on current input, not on the future input, system is causal.
f) 𝑦(𝑛) = 𝑥(𝑛)𝑢(𝑛)
Since the output depends on current input, not on the future input, system is causal.
g) 𝑦(𝑛) = 𝑥(𝑛) + 𝑛𝑥(𝑛 + 1)
Since the output depends on future input, system is noncausal.
h) 𝑦(𝑛) = 𝑥(2𝑛)
Since the output depends on current, past as well as future input, system is noncausal.
i) 𝑦(𝑛) = 𝑥(−𝑛)

Since the output depends on current, past as well as future input, system is noncausal.
j) 𝑦(𝑛) = 𝑥(𝑛) − 𝑥(𝑛 − 1)
Since the output depends on current, past input, system is causal.
k) 𝑦(𝑛) = 𝑛𝑥(𝑛)
Since the output depends on current input, not on the future input, system is causal.
l) 𝑦(𝑛) = 𝑥(𝑛2 )
Since the output depends on the future input, system is noncausal.
m) 𝑦(𝑛) = 𝑥 2 (n)
Since the output depends on current input, not on the future input, system is causal.
n) 𝑦(𝑛) = ∑𝑛𝑘=−∞ 𝑥(𝑘)
𝑛

𝑦(𝑛) = ∑ 𝑥(𝑘) = ⋯ … . 𝑥(𝑛 − 2) + 𝑥(𝑛 − 1) + 𝑥(𝑛)

𝑘=−∞

Since the output depends on current input, past input, system is causal
Stable versus unstable systems: Stability is an important property that must be considered in
any practical application of a system. An unstable system usually exhibits erratic and extreme
 Prepared by Dr Harish M S

behavior. Stability of a system is defined as, any arbitrary system is said to be bounded input-
bounded output (BIBO) stable if and only if every bounded input produces a bounded output. Let
us consider the input sequence 𝑥(𝑛), if input sequence is bounded, then it has a finite value at all
n. mathematically
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If this bounded input sequence is applied to the system, and the corresponding output sequence
is 𝑦(𝑛). For BIBO stable system, output sequence also bounded, and it has a finite value at all n.
mathematically
|𝑦(𝑛)| ≤ 𝑀𝑦 < ∞, for all n
Problem: Determine, whether the following systems represented by input output relations are
stable or unstable.
a) 𝑦(𝑛) = cos[𝑥 (𝑛)]
b) 𝑦(𝑛) = ∑𝑛+1
𝑘=−∞ 𝑥(𝑘)

c) 𝑦(𝑛) = 𝑥(𝑛)cos(𝜔0 𝑛)
d) 𝑦(𝑛) = 𝑥(−𝑛 + 2)
e) 𝑦(𝑛) = |𝑥(𝑛)|
f) 𝑦(𝑛) = 𝑥(𝑛)𝑢(𝑛)
g) 𝑦(𝑛) = 𝑥(𝑛) + 𝑛𝑥(𝑛 + 1)
h) 𝑦(𝑛) = 𝑥(2𝑛)
i) 𝑦(𝑛) = 𝑥(−𝑛)
j) 𝑦(𝑛) = 𝑥(𝑛) − 𝑥(𝑛 − 1)
k) 𝑦(𝑛) = 𝑛𝑥(𝑛)
l) 𝑦(𝑛) = 𝑥(𝑛2 )
m) 𝑦(𝑛) = 𝑥 2 (n)
n) 𝑦(𝑛) = 𝐴𝑥(𝑛) + 𝐵
o) 𝑦(𝑛) = 𝑒 𝑥(𝑛)
Solution:
a) 𝑦(𝑛) = cos[𝑥 (𝑛)]
Assume the input sequence is bounded, then
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
 Prepared by Dr Harish M S

𝑦(𝑛) = cos[𝑥(𝑛)]
Apply the absolute value function on both the sides to the above equation, we get
|𝑦(𝑛)| = |cos[𝑥(𝑛)]|
For all values of n, 𝑐𝑜𝑠[𝑥 (𝑛)] is between -1 and 1, thus |cos[𝑥(𝑛)]| < 1 for all n, thus system is
BIBO stable.
b) 𝑦(𝑛) = ∑𝑛+1
𝑘=−∞ 𝑥(𝑘)

Assume the input sequence is bounded, then

|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
𝑛+1

|𝑦(𝑛)| = ∑ 𝑥(𝑘)
𝑘=−∞

Apply the absolute value function on both the sides to the above equation, we get
𝑛+1

|𝑦(𝑛)| = | ∑ 𝑥(𝑘)|
𝑘=−∞
𝑛+1

|𝑦(𝑛)| ≤ ∑ |𝑥(𝑘)|
𝑘=−∞
𝑛+1

|𝑦(𝑛)| ≤ ∑ 𝑀𝑥
𝑘=−∞

For finite values of 𝑥(𝑛), summation diverges to infinity, the system output is unbounded, system
is unstable.
b) 𝑦(𝑛) = 𝑥(𝑛)cos(𝜔0 𝑛)
Assume the input sequence is bounded, then
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
𝑦(𝑛) = 𝑥(𝑛)cos(𝜔0 𝑛)
Apply the absolute value function on both the sides to the above equation, we get
|𝑦(𝑛)| = |𝑥(𝑛)cos(𝜔0 𝑛)|
|𝑦(𝑛)| ≤ |𝑥(𝑛)||cos(𝜔0 𝑛)|
|𝑦(𝑛)| ≤ 𝑀𝑥 |cos(𝜔0 𝑛)|
For finite values of 𝑥(𝑛), system output is bounded, system is stable.
 Prepared by Dr Harish M S

c) 𝑦(𝑛) = 𝑥(−𝑛 + 2)
Assume the input sequence is bounded, then
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
𝑦(𝑛) = 𝑥(−𝑛 + 2)
Apply the absolute value function on both the sides to the above equation, we get
|𝑦(𝑛)| = |𝑥(−𝑛 + 2)|
|𝑦(𝑛)| ≤ 𝑀𝑥
For finite values of 𝑥(𝑛), system output is bounded, system is stable.
d) 𝑦(𝑛) = |𝑥(𝑛)|
Assume the input sequence is bounded, then
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
|𝑦(𝑛)| = |𝑥(𝑛)|
|𝑦(𝑛)| ≤ 𝑀𝑥
For finite values of 𝑥(𝑛), system output is bounded, system is stable
e) 𝑦(𝑛) = 𝑥(𝑛)𝑢(𝑛)
Assume the input sequence is bounded, then
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
𝑦(𝑛) = 𝑥 (𝑛)𝑢(𝑛)
|𝑦(𝑛)| = |𝑥(𝑛)𝑢(𝑛)|
|𝑦(𝑛)| ≤ |𝑥(𝑛)||𝑢(𝑛)|
|𝑦(𝑛)| ≤ 𝑀𝑥 × 1
For finite values of 𝑥(𝑛), system output is bounded, system is stable
f) 𝑦(𝑛) = 𝑥(𝑛) + 𝑛𝑥(𝑛 + 1)
Assume the input sequence is bounded, then
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
𝑦(𝑛) = 𝑥(𝑛) + 𝑛𝑥(𝑛 + 1)
Apply the absolute value function on both the sides to the above equation, we get
 Prepared by Dr Harish M S

|𝑦(𝑛)| = |𝑥(𝑛) + 𝑛𝑥(𝑛 + 1)|

|𝑦(𝑛)| ≤ |𝑥(𝑛)| + |𝑛||𝑥(𝑛 − 1|
|𝑦(𝑛)| ≤ 𝑀𝑥 + |𝑛|𝑀𝑥
For finite values of 𝑥(𝑛), as n approaches to infinity, output diverges, system is unstable
e) 𝑦(𝑛) = 𝑥(2𝑛)
Assume the input sequence is bounded, then
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
𝑦(𝑛) = 𝑥(2𝑛)
Apply the absolute value function on both the sides to the above equation, we get
|𝑦(𝑛)| = |𝑥(2𝑛)|
|𝑦(𝑛)| ≤ |𝑥(2𝑛)||
|𝑦(𝑛)| ≤ 𝑀𝑥
For finite values of 𝑥(𝑛), system output is bounded, system is stable
f) 𝑦(𝑛) = 𝑥(−𝑛)
Assume the input sequence is bounded, then
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
𝑦(𝑛) = 𝑥(−𝑛)
Apply the absolute value function on both the sides to the above equation, we get
|𝑦(𝑛)| = |𝑥(−𝑛)|
|𝑦(𝑛)| ≤ |𝑥(−𝑛)||
|𝑦(𝑛)| ≤ 𝑀𝑥
For finite values of 𝑥(𝑛), system output is bounded, system is stable
g) 𝑦(𝑛) = 𝑥(𝑛) − 𝑥(𝑛 − 1)
Assume the input sequence is bounded, then
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
𝑦(𝑛) = 𝑥(𝑛) − 𝑥(𝑛 − 1)
Apply the absolute value function on both the sides to the above equation, we get
|𝑦(𝑛)| = |𝑥(𝑛) − 𝑥(𝑛 − 1)|
 Prepared by Dr Harish M S

