DISCRETE-TIME SIGNALS AND

SYSTEMS IN THE TIME-DOMAIN

PART A -- DISCRETE-TIME
SIGNALS

PART B -- DISCRETE-TIME
SYSTEMS

PART C -- TIME-DOMAIN
CHARACTERIZATION OF LTI
DISCRETE-TIME SYSTEMS
PART A:- DISCRETE-TIME
SIGNALS
TIME-DOMAIN REPRESENTATION
Types of Discrete-Time Signals
Finite & Infinite-Length Sequence
Right, Left & Double Sided Sequence
OPERATIONS ON SEQUENCES
Basic Operations
Combinations of Basic Operations
SAMPLING RATE ALTERATION
Up Sampling
Down Sampling
CLASSIFICATION OF SEQUENCES
TYPICAL SEQUENCES
SAMPLING PROCESS
Aliasing
 TIME-DOMAIN REPRESENTATION
Signals represented as sequences of numbers, called
samples

Sample value of typical signal or sequence denoted as x[n]

with n being an integer in the range - to +

x[n] defined only for integer values of n and undefined

for non-integer values of n

Discrete-time signal represented by {x[n]}

Discrete-time signal may also be written as a sequence of

numbers inside braces:
 TIME-DOMAIN REPRESENTATION
For Example

x[n] ...,0.95, 0.2, 2.17,1.1,0.2, 3.67,2.9, 0.8,4.1,......

In the above sequence

{x[-2]}=-0.95, {x[-1]}=-0.2, {x[0]}=2.17, {x[1]}=1.1 etc.

Arrow is placed under the sample at time index n = 0

Graphical representation of above discrete-time signal with

real-valued samples is as shown below:
 TIME-DOMAIN REPRESENTATION
In some applications, a discrete-time sequence {x[n]} may
be generated by periodically sampling a continuous-time
signal xa( t ) at uniform intervals of time
Here, n-th sample is given by
x n xa(t) t nT xa(nT), for n ..., 2, 1,0,1,2,...

Spacing T between two consecutive samples is called

sampling interval or sampling period
Reciprocal of sampling interval T, denoted as F T, and is
called the sampling frequency:
 TIME-DOMAIN REPRESENTATION
Whether or not sequence {x[n]} has been obtained by
sampling, quantity x[n] is called n th sample of sequence

{x[n]} is a real sequence,

If nth sample x[n] is real for all values of n
Otherwise, {x[n]} is a complex sequence

A complex sequence {x[n]} can be written as

{x[n]} = {xre[n]} + j {xim[n]}
where xre[n] and xim[n] are real & imaginary parts of x[n]

The complex conjugate sequence of {x[n]} is given by

{x*[n]} = {xre[n]} - j {xim[n]}

Often the braces are ignored to denote a sequence if there is

no ambiguity
 TIME-DOMAIN REPRESENTATION
Example
{x[n]}= {cos0.25n}
is a real sequence

{y[n]} = {ej 0.3n}

is a complex sequence
We can write
{y[n]} = {cos0.3n + j sin0.3n}
= {cos0.3n}+ j{sin0.3n}
Where
{yre[n]} = {cos0.3n}
{yim[n]} = {sin0.3n}

{w[n]} = {cos0.3n}- j{sin 0.3n}= {e- j 0.3n} is the complex

conjugate sequence of {y[n]}

i.e. {w[n]} = {y *[n]}

TYPES OF DISCRETE-TIME SIGNALS
Sampled-data signals in which samples are
continuous-valued
Digital signals in which samples are discrete-valued

Signals in a practical digital signal processing system are

digital signals obtained by quantizing the sample values
either by rounding or truncation
 FINITE & INFINITE-LENGTH
SEQUENCE
Discrete-time signal may be a finite length or an infinite-
length sequence

Finite-length (also called finite-duration or finite-extent)

sequence is defined only for a finite time interval
N1 < n < N2
where- < N1 and N2 < with N1 < N2

Length or duration of the above finite length sequence is

N = N2 - N1 +1
 FINITE & INFINITE-LENGTH
SEQUENCE
Example
x[n] n2 for 3 n 4 is a

finite length sequence of length n 4-(-3) 1 8

&
y[n] cos 0.4n is an infinite - length sequence
A length-N sequence is often referred to as an N-point
sequence