|𝑦(𝑛)| ≤ |𝑥(𝑛)| + |−𝑥(𝑛 − 1)|

|𝑦(𝑛)| ≤ 𝑀𝑥 + 𝑀𝑥
For finite values of 𝑥(𝑛), system output is bounded, system is stable
h) 𝑦(𝑛) = 𝑛𝑥(𝑛)
Assume the input sequence is bounded, then
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
𝑦(𝑛) = 𝑛𝑥(𝑛)
Apply the absolute value function on both the sides to the above equation, we get
|𝑦(𝑛)| = |𝑛𝑥(𝑛)|
|𝑦(𝑛)| ≤ |𝑛||𝑥(𝑛)|
|𝑦(𝑛)| ≤ |𝑛| 𝑀𝑥
For finite values of 𝑥(𝑛), as n approaches to infinity, output diverges, system is unstable
i) 𝑦(𝑛) = 𝑥(𝑛2 )
Assume the input sequence is bounded, then
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
𝑦(𝑛) = 𝑥(𝑛2 )
Apply the absolute value function on both the sides to the above equation, we get
|𝑦(𝑛)| = |𝑥(𝑛2 )|
|𝑦(𝑛)| ≤ |𝑥(𝑛2 )|
|𝑦(𝑛)| ≤ 𝑀𝑥
For finite values of 𝑥(𝑛), system output is bounded, system is stable
j) 𝑦(𝑛) = 𝑥 2 (n)
Assume the input sequence is bounded, then
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
𝑦(𝑛) = 𝑥(𝑛2 ) = 𝑥(𝑛)𝑥(𝑛)
Apply the absolute value function on both the sides to the above equation, we get
|𝑦(𝑛)| = |𝑥(𝑛)𝑥(𝑛)|
|𝑦(𝑛)| ≤ |𝑥(𝑛)||𝑥(𝑛)|
 Prepared by Dr Harish M S

|𝑦(𝑛)| ≤ 𝑀𝑥 𝑀𝑥
|𝑦(𝑛)| ≤ 𝑀𝑥2
For finite values of 𝑥(𝑛), system output is bounded, system is stable
k) 𝑦(𝑛) = 𝐴𝑥(𝑛) + 𝐵
Assume the input sequence is bounded, then
|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
𝑦(𝑛) = 𝐴𝑥(𝑛) + 𝐵
Apply the absolute value function on both the sides to the above equation, we get
|𝑦(𝑛)| = |𝐴𝑥(𝑛) + 𝐵|
|𝑦(𝑛)| ≤ 𝐴|𝑥(𝑛)| + |𝐵|
|𝑦(𝑛)| ≤ 𝐴𝑀𝑥 + 𝐵
For finite values of 𝑥(𝑛), system output is bounded, system is stable
l) 𝑦(𝑛) = 𝑒 𝑥(𝑛)

Assume the input sequence is bounded, then

|𝑥(𝑛)| ≤ 𝑀𝑥 < ∞, for all n
If the bounded input is applied to the system, then the corresponding output is
𝑦(𝑛) = 𝑒 𝑥(𝑛)
Apply the absolute value function on both the sides to the above equation, we get
|𝑦(𝑛)| = 𝑒 |𝑥(𝑛)|
As n changes, 𝑒 |𝑥(𝑛)| changes, for large value of|𝑥(𝑛)|, exponential function produces
unbounded output for bounded input, thus the system is unstable.
Analysis of Discrete Time Linear Time Invariant Systems
In this section we analyze the linear time invariant system. If the system is linear and time
invariant, then the system is known as Linear Time Invariant Systems (LTI) system.
Techniques for the Analysis of Linear Systems
There are two basic methods for analyzing the behavior or response of a LTI system to a given
input signal. One method is based on the direct solution of the input-output equation for the
system, which, in general, has the form
 Prepared by Dr Harish M S

𝑦(𝑛) = 𝑇[𝑦(𝑛 − 1), 𝑦(𝑛 − 2), … . , 𝑦(𝑛 − 𝑁), 𝑥(𝑛 − 1), … … . , 𝑥(𝑛 − 𝑀)]
Where 𝑇[. ] denotes the some function of the quantities in the brackets. The general linear
constant coefficient difference equation of the LTI system is
𝑁 𝑀

𝑦(𝑛) = − ∑ 𝑎𝑘 𝑦(𝑛 − 𝑘) + ∑ 𝑏𝑘 𝑥(𝑛 − 𝑘)

𝑘=1 𝑘=0

Where, 𝑥(𝑛) is the input to the system, 𝑦(𝑛) is the output of the system 𝑎𝑘 and 𝑏𝑘 are the
coefficients. This difference equation completely represents the LTI system. If we know the
initial condition of the system and the system input, we can find the output for of the system.
The second method for analyzing the behavior of a LTI system to a given input signal is first
decompose or resolve the input signal into a sum of elementary signals. The elementary signals
are selected so that the response of the system to each signal is easily determined. The using the
linearity property of the system, the response of the system to the elementary signals are added to
obtain the total response of the system to the given input.
Let us resolve the input signal 𝑥(𝑛) into a weighted sum of elementary signal {𝑥𝑘 (𝑛)}, so the
𝑥(𝑛) = ∑𝑘 𝑐𝑘 𝑥𝑘 (𝑛)
Where, {𝑐𝑘 } are the set of amplitudes in the decomposition of the signal 𝑥(𝑛). Now the response
of the system for the elementary signal component 𝑥𝑘 (𝑛)is 𝑦𝑘 (𝑛)
𝑦𝑘 (𝑛) ≡ 𝑇[𝑥𝑘 (𝑛)]
Assume that the system is relaxed and the response to 𝑐𝑘 𝑥𝑘 (𝑛) is𝑐𝑘 𝑦𝑘 (𝑛).
Finally, the total response to the input 𝑥(𝑛) is

𝑦(𝑛) = 𝑌[𝑥(𝑛)] = 𝑇 [∑ 𝑐𝑘 𝑥𝑘 (𝑛)] = ∑ 𝑐𝑘 𝑇[𝑥𝑘 (𝑛)] = ∑ 𝑐𝑘 𝑦𝑘 (𝑛)

𝑘 𝑘 𝑘

Resolutions of Discrete-time signals into Impulses

The unit sample sequence, also called the discrete-time impulse sequence or Dirac delta
sequence, is a fundamental concept in digital signal processing. It is defined as:
1, 𝑛=0
𝛿 (𝑛 ) = {
0, 𝑛≠0
Let us consider an arbitrary input signal 𝑥(𝑛) −∞ ≤𝑛 ≤∞
𝑥(𝑛) = {… … 𝑥(−2), 𝑥(−1), 𝑥(0), 𝑥(1), 𝑥(2), 𝑥(3), … … . }
Now each samples of 𝑥(𝑛) can be represented by impulse function, that is
 Prepared by Dr Harish M S

𝑥(0) = 𝑥(0)𝛿(𝑛)
𝑥(1) = 𝑥(1)𝛿(𝑛 − 1)
𝑥(−1) = 𝑥(−1)𝛿(𝑛 + 1)
𝑥(2) = 𝑥(2)𝛿(𝑛 − 2)
𝑥(−2) = 𝑥(−2)𝛿(𝑛 + 2)
.
.
∴ 𝑥(𝑛) = {… … + 𝑥(𝑛)𝛿(𝑛 + 2) + 𝑥(𝑛)𝛿(𝑛 + 1) + 𝑥(𝑛)𝛿(𝑛) + 𝑥(𝑛)𝛿(𝑛 − 1) + 𝑥(𝑛)𝛿(𝑛 − 2)
+ ⋯.}
∞

𝑥 (𝑛) = ∑ 𝑥(𝑘)𝛿(𝑛 − 𝑘) −∞≤𝑛 ≤∞

𝑘=−∞

We note that, input signal 𝑥(𝑛) is resolved into weighted sum of unit samples sequences. Thus
any signal can be decomposed or resolved into sum of the scaled shifted impulse function.
Example: Consider the finite duration sequence given as
2, 4, 0, 3
𝑥 (𝑛 ) = { }
↑
𝑥(−1) = 2, 𝑥(0) = 4, 𝑥(1) = 0, 𝑥(2) = 3
∞ 2