The length of a finite-length sequence can be increased by

zero-padding, i.e., by appending it with zeros
Example :- The length of above sequence can be increased
by zero-padding with 4 zero-valued
2 samples
n for 3 n 4
x[n]
0 for 5 n 8
 RIGHT, LEFT & DOUBLE SIDED
SEQUENCE
A finite length sequence can be considered as infinite
length sequence zero padding outside the defined length

There are three types of infinite length sequences

Right-sided sequence x[n]
has zero valued samples for n < N1
N1 can be +ve or ve finite integer
If N1 0 a right-sided sequence is called causal sequence

Left -sided sequence x[n]

has zero valued samples for n > N2
N2 can be +ve or ve finite integer
If N2 0 a left-sided sequence is called anti causal sequence

Double-sided sequence x[n]

x(n)0 for -n
 OPERATIONS ON SEQUENCES
A single-input, single-output discrete-time system operates
on a sequence,
called the input sequence,
according some prescribed rules and
develops another sequence, called the output sequence,
with more desirable properties

For example,
the input may be a signal corrupted with additive noise
Discrete-time system is designed to generate an output by
removing the noise component from the input
In most cases, the operation defining a particular discrete-
time system is composed of some basic operations
 BASIC OPERATIONS
Multiplication operation :- Scalar multiplication

x [n] A y[n] =A x [n]

Time-shifting operation:
y[n] = x[n - N]
where N is an integer
If N > 0, it is delaying operation

x [n] Z-1 y[n] = x [n-1]

If N < 0, it is an advance operation

x [n] Z y[n] = x [n-1]

 BASIC OPERATIONS
Time-reversal (folding) operation:

y[n] = x[-n]

Branching operation:
Used to provide multiple copies of a sequence

x[n] x[n]

x[n]
 ELEMENTARY OPERATIONS
Product / (modulation) operation:

x[n] X y[n]=x[n]w[n]

w[n]
Application is in forming a finite-length sequence from an
infinite-length sequence
by multiplying the latter with a finite-length sequence called
an window sequence
Process called windowing

Addition operation:
x[n] + y[n]=x[n]+w[n]

w[n]
 BASIC OPERATIONS
Example 2.1
Consider the two following sequences both defined for 0 < n
<4

{a[n]} = {3, 4, 6, - 9, 0}

{b[n]} = {2, -1, 4, 5, - 3}

Generate following new sequences from the above two sequences

{c[n]} = {a[n]b[n]}

{d[n]}= {a[n] + b[n]}

{e[n]} = 3/2{a[n]}
 BASIC OPERATIONS
Solution 2.1
operations are as follows:

{c[n]} = {a[n] b[n]}

= {3 x 2, 4 x- 1, 6 x 4, - 9 x 5, 0 x - 3 }
= {6, -4, 24, -45, 0}

{d[n]} = {a[n] + b[n]}

= {3 + 2, 4 + - 1, + 4, - 9 + 5, 0 + - 3 }
= {5, 3, 10, - 4, - 3}

{e[n]} = 3/2{a[n]}
= {4.5, 6, 9, 13.5, 0}
 BASIC OPERATIONS
operations on two or more sequences can be carried out
ONLY IF
all sequences involved are of same length
and
defined for the same range of the time index n

However if sequences are not of same length, in some

situations, this problem can be circumvented by

Appending zero-valued samples to the sequence(s) of smaller

lengths to make all sequences have the same range of the
time index
 BASIC OPERATIONS
Example 2.2
Consider the sequence

{f [n]} ={- 2, 1, - 3} defined for 0 < n <2

{a[n]} = {3, 4, 6, - 9, 0} defined for 0 < n < 4

Generate following new sequence from the above two sequences

{g[n]} = {a[n]}+{f[n]}
 BASIC OPERATIONS
Solution 2.2
We cannot add the length-3 sequence to the length-5
sequence

Therefore first append with 2 zero-valued samples to {f[n]}

resulting in a length-5 sequence

Then

{f[n]} ={- 2, 1, -3, 0, 0}

Solution is as follows:

{g[n]} = {a[n]}+{ f[n]} ={1, 5, 3, - 9, 0}

 COMBINATIONS OF BASIC
OPERATIONS
Example 2.3
Draw a block doagram of following Discrete Time system

y[n] =1x[n] + 2 x[n - 1] + 3 x[n - 2] + 4 x[n 3]