𝑥(𝑛) = ∑ 𝑥(𝑘 )𝛿(𝑛 − 𝑘) = ∑ 𝑥(𝑘)𝛿(𝑛 − 𝑘) −1 ≤𝑛 ≤2

𝑘=−∞ 𝑘=−1

𝑥(𝑛) = 𝑥(−1)𝛿(𝑛 + 1) + 𝑥(0)𝛿(𝑛) + 𝑥(1)𝛿(𝑛 − 1) + 𝑥(2)𝛿(𝑛 − 2)

𝑥(𝑛) = 2𝛿(𝑛 + 1) + 4𝛿 (𝑛) + 0𝛿 (𝑛 − 1) + 3𝛿(𝑛 − 2)
𝑥(𝑛) = 2𝛿(𝑛 + 1) + 4𝛿 (𝑛) + 3𝛿(𝑛 − 2)
Response of LTI Systems to Arbitrary inputs: The convolution sum
Having resolved an arbitrary input signal 𝑥(𝑛) into a sum of impulses, we can determine the
output 𝑦(𝑛) of relaxed LTI system for any input signal. Let ℎ(𝑛) be a response of the LTI system
for the unit impulse input 𝛿(𝑛). And ℎ(𝑛 − 𝑘)be a response of the LTI system for the shifted unit
impulse input 𝛿(𝑛 − 𝑘) as shown in figure 1.38. The term ℎ(𝑛) or ℎ(𝑛 − 𝑘 ) are known as
impulse response of the LTI system, it is the response due to impulse input. Mathematically this
can be expressed as
ℎ(𝑛) = 𝑇[𝛿 (𝑛)]
ℎ(𝑛 − 𝑘) = 𝑇[𝛿 (𝑛 − 𝑘)]
 Prepared by Dr Harish M S

For the same LTI system, if we apply input 𝑥(𝑛) = ∑∞

𝑘=−∞ 𝑥 (𝑘 )𝛿(𝑛 − 𝑘), then the

corresponding output is 𝑦(𝑛) as shown in figure 1.39.

∴ 𝑦(𝑛) = 𝑇[𝑥(𝑛)] = 𝑇[ ∑ 𝑥(𝑘)𝛿 (𝑛 − 𝑘)]

𝑘=−∞

If the system is linear, then

∞

𝑦(𝑛) = ∑ 𝑥 (𝑘)𝑇[𝛿(𝑛 − 𝑘)]

𝑘=−∞
∞

𝑦 (𝑛 ) = ∑ 𝑥 (𝑘 )ℎ(𝑛 − 𝑘 ) −∞≤𝑛≤∞
𝑘=−∞

𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛)

This expression gives the response 𝑦(𝑛) of the LTI system for the arbitrary input 𝑥(𝑛) called
convolution sum. We can say that the input 𝑥(𝑛) is convolved with impulse response ℎ(𝑛)to
yield the output 𝑦(𝑛). For 𝑦(𝑛) computation, following four steps are involved.
1. Folding: ℎ(𝑘) is obtained from ℎ(𝑛) by replacing 𝑛 by 𝑘. Fold ℎ(𝑘) about
𝑘 = 0 to obtain ℎ(−𝑘).
2. Shifting: shift ℎ(−𝑘) by n to right or left to obtain ℎ(𝑛 − 𝑘).
3. Multiplication: Multiply 𝑥(𝑘) and ℎ(𝑛 − 𝑘) obtain the product sequence.
4. Summation: sum of all the values of product sequence to obtain the output
sequence.
Explanation: To compute 𝑦(0)
∞

𝑦(0) = ∑ 𝑥(𝑘)ℎ(−𝑘)
𝑘=−∞
 Prepared by Dr Harish M S

The input 𝑥(𝑛) is changed to 𝑥(𝑘) by substituting 𝑛 = 𝑘. The signal ℎ(𝑘) is obtained from ℎ(𝑛)
by substituting 𝑛 = 𝑘 and fold ℎ(𝑘) we get ℎ(−𝑘) . And multiply 𝑥(𝑘) and ℎ(−𝑘) sample by
sample and added together we get 𝑦(0)
To compute the output sample 𝑦(1), 𝑦(1) = ∑∞
𝑘=−∞ 𝑥 (𝑘 )ℎ (1 − 𝑘 )

Shift ℎ(−𝑘) one sample towards right we get ℎ(1 − 𝑘). And multiply 𝑥(𝑘) and ℎ(1 − 𝑘)
sample by sample and added together we get 𝑦(1) .
This process is continued till all the output samples are computed. This computation is called
convolution sum, and ℎ(𝑘) is shifted left or right linearly, thus this convolution is also known as
linear convolution. The symbol ∗ represents the linear convolution
1, 2, 1, −1
Problem: The impulse response of the LTI system is ℎ(𝑛) = ( ). Determine the
↑
1, 2, 3, 1
response of system to the input signal 𝑥(𝑛) = ( ).
↑
Solution: The LTI as shown in below figure.

𝑦 (𝑛 ) = 𝑥 (𝑛 ) ∗ ℎ(𝑛 ) = ∑ 𝑥 (𝑘 )ℎ (𝑛 − 𝑘 ) −∞≤𝑛≤∞
𝑘=−∞

1, 2, 3, 1 1, 2, 1, −1
𝑥 (𝑛 ) = ( ) 0 ≤ 𝑛 ≤ 3 , and ℎ(𝑛) = ( ) −1≤𝑛 ≤2
↑ ↑
1, 2, 3, 1 1, 2, 1, −1
𝑥 (𝑘 ) = ( ) 0 ≤ 𝑘 ≤ 3 , and ℎ(𝑘) = ( ) −1 ≤𝑘 ≤2
↑ ↑
There are different methods are there to compute the output 𝑦(𝑛)
a. Analytical method
∞

𝑦( 𝑛 ) = ∑ 𝑥 ( 𝑘 ) ℎ ( 𝑛 − 𝑘 ) −∞ ≤𝑛 ≤ ∞
𝑘=−∞

Let the number of samples in 𝑥(𝑛) is 𝑁1 = 4 and number of samples in ℎ(𝑛) is 𝑁2 = 4

Total number of samples in 𝑦(𝑛) is given by 𝑁 = 𝑁1 + 𝑁2 − 1 = 4 + 4 − 1 = 7, thus, there are
seven samples in 𝑦(𝑛).
Starting point of 𝑦(𝑛) is 𝑦𝑙 = 𝑥𝑙 +ℎ𝑙 = 0 + (−1) = −1.
Terminating point of 𝑦(𝑛) is 𝑦ℎ = 𝑥ℎ +ℎℎ = 3 + 2 = 5.
 Prepared by Dr Harish M S

𝑦(−1), 𝑦(0), 𝑦(1), 𝑦(2), 𝑦(3), 𝑦(4), 𝑦(5)

The seven samples of 𝑦(𝑛) are 𝑦(𝑛) = ( )
↑
3

∴ 𝑦 ( 𝑛 ) = ∑ 𝑥 (𝑘 )ℎ(𝑛 − 𝑘 ) −1 ≤𝑛 ≤5
𝑘=−1

𝑦(𝑛) = 𝑥(−1)ℎ(𝑛 + 1) + 𝑥(0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2) + 𝑥(3)ℎ(𝑛 − 3) , −1 ≤𝑛 ≤5

𝑦(−1) = 𝑥(−1)ℎ(−1 + 1) + 𝑥(0)ℎ(−1) + 𝑥(1)ℎ(−1 − 1) + 𝑥(2)ℎ(−1 − 2) + 𝑥(3)ℎ(−1 − 3)
𝑦(−1) = 𝑥(−1)ℎ(0) + 𝑥(0)ℎ(−1) + 𝑥(1)ℎ(−2) + 𝑥(2)ℎ(−3) + 𝑥(3)ℎ(−4)
𝑦(−1) = 0 × 2 + 1 × 1 + 2 × 0 + 3 × 0 + 1 × 0 = 1
𝑦(0) = 𝑥(−1)ℎ(0 + 1) + 𝑥(0)ℎ(0) + 𝑥(1)ℎ(0 − 1) + 𝑥(2)ℎ(0 − 2) + 𝑥(3)ℎ(0 − 3)
𝑦(0) = 𝑥(−1)ℎ(1) + 𝑥(0)ℎ(0) + 𝑥(1)ℎ(−1) + 𝑥(2)ℎ(−2) + 𝑥(3)ℎ(−3)
𝑦(0) = 0 × 1 + 1 × 2 + 2 × 1 + 3 × 0 + 1 × 0 = 4
𝑦(1) = 𝑥(−1)ℎ(1 + 1) + 𝑥(0)ℎ(1) + 𝑥(1)ℎ(1 − 1) + 𝑥(2)ℎ(1 − 2) + 𝑥(3)ℎ(1 − 3)
𝑦(1) = 𝑥(−1)ℎ(2) + 𝑥(0)ℎ(1) + 𝑥(1)ℎ(0) + 𝑥(2)ℎ(−1) + 𝑥(3)ℎ(−2)
𝑦(1) = 0 × −1 + 1 × 1 + 2 × 2 + 3 × 1 + 1 × 0 = 8
𝑦(2) = 𝑥(−1)ℎ(2 + 1) + 𝑥(0)ℎ(2) + 𝑥(1)ℎ(2 − 1) + 𝑥(2)ℎ(2 − 2) + 𝑥(3)ℎ(2 − 3)
𝑦(2) = 𝑥(−1)ℎ(3) + 𝑥(0)ℎ(2) + 𝑥(1)ℎ(1) + 𝑥(2)ℎ(0) + 𝑥(3)ℎ(−1)
𝑦(2) = 0 × 0 + 1 × −1 + 2 × 1 + 3 × 2 + 1 × 1 = 8
𝑦(3) = 𝑥(−1)ℎ(3 + 1) + 𝑥(0)ℎ(3) + 𝑥(1)ℎ(3 − 1) + 𝑥(2)ℎ(3 − 2) + 𝑥(3)ℎ(3 − 3)
𝑦(3) = 0 × 0 + 1 × 0 + 2 × −1 + 3 × 1 + 1 × 2 = 3
𝑦(4) = 𝑥(−1)ℎ(4 + 1) + 𝑥(0)ℎ(4) + 𝑥(1)ℎ(4 − 1) + 𝑥(2)ℎ(4 − 2) + 𝑥(3)ℎ(4 − 3)
𝑦(4) = 𝑥(−1)ℎ(5) + 𝑥(0)ℎ(4) + 𝑥(1)ℎ(3) + 𝑥(2)ℎ(2) + 𝑥(3)ℎ(1)
𝑦(4) = 0 × 0 + 1 × 0 + 2 × 0 + 3 × −1 + 1 × 1 = −2
𝑦(5) = 𝑥(−1)ℎ(5 + 1) + 𝑥(0)ℎ(5) + 𝑥(1)ℎ(5 − 1) + 𝑥(2)ℎ(5 − 2) + 𝑥(3)ℎ(5 − 3)
𝑦(3) = 𝑥(−1)ℎ(6) + 𝑥(0)ℎ(5) + 𝑥(1)ℎ(4) + 𝑥(2)ℎ(3) + 𝑥(3)ℎ(2)
𝑦(3) = 0 × 0 + 1 × 0 + 2 × 0 + 3 × 0 + 1 × −1 = −1
𝑦(−1), 𝑦(0), 𝑦(1), 𝑦(2), 𝑦(3), 𝑦(4), 𝑦(5) 1, 4, 8, 8, 3, −2, −1
𝑦 (𝑛 ) = ( )=( )
↑ ↑
b. Graphical method
∞

𝑦( 𝑛 ) = ∑ 𝑥 ( 𝑘 ) ℎ ( 𝑛 − 𝑘 ) −∞ ≤𝑛 ≤ ∞
𝑘=−∞

Let the number of samples in 𝑥(𝑛) is 𝑁1 = 4 and number of samples in ℎ(𝑛) is 𝑁2 = 4

Total number of samples in 𝑦(𝑛) is given by 𝑁 = 𝑁1 + 𝑁2 − 1 = 4 + 4 − 1 = 7, thus, there are
seven samples in 𝑦(𝑛).
Starting point of 𝑦(𝑛) is 𝑦𝑙 = 𝑥𝑙 +ℎ𝑙 = 0 + (−1) = −1.
 Prepared by Dr Harish M S

Terminating point of 𝑦(𝑛) is 𝑦ℎ = 𝑥ℎ +ℎℎ = 3 + 2 = 5.

𝑦(−1), 𝑦(0), 𝑦(1), 𝑦(2), 𝑦(3), 𝑦(4), 𝑦(5)
The seven samples of 𝑦(𝑛) are 𝑦(𝑛) = ( )
↑
 Prepared by Dr Harish M S

a. To compute 𝑦(0), refer figure 1 and figure 2, note that two samples are overlapped,
therefore
𝑦 (0) = 1 × 2 + 2 × 1 = 4
b. To compute 𝑦(−1), refer figure 1 and figure 3, note that only one sample is
overlapped, therefore
𝑦(−1) = 1 × 1 = 1
c. To compute 𝑦(1), refer figure 1 and figure 4, note that three samples are overlapped,
therefore
𝑦 (1) = 1 × 1 + 2 × 2 + 1 × 3 = 8
d. To compute 𝑦(2), refer figure 1 and figure 5, note that four samples are overlapped,
therefore
𝑦(2) = 1 × −1 + 1 × 2 + 2 × 3 + 1 × 1 = 8
e. To compute 𝑦(3), refer figure 1 and figure 6, note that three samples are overlapped,
therefore
𝑦(3) = −1 × 2 + 1 × 3 + 2 × 1 = 3
f. To compute 𝑦(4), refer figure 1 and figure 7, note that two samples are overlapped,
therefore
𝑦(4) = −1 × 3 + 1 × 1 = −2
g. To compute 𝑦(5), refer figure 1 and figure 8, note that only one samples is
overlapped, therefore
𝑦(5) = −1 × 1 = −1
 Prepared by Dr Harish M S

𝑦(−1), 𝑦(0), 𝑦(1), 𝑦(2), 𝑦(3), 𝑦(4), 𝑦(5) 1, 4, 8, 8, 3, −2, −1

𝑦 (𝑛 ) = ( )=( )
↑ ↑
Problem: Determine the output 𝑦(𝑛)of a relaxed linear time-invariant system with impulse
response ℎ(𝑛) = 𝑎𝑛 𝑢(𝑛) |𝑎| < 1, when the input is a step sequence.
Solution: The LTI system as shown in below figure.

𝑦 (𝑛 ) = 𝑥 (𝑛 ) ∗ ℎ(𝑛 ) = ∑ 𝑥 (𝑘 )ℎ (𝑛 − 𝑘 ) −∞≤𝑛≤∞
𝑘=−∞

1, 𝑛 ≤ 0
Input 𝑥(𝑛) is a step input, it is defined as 𝑥(𝑛) = 𝑢(𝑛) = {
0, 𝑛 < 0
𝑎𝑛 , 𝑛 ≥ 0
And impulse response of the system is ℎ(𝑛) = 𝑎𝑛 𝑢(𝑛) = {
0, 𝑛 < 0

Let the number of samples in 𝑥(𝑛) is 𝑁1 = ∞and number of samples in ℎ(𝑛) is 𝑁2 = ∞

Total number of samples in 𝑦(𝑛) is given by 𝑁 = 𝑁1 + 𝑁2 − 1 = ∞ + ∞ − 1 = ∞ thus; there
are infinite numbers of samples in 𝑦(𝑛).
Starting point of 𝑦(𝑛) is 𝑦𝑙 = 𝑥𝑙 +ℎ𝑙 = 0 + 0 = 0.
Terminating point of 𝑦(𝑛) is 𝑦ℎ = 𝑥ℎ +ℎℎ = ∞ + ∞ = ∞.
𝑦(0), 𝑦(1), 𝑦(2), 𝑦(3), … … ..
The infinite samples in 𝑦(𝑛), they are 𝑦(𝑛) = ( )
↑
∞

𝑦 (𝑛 ) = ∑ 𝑥 (𝑘 )ℎ (𝑛 − 𝑘 ) −∞≤𝑛≤∞
𝑘=−∞

When 𝑛 = 0, only one sample is overlapped i.e.

 Prepared by Dr Harish M S

When 𝑛 = 1, only two samples are overlapped i.e.