Solution 2.3
 SAMPLING RATE ALTERATION
Employed to generate new sequence y[n] with sampling
rate FT higher or lower than that of sampling rate FT of
given sequence x[n] '
FT
R
Sampling rate alteration ratio is FT

If R > 1, the process called interpolation

Operation required to be carried out is called as Up sampling

If R < 1, the process called decimation

Operation required to be carried out is called as Down sampling
 UP SAMPLING
In up-sampling by an integer factor L > 1,
L-1 equidistant zero-valued samples are inserted by the up-
sampler between each two consecutive samples of the input
sequence x[n]
n
x for n 0, L, 2 L,....
xu [n] L

0 otherwise

x [n] L xu[n]

This operator is also called as sampling rate expander

 UP SAMPLING
Example of up sampling
 DOWN SAMPLING
In down-sampling by an integer factor M > 1,
every Mth samples of input sequence are kept and M -1 in-
between samples are removed

xu [n] x[nM ]

x [n] M xd[n]

This operator is also called as Sampling Rate Compressor

 DOWN SAMPLING
Example of down sampling
 CLASSIFICATION OF SEQUENCES
Based on Symmetry
Conjugate-symmetric sequence Vs Conjugate-anti symmetric
sequence

even sequence Vs Odd Sequence

Energy Vs Power sequence

Bounded, absolutely summable, square summable

 CONJUGATE-SYMMETRIC
SEQUENCE
Conjugate-symmetric sequence:
Sequence x[n] is called conjugate symmetric sequence if

x[n] = x *[-n]
i.e
Let x[n] = xre[n] + j xre[n] Then x*[n] = xre[n] - j xre[n]
If x[n] = xre[-n] - j xre[-n] Then x[n] is conjugate symmetric
sequence
 CONJUGATE-SYMMETRIC
SEQUENCE
Real Conjugate-symmetric sequence:
If x[n] is real and Conjugate-symmetric sequence then it is
called as even sequence
 CONJUGATE-ANTISYMMETRIC
SEQUENCE
Conjugate-anti symmetric sequence:
Sequence x[n] is called conjugate anti symmetric sequence if

x[n] = - x *[-n]

If x[n] is real and Conjugate-anti symmetric sequence then it is

called as odd sequence
 CLASSIFICATION BASED ON
SYMMETRY
From definition if conjugate-symmetric sequence is {x[n]},
then
x[0] must be a
real number

For a conjugate anti-symmetric sequence {y[n]},

y[0] must be
an imaginary number

For an odd sequence {w[n]},

w[0]
must be = 0
 DECOMPOSITION OF SEQUENCE
Any complex sequence can be expressed as a sum of its
conjugate-symmetric part and its conjugate-antisymmetric
part:

i.e x[n] = xcs [n] + xca[n]

where
1
2

xcs [n] x[ n ] x *
[ n ]

& 2.20
1
xca [n] x[n] x*[ n]
2
 DECOMPOSITION OF SEQUENCE
Example 2.4
Consider the following a finite length sequence

{g[n]} = {0, 1+j4, -2+j3, 4-j2, -5-j6, -j2, 3} defined for -3 < n < 3

Determine conjugate symmetric and antisymmetric component of {g[n]}

Solution 2.4
Step 1:- Find Conjugate sequence
{g * [n]} = {0, 1-j4, -2- j3, 4+j2, -5+j6, j2, 3}
Step2:- Find Time-reversed version
{g * [-n]} = {3, j2, -5+j6, 4+j2, -2- j3, 1- j4, 0}
Step3:- conjugate symmetric & antisymmetric component by equation
2.20
{gcs [n]} = 1/2 X ({g[n] + g* [-n]} )
= {1.5, 0.5+j3, -3.5+j4.5, 4, -3.5-j4.5, 0.5-j3, 1.5}
{gca [n]} = 1/2 X ({g[n] - g* [-n]} )
={-1.5, 0.5+j, 1.5-j1.5, - j2, -1.5-j1.5, -0.5-j, 1.5}
 DECOMPOSITION OF SEQUENCE
Any real sequence can be expressed as a sum of its even
part and its odd part:

x[n] = xev[n] + xod [n]

where

1
xev [n] x[ n ] x[ n ]
2

& 2.22
1
xod [n] x[n] x[n]
2
 DECOMPOSITION OF SEQUENCE
Example 2.5
Consider the following a finite length sequence