Output is from 0 to infinity, in general, therefore

𝑛

𝑦 (𝑛 ) = ∑ 𝑥 (𝑘 )ℎ (𝑛 − 𝑘 ), 0≤𝑛≤∞
𝑘=0
𝑛 𝑛 𝑛 1
1 1 − (𝑎)𝑛+1
𝑦(𝑛) = ∑ 1 × 𝑎𝑛−𝑘 = 𝑎𝑛 ∑ 𝑎−𝑘 𝑛 𝑛
=𝑎 ∑ 𝑘 = 𝑎 { 1 }
𝑎 1−
𝑘=0 𝑘=0 𝑘=0 𝑎

1 𝑛+1 1 𝑛+1
1 − (𝑎 ) 1 − (𝑎 ) 𝑎𝑛+1 − 1 1 − 𝑎𝑛+1
𝑦 (𝑛 ) = 𝑎 𝑛 { 𝑎−1 } = 𝑎𝑛+1 [ ]= [ ]=
𝑎−1 𝑎−1 1−𝑎
𝑎
 Prepared by Dr Harish M S

𝑦(0) = 1
𝑌(1) = 1 + 𝑎
𝑦(2) = 1 + 𝑎 + 𝑎2
..
.
1
𝑦(∞) =
1−𝑎
1 − 𝑎𝑛+1
𝑦 (𝑛 ) = 0≤𝑛≤∞
1−𝑎

Problem: Compute𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛), where 𝑥(𝑛) = 1 + 𝑛, 0 ≤ 𝑛 ≤ 2 and ℎ(𝑛) = 1 + 𝑛, 0 ≤ 𝑛 ≤ 3.

Solution:

1. Analytical method
1, 2, 3
Input 𝑥(𝑛) is defined as 𝑥(𝑛) = ( )
↑
1, 2, 3, 4
And impulse response of the system is ℎ(𝑛) = ( )
↑
∞

𝑦 (𝑛 ) = 𝑥 (𝑛 ) ∗ ℎ(𝑛 ) = ∑ 𝑥 (𝑘 )ℎ (𝑛 − 𝑘 ) −∞≤𝑛≤∞
𝑘=−∞

Let the number of samples in 𝑥(𝑛) is 𝑁1 = 3 and number of samples in ℎ(𝑛) is 𝑁2 = 4

Total number of samples in 𝑦(𝑛) is given by 𝑁 = 𝑁1 + 𝑁2 − 1 = 3 + 4 − 1 = 6 thus; there are
six numbers of samples in 𝑦(𝑛).
Starting point of 𝑦(𝑛) is 𝑦𝑙 = 𝑥𝑙 +ℎ𝑙 = 0 + 0 = 0.
Terminating point of 𝑦(𝑛) is 𝑦ℎ = 𝑥ℎ +ℎℎ = 2 + 3 = 5.
3

∴ 𝑦(𝑛) = ∑ 𝑥(𝑘)ℎ(𝑛 − 𝑘) 0≤𝑛≤5

𝑘=0

𝑦(𝑛) = 𝑥(0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥 (2)ℎ(𝑛 − 2) + 𝑥(3)ℎ(𝑛 − 3) 0≤𝑛≤5

𝑦(0) = 𝑥(0)ℎ(0) + 𝑥 (1)ℎ(0 − 1) + 𝑥(2)ℎ(0 − 2) + 𝑥(3)ℎ(0 − 3)
 Prepared by Dr Harish M S

𝑦(0) = 𝑥(0)ℎ(0) + 𝑥(1)ℎ(−1) + 𝑥(2)ℎ(−2) + 𝑥(3)ℎ(−3)

𝑦 (0) = 1 × 1 + 1 × 0 + 2 × 0 + 0 × 0 = 1

𝑦(𝑛) = 𝑥(0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2) + 𝑥(3)ℎ(𝑛 − 3)

𝑦(1) = 𝑥(0)ℎ(1) + 𝑥(1)ℎ(1 − 1) + 𝑥(2)ℎ(1 − 2) + 𝑥(3)ℎ(1 − 3)

𝑦(1) = 𝑥(0)ℎ(1) + 𝑥(1)ℎ(0) + 𝑥(2)ℎ(−1) + 𝑥(3)ℎ(−2)

𝑦 (1) = 1 × 2 + 2 × 1 + 3 × 0 + 0 × 0 = 4

𝑦(𝑛) = 𝑥(0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2) + 𝑥(3)ℎ(𝑛 − 3)

𝑦(2) = 𝑥(0)ℎ(2) + 𝑥 (1)ℎ(2 − 1) + 𝑥(2)ℎ(2 − 2) + 𝑥(3)ℎ(2 − 3)

𝑦(2) = 𝑥(0)ℎ(2) + 𝑥(1)ℎ(1) + 𝑥(2)ℎ(0) + 𝑥(3)ℎ(−1)

𝑦(2) = 1 × 3 + 2 × 2 + 3 × 1 + 0 × 0 = 10

𝑦(𝑛) = 𝑥(0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2) + 𝑥(3)ℎ(𝑛 − 3)

𝑦(3) = 𝑥(0)ℎ(3) + 𝑥 (1)ℎ(3 − 1) + 𝑥(2)ℎ(3 − 2) + 𝑥(3)ℎ(3 − 3)

𝑦(3) = 𝑥 (0)ℎ(3) + 𝑥(1)ℎ(2) + 𝑥(2)ℎ(1) + 𝑥(3)ℎ(0)

𝑦(3) = 1 × 4 + 2 × 3 + 3 × 2 + 0 × 1 = 16

𝑦(𝑛) = 𝑥(0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2) + 𝑥(3)ℎ(𝑛 − 3)

𝑦(4) = 𝑥(0)ℎ(4) + 𝑥(1)ℎ(4 − 1) + 𝑥(2)ℎ(4 − 2) + 𝑥(3)ℎ(4 − 3)

𝑦(4) = 𝑥(0)ℎ(4) + 𝑥(1)ℎ(3) + 𝑥(2)ℎ(2) + 𝑥(3)ℎ(1)

𝑦(4) = 1 × 0 + 2 × 4 + 3 × 3 + 0 × 2 = 17

𝑦(𝑛) = 𝑥(0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2) + 𝑥(3)ℎ(𝑛 − 3)

𝑦(5) = 𝑥(0)ℎ(5) + 𝑥(1)ℎ(5 − 1) + 𝑥(2)ℎ(5 − 2) + 𝑥(3)ℎ(5 − 3)

𝑦(5) = 𝑥(0)ℎ(5) + 𝑥(1)ℎ(4) + 𝑥(2)ℎ(3) + 𝑥(3)ℎ(2)

𝑦(5) = 1 × 0 + 2 × 0 + 3 × 4 + 0 × 3 = 12

𝑦(𝑛) = (𝑦(0), 𝑦(1), 𝑦(2), 𝑦(3), 𝑦(4), 𝑦(5)) = ( 1, 4,10, 16, 17, 12)

2. Graphical method

Do it your self
1
Problem: Compute 𝑦(𝑛) = 𝑢(𝑛) ∗ (2)𝑛 [𝑢(𝑛) − 𝑢(𝑛 − 4)]
 Prepared by Dr Harish M S

1
Solution: 𝑦(𝑛) = 𝑢(𝑛) ∗ (2)𝑛 [𝑢(𝑛) − 𝑢(𝑛 − 4)] = 𝑥(𝑛) ∗ ℎ(𝑛)
1
1, 𝑛 ≥ 0 1 ( )𝑛 , 0 ≤ 𝑛 ≤ 3
𝑥 (𝑛 ) = 𝑢 (𝑛 ) = { And ℎ(𝑛) = (2)𝑛 [𝑢(𝑛) − 𝑢(𝑛 − 4)] = { 2
0, 𝑛 < 0 0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

𝑦 (𝑛 ) = 𝑥 (𝑛 ) ∗ ℎ(𝑛 ) = ∑ 𝑥 (𝑘 )ℎ (𝑛 − 𝑘 ) −∞≤𝑛≤∞
𝑘=−∞

Let number of samples in 𝑥(𝑛) is 𝑁1 = ∞ and number of samples in ℎ(𝑛) is 𝑁2 = 4

Total number of samples in 𝑦(𝑛) is given by 𝑁 = 𝑁1 + 𝑁2 − 1 = ∞ + 4 − 1 = ∞ thus; there are
infinite numbers of samples in 𝑦(𝑛).
Starting point of 𝑦(𝑛) is 𝑦𝑙 = 𝑥𝑙 +ℎ𝑙 = 0 + 0 = 0.
Terminating point of 𝑦(𝑛) is 𝑦ℎ = 𝑥ℎ +ℎℎ = ∞ + 3 = ∞.
∞

𝑦 (𝑛 ) = ∑ 𝑥 (𝑘 ) ℎ(𝑛 − 𝑘 ) 0≤𝑛≤∞
𝑘=0

𝑦(𝑛) = 𝑥 (0)ℎ(𝑛 − 0) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2) + 𝑥(3)ℎ(𝑛 − 3) + ⋯.

From figure1,𝑎𝑡 𝑛 = 0 only one sample is overlapped, 𝑦(0) = 1 × 1 = 1

1
From figure2,𝑎𝑡 𝑛 = 1 two samples are is overlapped, 𝑦(1) = 1 × 2 + 1 × 1 = 1.5
 Prepared by Dr Harish M S

1 1
From figure3,𝑎𝑡 𝑛 = 2 three samples are overlapped, 𝑦(2) = 1 × 4 + 1 × 2 + 1 × 1 = 1.75
1 1 1
From figure4,𝑎𝑡 𝑛 = 3 four samples are overlapped, 𝑦(3) = 1 × 8 + 1 × 4 + 1 × 2 + 1 × 1 =

1.875
From figure 5,𝑎𝑡 𝑛 > 3 four samples are overlapped,
1
𝑦(𝑛) = ∑𝑛𝑘=3−3 𝑥(𝑘)ℎ(𝑛 − 𝑘) = ∑𝑛𝑘=𝑛−3 1 × (2)𝑛−𝑘
𝑛 𝑛
1 1 1
𝑦(𝑛) = ∑ ( )𝑛−𝑘 = ∑ ( )𝑛 ( )−𝑘
2 2 2
𝑘=𝑛−3 𝑘=𝑛−3
𝑛
1
𝑦(𝑛) = ( )𝑛 ∑ (2)𝑘
2
𝑘=𝑛−3

Put 𝑙 = 𝑘 − 𝑛 + 3, 𝑘 =𝑙+𝑛−3
If the lower limit 𝑘 = 𝑛 − 3, then𝑙 = 0, if upper limit k=n, then 𝑙 = 3
3 3
1 1
𝑦(𝑛) = ( )𝑛 ∑(2)𝑙+𝑛−3 = ( )𝑛 ∑(2)𝑙 (2)𝑛 (2)−3
2 2
𝑙=0 𝑙=0
3
1 1 1 − 24 1 1 − 16 15
𝑦(𝑛) = ( )𝑛 (2)𝑛 (2)−3 ∑(2)𝑙 = [ ]= [ ]= = 1.875
2 8 1−2 5 −1 8
𝑙=0

1, 𝑛 = 0
1.5, 𝑛 = 1
∴ 𝑦 (𝑛 ) = {
1.75, 𝑛=2
1.875, 𝑛 ≥ 3
Problem: If𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛), show that ∑𝑦 = ∑𝑥 ∑ℎ where ∑x = ∑∞n=−∞ x(n)
 Prepared by Dr Harish M S

Solution: We know that 𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛) = ∑∞𝑘=0 𝑥(𝑘)ℎ(𝑛 − 𝑘) 0≤𝑛≤∞

∞

𝑦(𝑛) = ∑ 𝑥(𝑘)ℎ(𝑛 − 𝑘) = ∑ ℎ(𝑘)𝑥(𝑛 − 𝑘)

𝑘=0

Take summation on both sides, we get

∞ ∞ ∞

∑ 𝑦(𝑛) = ∑ ∑ ℎ(𝑘)𝑥(𝑛 − 𝑘)
𝑛=−∞ 𝑛=−∞ 𝑘=0

Inter change the order of summation, we get

∞ ∞ ∞

∑ 𝑦(𝑛) = ∑ ℎ(𝑘) ∑ 𝑥(𝑛 − 𝑘)

𝑛=−∞ 𝑘=−∞ 𝑛=−∞

Put 𝑛 − 𝑘 = 𝑙 in the second summation. If th lower limit 𝑛 = −∞ then 𝑙 = −∞, if the upper
limit 𝑛 = −∞ then 𝑙 = ∞.
∞ ∞ ∞

∑ 𝑦(𝑛) = ∑ ℎ(𝑘) ∑ 𝑥(𝑙 )

𝑛=−∞ 𝑘=−∞ 𝑙=−∞

Since 𝑙 is the dummy variable, then

∞ ∞ ∞ ∞ ∞

∑ 𝑦(𝑛) = ∑ ℎ(𝑘) ∑ 𝑥(𝑛) = ∑ 𝑥(𝑛) ∑ ℎ(𝑘)

𝑛=−∞ 𝑘=−∞ 𝑛=−∞ 𝑛=−∞ 𝑘=−∞

∴ ∑ = ∑∑
𝑦 𝑥 ℎ

1, 1, 0, 1, 1 1, −2, −3,4
Problem: Compute𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛), if 𝑥(𝑛) = ( ) and ℎ(𝑛) = ( ) and
↑ ↑
verify the result using the relation ∑𝑦 = ∑𝑥 ∑ℎ
Solution: We know that 𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛) = ∑∞𝑘=0 𝑥(𝑘)ℎ(𝑛 − 𝑘) 0≤𝑛≤∞
Let the number of samples in 𝑥(𝑛) is 𝑁1 = 5 and number of samples in ℎ(𝑛) is 𝑁2 = 4
Total number of samples in 𝑦(𝑛) is given by 𝑁 = 𝑁1 + 𝑁2 − 1 = 5 + 4 − 1 = 8 thus; there are
eight samples in 𝑦(𝑛).
Starting point of 𝑦(𝑛) is 𝑦𝑙 = 𝑥𝑙 +ℎ𝑙 = −2 − 3 = −5.
Terminating point of 𝑦(𝑛) is 𝑦ℎ = 𝑥ℎ +ℎℎ = 2 + 0 = 2.
2

∴ 𝑦 (𝑛 ) = ∑ 𝑥 (𝑘 ) ℎ(𝑛 − 𝑘 ) −5≤𝑛≤2
𝑘=−3
 Prepared by Dr Harish M S

𝑦(𝑛) = x(−3)h(n + 3) + x(−2)h(n + 2) + x(−1)h(n + 1) + x(0)h(n) + x(1)h(n − 1)

+ x(2)h(n − 2) −5≤n≤2
𝑦(−5) = x(−3)h(−2) + x(−2)h(−3) + x(−1)h(−4) + x(0)h(−5) + x(1)h(−6)
+ x(2)h(−7)
𝑦(−5) = 0 × −2 + 1 × 1 + 1 × 0 + 0 × 0 + 1 × 0 + 1 × 0 = 1
𝑦(−4) = x(−3)h(−1) + x(−2)h(−2) + x(−1)h(−3) + x(0)h(−4) + x(1)h(−5)
+ x(2)h(−6)
𝑦(−4) = 0 × −3 + 1 × −2 + 1 × 1 + 0 × 0 + 1 × 0 + 1 × 0 = −1
𝑦(−3) = x(−3)h(0) + x(−2)h(−1) + x(−1)h(−2) + x(0)h(−3) + x(1)h(−4) + x(2)h(−5)
𝑦(−3) = 0 × 4 ± 5 + 1 × −2 + 0 × 1 + 1 × 0 + 1 × 0 = −5
𝑦(−2) = x(−3)h(1) + x(−2)h(0) + x(−1)h(−1) + x(0)h(−2) + x(1)h(−3) + x(2)h(−4)
𝑦(−2) = 0 × 0 + 1 × 4 + 1 × −3 + 0 × −2 + 1 × 1 + 1 × 0 = 2
𝑦(−1) = x(−3)h(2) + x(−2)h(1) + x(−1)h(0) + x(0)h(−1) + x(1)h(−2)
+ x(2)h(−3)
𝑦(−1) = 0 × 0 + 1 × 0 + 1 × 4 + 0 × −3 + 1 × −2 + 1 × 1 = 3
𝑦(0) = x(−3)h(3) + x(−2)h(2) + x(−1)h(1) + x(0)h(0) + x(1)h(−1) + x(2)h(−2)
𝑦(0) = 0 × 0 + 1 × 0 + 1 × 0 + 0 × 4 + 1 × −3 + 1 × −2 = −5
𝑦(1) = x(−3)h(4) + x(−2)h(3) + x(−1)h(2) + x(0)h(1) + x(1)h(0) + x(2)h(−1)
𝑦(1) = 0 × 0 + 1 × 0 + 1 × 0 + 0 × 0 + 1 × 4 + 1 × −3 = 1
𝑦(2) = x(−3)h(5) + x(−2)h(4) + x(−1)h(3) + x(0)h(2) + x(1)h(1) + x(2)h(0)
𝑦(2) = 0 × 0 + 1 × 0 + 1 × 0 + 0 × 0 + 1 × 0 + 1 × 4 = 4
1, −1, −5, 2, 3, −5, 1, 4
𝑦 (𝑛 ) = ( )
↑
Verification ∑𝑦 = ∑𝑥 ∑ℎ
LHS ∑𝑦 = ∑2𝑛=−5 𝑦(𝑛) = 1 − 1 − 5 + 2 + 3 − 5 + 1 + 4 = 0
2