{u[n]} = {1+j4, -2+j3, 4-j2, -5-j6} defined for 0 n 3

Determine conjugate symmetric and antisymmetric component of

{u[n]}

Solution 2.5
Step 1:- Conjugate sequence is
{u* [n]} = {1-j4, -2-j3, 4+j2, -5+j6}
Step2:- Time-reversed version is
Sequence is undefined for x[-3], x[-2], x[-1] ???????

Step3:- conjugate symmetric & antisymmetric component by

equation 2.20
????????
 DECOMPOSITION OF SEQUENCE
If length-N sequence x[n], is defined for 0 n N -1,
Then Above definitions of symmetry and Equation 2.20 are not
applicable

The definition of symmetry in case of finite one sided

sequence is given using MODULO Operation

Such symmetry is known as

periodic conjugate-symmetric
&
periodic conjugate-antisymmetric

Sequence x[n] can be expressed as :

i.e x[n] = xpcs [n] + xpca[n] for 0 < n < N -1

 MODULO OPERATION
Let
{u[n]} ={a1, a2, a3, a4, , aN} defined for 0 n N-
1
Graphical representation will be
a1 a2 a3 a4 aN

0 1 2 3 N-1

We can consider the sequence to be periodic beyond this range

Graphical representation will be
aN-2 aN-1 aN a1 a2 a3 a4 aN a1 a2 a3 a4

-3 -2 -1 0 1 2 3 N-1 N N+1 N+2 N+3

Now u[<-0>N]=u[0], u[<-1>N]=u[N-1], u[<-2>N]=u[N-2] & so

on
 MODULO OPERATION
Solution 2.5
Step 1:- Conjugate sequence is
{u* [n]} = {1-j4, -2-j3, 4+j2, -5+j6}
Step2:- Time-reversed version of the above is

1-j4 -2-j3 4+j2 -5+j6 1-j4 -2-j3 4+j2 -5+j6

-4 -3 -2 -1 0 1 2 3

u* [<-0>4] = u* [0] = 1-j4

u* [<-1>4] = u* [3] = -5+j6

u* [<-2>4] = u* [2] = 4+j2

u* [<-3>4] = u* [1] = -2-j3

Hence
{u* [<-n>4]} = {1-j4, -5+j6, 4+j2, -2-j3}
 DECOMPOSITION OF SEQUENCE
If length-N sequence x[n], is defined for 0 n N -1,
Then
Sequence x[n] can be expressed as :

i.e x[n] = xpcs [n] + xpca[n] for 0 n N -1

1
x pcs [n]
2
x[ n ] x *
[ n N
]

& 2.24
1
x pca [n] x[n] x*[ n N ]
2
 MODULO OPERATION
Solution 2.5
{u[n]} = {1+j4, -2+j3, 4-j2, -5-j6}
Step 1:- Conjugate sequence is then given by
{u* [n]} = {1-j4, -2-j3, 4+j2, -5+j6}
Step2:- Time-reversed version of the above is
{u* [<-n>4]} = {1-j4, -5+j6, 4+j2, -2-j3}
Step3:- conjugate symmetric & antisymmetric component by
equation 2.24
{upcs [n]} = 1/2 X ({u[n] + u* [<-n>4]} )
= { 1, - 3.5+J4.5, 4, -3.5-J4.5}
{upca [n]} = 1/2 X ({u[n] - u* [<-n>4]} )
={J4, 1.5-J1.5, -J2, -1.5-J1.5}

It can be easily verified that

upcs [n] = u*pcs [<-n>4]
& upca [n] = -u*pca [<-n>4]
 PERIODIC SYMMETRY FOR REAL
SEQUENCE
For a real sequence,

The periodic conjugate symmetric part, is a real sequence

And is called the periodic even part xpe[n]

The periodic conjugate antisymmetric part, is a real sequence

And is called the periodic odd part xpo[n]
PERIODIC & APERIODIC SEQUENCE
A sequence (n) is called a periodic sequence with a period
N if it satisfyes
(n)=x (n+kN) for all n
where N is a positive integer and k is any integer