∑ = ∑ 𝑥 (𝑛 ) = 1 + 1 + 0 + 1 + 1 = 4
𝑥 𝑛=−2
0

∑ = ∑ ℎ(𝑛 ) = 1 − 2 − 3 + 4 = 0
ℎ 𝑛=−3

RHS ∑𝑥 ∑ℎ =4×0=0
 Prepared by Dr Harish M S

Therefore LHS=RHS
Problem: Compute 𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛), if
1
𝑛, 0 ≤ 𝑛 ≤ 6 1, − 2 ≤ 𝑛 ≤ 2
𝑥 ( 𝑛 ) = {3 And ℎ(𝑛) = {
0, 𝑒𝑙𝑠𝑒𝑤ℎ𝑒𝑟𝑒 0, 𝑒𝑙𝑠𝑒𝑤ℎ𝑒𝑟𝑒

Solution: 𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛) = ∑∞

𝑘=−∞ 𝑥 (𝑘 ) ℎ(𝑛 − 𝑘 ) −∞ ≤𝑛 ≤ ∞
1 2 4 5
0, 3 , 3 , 1, 3 , 3 , 2 1, 1, 1, 1, 1
𝑥 (𝑛 ) = ( ) And ℎ(𝑛) ( )
↑ ↑
Let the number of samples in 𝑥(𝑛) is 𝑁1 = 7 and number of samples in ℎ(𝑛) is 𝑁2 = 5
Total number of samples in 𝑦(𝑛) is given by 𝑁 = 𝑁1 + 𝑁2 − 1 = 7 + 5 − 1 = 11 thus; there are
eleven samples in 𝑦(𝑛).
Starting point of 𝑦(𝑛) is 𝑦𝑙 = 𝑥𝑙 +ℎ𝑙 = 0 − 2 = −2.
Terminating point of 𝑦(𝑛) is 𝑦ℎ = 𝑥ℎ +ℎℎ = 6 + 2 = 8.
6

∴ 𝑦 (𝑛 ) = ∑ 𝑥 (𝑘 ) ℎ(𝑛 − 𝑘 ) −2≤𝑛≤8
𝑘=−2

𝑦(𝑛) = 𝑥(−2)ℎ(𝑛 + 2) + 𝑥 (−1)ℎ(𝑛 + 1) + 𝑥 (0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2)

+ 𝑥(3)ℎ(𝑛 − 3) + 𝑥(4)ℎ(𝑛 − 4) + 𝑥(5)ℎ(𝑛 − 5) + 𝑥(6)ℎ(𝑛 − 6)
𝑦(−2) = 𝑥 (−2)ℎ(−2 + 2) + 𝑥(−1)ℎ(−2 + 1) + 𝑥(0)ℎ(−2) + 𝑥(1)ℎ(−2 − 1)
+ 𝑥 (2)ℎ(−2 − 2) + 𝑥(3)ℎ(−2 − 3) + 𝑥(4)ℎ(−2 − 4) + 𝑥(5)ℎ(−2 − 5)
+ 𝑥 (6)ℎ(−2 − 6)
1 2 4 5
𝑦(−2) = 0 × 1 + 0 × 1 + 0 × 1 + × 0 + × 0 + 1 × 0 + × 0 + × 0 + 2 × 0 = 0
3 3 3 2
𝑦(𝑛) = 𝑥(−2)ℎ(𝑛 + 2) + 𝑥 (−1)ℎ(𝑛 + 1) + 𝑥 (0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2)
+ 𝑥(3)ℎ(𝑛 − 3) + 𝑥(4)ℎ(𝑛 − 4) + 𝑥(5)ℎ(𝑛 − 5) + 𝑥(6)ℎ(𝑛 − 6)
𝑦(−1) = 𝑥 (−2)ℎ(−1 + 2) + 𝑥(−1)ℎ(−1 + 1) + 𝑥(0)ℎ(−1) + 𝑥(1)ℎ(−1 − 1)
+ 𝑥 (2)ℎ(−1 − 2) + 𝑥(3)ℎ(−1 − 3) + 𝑥(4)ℎ(−1 − 4) + 𝑥(5)ℎ(−1 − 5)
+ 𝑥 (6)ℎ(−1 − 6)
1 2 4 5 1
𝑦(−1) = 0 × 1 + 0 × 1 + 0 × 1 + × 1 + × 0 + 1 × 0 + × 0 + 0 + 2 × 0 =
3 3 3 3 3
𝑦(𝑛) = 𝑥(−2)ℎ(𝑛 + 2) + 𝑥 (−1)ℎ(𝑛 + 1) + 𝑥 (0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2)
+ 𝑥(3)ℎ(𝑛 − 3) + 𝑥(4)ℎ(𝑛 − 4) + 𝑥(5)ℎ(𝑛 − 5) + 𝑥(6)ℎ(𝑛 − 6)
 Prepared by Dr Harish M S

𝑦(0) = 𝑥(−2)ℎ(0 + 2) + 𝑥 (−1)ℎ(0 + 1) + 𝑥 (0)ℎ(0) + 𝑥(1)ℎ(0 − 1) + 𝑥(2)ℎ(0 − 2)

+ 𝑥(3)ℎ(0 − 3) + 𝑥(4)ℎ(0 − 4) + 𝑥(5)ℎ(0 − 5) + 𝑥(6)ℎ(0 − 6)
1 2 4 5
𝑦 (0 ) = 0 × 1 + 0 × 1 + 0 × 1 +
× 1 + × 1 + 1 × 0 + × ℎ (0 − 4 ) + × 0 + 2 × 0 = 1
3 3 3 3
𝑦(𝑛) = 𝑥(−2)ℎ(𝑛 + 2) + 𝑥 (−1)ℎ(𝑛 + 1) + 𝑥 (0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2)
+ 𝑥(3)ℎ(𝑛 − 3) + 𝑥(4)ℎ(𝑛 − 4) + 𝑥(5)ℎ(𝑛 − 5) + 𝑥(6)ℎ(𝑛 − 6)
𝑦(1) = 𝑥(−2)ℎ(1 + 2) + 𝑥 (−1)ℎ(1 + 1) + 𝑥 (0)ℎ(1) + 𝑥(1)ℎ(1 − 1) + 𝑥(2)ℎ(1 − 2)
+ 𝑥(3)ℎ(1 − 3) + 𝑥(4)ℎ(1 − 4) + 𝑥(5)ℎ(1 − 5) + 𝑥(6)ℎ(1 − 6)
1 2 4 5
𝑦 (1) = 0 × 0 + 0 × 1 + 0 × 1 + × 1 + × 1 + 1 × 1 + × 0 + × 0 + 2 × 0 = 2
3 3 3 3
𝑦(𝑛) = 𝑥(−2)ℎ(𝑛 + 2) + 𝑥 (−1)ℎ(𝑛 + 1) + 𝑥 (0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2)
+ 𝑥(3)ℎ(𝑛 − 3) + 𝑥(4)ℎ(𝑛 − 4) + 𝑥(5)ℎ(𝑛 − 5) + 𝑥(6)ℎ(𝑛 − 6)
𝑦(2) = 𝑥(−2)ℎ(2 + 2) + 𝑥 (−1)ℎ(2 + 1) + 𝑥 (0)ℎ(2) + 𝑥(1)ℎ(2 − 1) + 𝑥(2)ℎ(2 − 2)
+ 𝑥(3)ℎ(2 − 3) + 𝑥(4)ℎ(2 − 4) + 𝑥(5)ℎ(2 − 5) + 𝑥(6)ℎ(2 − 6)
1 2 4 5 10
𝑦 (2) = 0 × 0 + 0 × 0 + 0 × 1 +×1+ ×1+1×1+ ×1+ ×0+2×0=
3 3 3 3 3
𝑦(𝑛) = 𝑥(−2)ℎ(𝑛 + 2) + 𝑥 (−1)ℎ(𝑛 + 1) + 𝑥 (0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2)
+ 𝑥(3)ℎ(𝑛 − 3) + 𝑥(4)ℎ(𝑛 − 4) + 𝑥(5)ℎ(𝑛 − 5) + 𝑥(6)ℎ(𝑛 − 6)
𝑦(3) = 𝑥(−2)ℎ(3 + 2) + 𝑥 (−1)ℎ(3 + 1) + 𝑥 (0)ℎ(3) + 𝑥(1)ℎ(3 − 1) + 𝑥(2)ℎ(3 − 2)
+ 𝑥(3)ℎ(3 − 3) + 𝑥(4)ℎ(3 − 4) + 𝑥(5)ℎ(3 − 5) + 𝑥(6)ℎ(3 − 6)
1 2 4 5
𝑦 (3) = 0 × 0 + 0 × 0 + 0 × 1 + × 1 + × 1 + 1 × 1 + × 1 + × 1 + 2 × 0 = 5
3 3 3 3
𝑦(𝑛) = 𝑥(−2)ℎ(𝑛 + 2) + 𝑥 (−1)ℎ(𝑛 + 1) + 𝑥 (0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2)
+ 𝑥(3)ℎ(𝑛 − 3) + 𝑥(4)ℎ(𝑛 − 4) + 𝑥(5)ℎ(𝑛 − 5) + 𝑥(6)ℎ(𝑛 − 6)
𝑦(4) = 𝑥(−2)ℎ(4 + 2) + 𝑥 (−1)ℎ(4 + 1) + 𝑥 (0)ℎ(4) + 𝑥(1)ℎ(4 − 1) + 𝑥(2)ℎ(4 − 2)
+ 𝑥(3)ℎ(4 − 3) + 𝑥(4)ℎ(4 − 4) + 𝑥(5)ℎ(4 − 5) + 𝑥(6)ℎ(4 − 6)
1 2 4 5 20
𝑦 (4) = 0 × 0 + 0 × 0 + 0 × 0 + ×0+ ×1+1×1+ ×1+ ×1+2×1=
3 3 3 3 3

𝑦(𝑛) = 𝑥(−2)ℎ(𝑛 + 2) + 𝑥 (−1)ℎ(𝑛 + 1) + 𝑥 (0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2)

+ 𝑥(3)ℎ(𝑛 − 3) + 𝑥(4)ℎ(𝑛 − 4) + 𝑥(5)ℎ(𝑛 − 5) + 𝑥(6)ℎ(𝑛 − 6)
 Prepared by Dr Harish M S

𝑦(5) = 𝑥(−2)ℎ(5 + 2) + 𝑥 (−1)ℎ(5 + 1) + 𝑥 (0)ℎ(5) + 𝑥(1)ℎ(5 − 1) + 𝑥(2)ℎ(5 − 2)

+ 𝑥(3)ℎ(5 − 3) + 𝑥(4)ℎ(5 − 4) + 𝑥(5)ℎ(5 − 5) + 𝑥(6)ℎ(5 − 6)
1 2 4 5
𝑦 (5) = 0 × 0 + 0 × 0 + 0 × 0 + × 0 + × 0 + 1 × 1 + × 1 + × 1 + 2 × 1 = 6
3 3 3 3
𝑦(𝑛) = 𝑥(−2)ℎ(𝑛 + 2) + 𝑥 (−1)ℎ(𝑛 + 1) + 𝑥 (0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2)
+ 𝑥(3)ℎ(𝑛 − 3) + 𝑥(4)ℎ(𝑛 − 4) + 𝑥(5)ℎ(𝑛 − 5) + 𝑥(6)ℎ(𝑛 − 6)
𝑦(6) = 𝑥(−2)ℎ(6 + 2) + 𝑥 (−1)ℎ(6 + 1) + 𝑥 (0)ℎ(6) + 𝑥(1)ℎ(6 − 1) + 𝑥(2)ℎ(6 − 2)
+ 𝑥(3)ℎ(6 − 3) + 𝑥(4)ℎ(6 − 4) + 𝑥(5)ℎ(6 − 5) + 𝑥(6)ℎ(6 − 6)
1 2 4 5
𝑦 (6) = 0 × 0 + 0 × 0 + 0 × 0 + × 0 + × 0 + 1 × 0 + × 1 + × 1 + 2 × 1 = 5
3 3 3 3
𝑦(𝑛) = 𝑥(−2)ℎ(𝑛 + 2) + 𝑥 (−1)ℎ(𝑛 + 1) + 𝑥 (0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2)
+ 𝑥(3)ℎ(𝑛 − 3) + 𝑥(4)ℎ(𝑛 − 4) + 𝑥(5)ℎ(𝑛 − 5) + 𝑥(6)ℎ(𝑛 − 6)
𝑦(7) = 𝑥(−2)ℎ(7 + 2) + 𝑥 (−1)ℎ(7 + 1) + 𝑥 (0)ℎ(7) + 𝑥(1)ℎ(7 − 1) + 𝑥(2)ℎ(7 − 2)
+ 𝑥(3)ℎ(7 − 3) + 𝑥(4)ℎ(7 − 4) + 𝑥(5)ℎ(7 − 5) + 𝑥(6)ℎ(7 − 6)
1 2 4 5 11
𝑦 (7) = 0 × 0 + 0 × 0 + 0 × 0 + ×0+ ×0+1×0+ ×0+ ×1+2×1=
3 3 3 3 3
𝑦(𝑛) = 𝑥(−2)ℎ(𝑛 + 2) + 𝑥 (−1)ℎ(𝑛 + 1) + 𝑥 (0)ℎ(𝑛) + 𝑥(1)ℎ(𝑛 − 1) + 𝑥(2)ℎ(𝑛 − 2)
+ 𝑥(3)ℎ(𝑛 − 3) + 𝑥(4)ℎ(𝑛 − 4) + 𝑥(5)ℎ(𝑛 − 5) + 𝑥(6)ℎ(𝑛 − 6)
𝑦(8) = 𝑥(−2)ℎ(8 + 2) + 𝑥 (−1)ℎ(8 + 1) + 𝑥 (0)ℎ(8) + 𝑥(1)ℎ(8 − 1) + 𝑥(2)ℎ(8 − 2)
+ 𝑥(3)ℎ(8 − 3) + 𝑥(4)ℎ(8 − 4) + 𝑥(5)ℎ(8 − 5) + 𝑥(6)ℎ(8 − 6)
1 2 4 5
𝑦(8) = 0 × 0 + 0 × 0 + 0 × 0 + × 0 + × 0 + 1 × 1 + × 1 + × 0 + 2 × 1 = 2
3 3 3 3
1 10 20 11
0,
∴ 𝑦 (𝑛 ) = ( 3 , 1, 2, , 5, , 6,5, ,2
3 3 3 )
↑
Problem: Compute 𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛), if
1 1
𝑥(𝑛) = (2)𝑛 𝑢(𝑛) And ℎ(𝑛) = (4)𝑛 𝑢(𝑛)

Solution: 𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛) = ∑∞

𝑘=−∞ 𝑥 (𝑘 ) ℎ(𝑛 − 𝑘 ) −∞ ≤𝑛 ≤ ∞
1 𝑛 1
1 𝑛 (2) , 𝑛≥0 1 𝑛 (4)𝑛 , 𝑛≥0
𝑥(𝑛) = (2) 𝑢(𝑛) = { And ℎ(𝑛) = (4) 𝑢(𝑛) = {
0 , 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
Let the number of samples in 𝑥(𝑛) is 𝑁1 = ∞ and number of samples in ℎ(𝑛) is 𝑁2 = ∞
 Prepared by Dr Harish M S

Total number of samples in 𝑦(𝑛) is given by 𝑁 = 𝑁1 + 𝑁2 − 1 = ∞ + ∞ − 1 = ∞ thus; there

are eleven samples in 𝑦(𝑛).
Starting point of 𝑦(𝑛) is 𝑦𝑙 = 𝑥𝑙 +ℎ𝑙 = 0 + 0 = 0.
Terminating point of 𝑦(𝑛) is 𝑦ℎ = 𝑥ℎ +ℎℎ = ∞ + ∞ = ∞.

∴ 𝑦(𝑛) = ∑ 𝑥(𝑘)ℎ(𝑛 − 𝑘) 0≤𝑛≤∞

𝑘=0
𝑛 𝑛
1 1 1 1 1
𝑦(𝑛) = ∑ ( )𝑘 ( )𝑛−𝑘 = ∑ ( )𝑘 ( )𝑛 ( )−𝑘
2 4 2 4 4
𝑘=0 𝑘=0
𝑛 𝑛
1 1 1 1 1 − 2𝑛+1
𝑦 ( 𝑛 ) = ( )𝑛 ∑ ( )𝑘 (4 ) = ( )𝑛 ∑ 2𝑘 = ( )𝑛 [
𝑘
]
4 2 4 4 1−2
𝑘=0 𝑘=0

1 1 2 1 1 1
𝑦(𝑛) = ( )𝑛 [2𝑛+1 − 1] = ( )𝑛 [2 × 2𝑛 − 1] = 2( )𝑛 − ( )𝑛 = 2( )𝑛 − ( )𝑛
4 4 4 4 2 4
1 1
𝑦(𝑛) = 2( )𝑛 − ( )𝑛 , 0≤𝑛≤∞
2 4