Smallest value of N satisfying above condition is called the

fundamental period

A sequence not satisfying the periodicity condition is called

an aperiodic sequence
 ENERGY & POWER SEQUENCE
Total energy of a sequence x(n) is defined by
2
x x[n]
n

An infinite length sequence with finite sample values may or

may not have finite energy
Example
1n , n 1 1 , n 1
Let x1[n] Let x1[n] n
0 n 0 0 n 0
2

Energy ( x1 ) 1
n
Energy ( x1 ) 1n
n 1 n 1
2 This does not converges
This converges to
6 i.e
i.e x1[n] has infinite energy
infinite length sequence x1[n] has finite energy

A finite length sequence with finite sample values has finite

energy
 ENERGY & POWER SEQUENCE
The average power of a periodic sequence with period N is
given by 2
1 N1
Average Power ( Px ) ~x (n)
N n 0

The average power of an aperiodic sequence

2 is defined by
1 K
Average Power ( Px ) lim x[n]
K 2 K 1 n K

The average power of an aperiodic sequence can be related

to its power by defining the 2
K energy over a finite interval K
< n <K x K x[n]
n K
Then
1
Px lim xK
K 2K 1
 ENERGY & POWER SEQUENCE
Infinite energy signal with finite power is called Power
signal
A perodic sequence which has a finite average power but
infinite energy

Finite energy signal with zero average power is called

energy signal
Finite length sequence which has finite energy but zero
average power
 OTHER CLASSIFICATION
A sequence x(n) is said to be bounded if each of its sample is
of magnitude less than or equal to a finite positive number B x

x[n] Bx

A sequence x(n) is said to be absolutely summable if

x[n]
n

A sequence x(n) is said to be square summable if

2
x[n]
n

Square summable sequence has a finite energy & is energy signal if

it has zero power
 TYPICAL SEQUENCES
Unit Sample Sequence (n) (discrete time impulse) is defined
by
1, n 0
[n]
0, n 0
-5 -4 -3 -2 -1 0 1 2 3 4 5

Unit sample sequence shifted by k sample is given by

1, n k
[n k ]
0, n k
-5 -4 -3 -2 -1 0 1 2 k

Any arbitrary sequence can be represented as sum of weighted

(n) vs. (t)
time shifted unit sample sequence
Their energies are both equal to 1.
Certain class of discrete time system is completely characterized in
time domain by its output response to a unit impulse input
But [n] is of engineering value,
Knowing a unit impulse response of a system, we can calculate its
Where
responseas (t)
to any only
arbitrary has
input meaning in theory
sequence
 TYPICAL SEQUENCES
Unit Step Sequence denoted by [n] defined by
1, n 0
[n]
0, n0 -5 -4 -3 -2 -1 0 1 2 3 4 5

Unit step sequence shifted by k sample is given by

1, n k
[n k ]
0, nk -5 -4 -3 -2 -1 0 1 2 3 4 5

Unit sample sequence and Unit step sequence are related

as follows:-
n
[ n] [ k ] [ n k ]
k k 0

[n] [n] [n 1]
 TYPICAL SEQUENCES
Real sinusoidal sequence
x[n]=Acos(0 n+ ), - < n <
where A is the amplitude,0 is the angular frequency, and is the phase
of x(n)
Can be alternately written as
x[n]=xi[n]+xq[n]
where xi[n] = Acos( )cos(0 n) & xq[n] = -Asin( )sin0 n) are inphase
and quadrature component of x[n]

x[n] A n sequence
Exponential where A & can be real or complex
In general we can write A A e j & e( 0 j0 )
x[n] A e j e n ( 0 j0 ) xre [n] jxim [n]
Where xre [n] A e 0 cos( 0 n )
xim [n] A e 0 sin( 0 n )
COMPLEX EXPONENTIALS
 USE OF TYPICAL SEQUENCES
An arbitrary sequence can be represented in the time-
domain as a weighted sum of some basic sequence and its
delayed (advanced) Versions

x(n) x(k ) (n k )
k

Another interpretation is that the above equation can be

viewed as a convolution of x(n) and (n)
 SAMPLING PROCESS
Often Discrete time signal is obtained by uniform sampling
of a continuous time signal xa(t)
Relation between the two is given by
x n xa(t) t nT xa(nT), for n ..., 2, 1,0,1,2,...
Where continuous time variable t is related to discrete time variable n
only at discrete time instant tn . The relationship is given by
n 2n Where Sampling frequency FT 1/T
t n nT
FT T and Angular Sampling frequency T 2FT
For example
Let continuous time signal be
xa (t) A cos(2 f 0t ) A cos( 0t )

The corresponding discrete time signal is

x[n] A cos(0 nT )
20
A cos n
T
A cos 0 n ,
20
Where 0 T 0T
 SAMPLING PROCESS
Example 2.5
Let continuous time signal be a cosine function of 3 Hz frequency. Derive a
discrete time sequence obtained by sampling it at 10Hz
Solution 2.5
x(t)=cos(2ft)=cos(6t) x[n]=cos(2f n/FT)=cos(0.6n) for n=0,1,2,

x[n]={1, -0.3090, -0.8090, 0.8090, 0.3090, -1)

 SAMPLING PROCESS
Example 2.5 (contd)
Now let there be another continuous time signal in the form of cosine function of
7 Hz frequency. Derive a discrete time sequence obtained by sampling it at 10Hz
Solution 2.5
x(t)=cos(2ft)=cos(14t) x[n]=cos(2f n/FT)=cos(1.4n) for n=0,1,2,

x[n]={1, -0.3090, -0.8090, 0.8090, 0.3090, -1)

 ALIASING
From above example
In both cases, the sequences generated are identical, & hence
unique Continuous time function can not be associate to this
sequence.
Same sequence will be generated for cosine function of
frequency 13Hz
In fact all cosine functions of frequencies (10k3)Hz where k is
any nonnegative integer will lead to the same sequence.
The family of continuous time sinusoidal
In general we can say
xa, k (t) A cos 0t kT t ,
For k 0 , 1, 2,...
Leads to identical sampled sequence : -
xa, k (t)
A cos 0 kT nT
A cos

2 0 kT n


T

2 0
A cos n

T

A cos 0n
x[n]
 ALIASING
In other word
for cos(n), = 2r 0 all r appear same after sampling
i.e We say that a larger appears aliased to a lower frequency

Phenomenon of continuous time sinusoidal signal of

higher frequency acquiring identity of a sinusoidal
sequence of lower frequency after sampling is called
ALISING
2 0
0
But T

If T > 2| 0|
then 0 will be in range - < < implying no aliasing
If T < 20
then 0 will fold into lower digital frequency implying aliasing
Hence to prevent aliasing, Principal value of digital time
frequency should be 0 | 0| (i.e. less than one half cycle per
sample)
PART B:- DISCRETE-TIME
SYSTEM
LINEAR SYSTEM
SHIFT (TIME)-INVARIANT
SYSTEM
LINEAR TIME-INVARIANT
SYSTEM
CAUSAL SYSTEM
STABLE SYSTEM
PASSIVE AND LOSSLESS
SYSTEMS
 DISCRETE TIME SYSTEM
A discrete-time system processes
a given input sequence x[n]
to generates an output sequence y[n]
with more desirable properties

Discrete-time system can be

2-input, 1-output discrete-time systems
Modulator, adder
1-input, 1-output discrete-time systems -
Multiplier, unit delay, unit advance is a single-input, single-output
system:
Output sequence is generated sequentially beginning with
certain value on n (say n0)

In practical discrete-time system, all signals are digital &

output generated are also digital. Such systems are usually
called as digital filters
 DISCRETE TIME SYSTEMS:-
EXAMPLES 1
Accumulator
Cumulatively adds, i.e., it accumulates all input sample values
Output y[n] at time instant n is
sum of input sample x[n] at time instant n and
previous output y[n-1] at time instant n-1
which is the sum of all previous input sample values from - to n-1

n
y[n] x[l ]
l
n 1

l
x[l ] x[n]
y[n 1] x[n]

Input-output relation can also be written in the form

1 n n
y[ n] x[l ] x[l ] y[ 1] x[l ]
l l 0 l 0

This form is used for a causal input sequence, in which case y[-
1] is called the initial condition
 DISCRETE TIME SYSTEMS:-
EXAMPLES 2
M-point moving-average Filter
Used in smoothing random variations in data

1 M 1
y[n] x[ n k ]
M k 0

An application
Consider a signal s[n] corrupted by noise d[n] for n0. Resulting in
measured data x[n]=s[n]+d[n]
Let s[n] = 2[n(0.9)n ], d[n] - random signal
 DISCRETE TIME SYSTEMS:-
EXAMPLES 3
Linear interpolation
Employed to estimate sample values between pairs of
adjacent sample values of a discrete-time sequence
An application
Factor-of-4 interpolation
For 2 factor interpolation
1
y[n] xu [n] xu [n 1] xu [n 1]
0 1 . 2 3 2

0 1 . 2 3

0 1 . 2 3
 DISCRETE TIME SYSTEMS:-
EXAMPLES 4
Median Filter
 DISCRETE-TIME SYSTEMS:
CLASSIFICATION
Linear System

Shift-Invariant System

Causal System

Stable System

Passive and Lossless Systems

 LINEAR DISCRETE-TIME SYSTEMS
Definition
If
y1[n] is output due to input x1[n]
and
y2[n] is output due to input x2[n]

then
For input x [n] = x1[n] + x2[n]

Output is given by y[n] = y1[n] + y2[n]

Above property must hold for any arbitrary constants and

and for all possible inputs x1[n] and x2[n]

i.e systems obey superposition

LINEAR DISCRETE-TIME SYSTEMS
LINEAR DISCRETE-TIME SYSTEMS
 SHIFT-INVARIANT SYSTEM
Definition
If
if y[n] is the response to an input x[n]

then
For input x[n] = x1[n - no]

Output is given by y[n] = y1[n - no] where no is any positive or negative

integer

Above relation must hold for any arbitrary input x[n] & its
corresponding output y[n]

In case of sequences & systems with indices n related to discrete

instants of time, above property is called time-invariance
property

Time-invariance property ensures that for specified input, output is

independent of the time the input is being applied
SHIFT-INVARIANT SYSTEM
SHIFT-INVARIANT SYSTEM
 LINEAR TIME-INVARIANT (LTI)
SYSTEM
System satisfying both linearity and time-invariance
property

LTI systems are mathematically easy to analyze and

characterize, and consequently, easy to design

Highly useful signal processing algorithms have been

developed utilizing this class of systems over the last
several decades
 CAUSAL SYSTEM
Definition
noth output sample y[no] depends only on
input samples x[n] for n no
and does not depend on input samples for n > no

Let y1[n] and y2[n] be the responses of a causal discrete-time

system to the inputs x1[n] and x2[n], respectively
Then
x1[n] = x2[n] for n < N
implies also that
y1[n] = y2[n] for n < N

For a causal system, changes in output samples do not

precede changes in the input samples
CAUSALITY EXAMPLE
 STABLE SYSTEM
Definition
There are various definitions of stability

We consider here bounded-input, bounded-output (BIBO)

stability

If y[n] is the response to an input x[n] and if

x[n] Bx for all values of n

Then

y[n] By for all values of n

STABLE SYSTEM
 PASSIVE AND LOSSLESS SYSTEMS
A discrete-time system is defined to be passive if, for
every finite-energy input x[n], the output y[n] has, at
most, the same energy,
2 2
i.e. y[n] x[n]
n n

For a lossless system, the above inequality is satisfied

with an equal sign for every input

Example
Consider the discrete-time system defined by y[n] = a x[n - N]
with N a positive integer
Its output energy is given by
2 2
2
y[n] x[n]
n n

Hence, it is a passive system if ||< 1 and is a lossless system

if ||<=1
PART C:-
TIME-DOMAIN CHARACTERIZATION
OF LTI
IMPULSE AND STEP RESPONSES
SIMPLE INTERCONNECTION
SCHEMES
STABILITY CONDITION
CAUSALITY CONDITION
 IMPULSE AND STEP RESPONSES
The response of a discrete-time system to a unit sample
sequence {d[n]} is called the unit sample response or
simply, impulse response, and is denoted by {h[n]}

The response of a discrete-time system to a unit step

sequence {m[n]} is called the unit step response or
simply, the step response, and is denoted by {s[n]}
 IMPULSE RESPONSE
Example Find out the impulse response of the system

y[n] = a1x[n] +a2 x[n -1]+a3x[n - 2]+a4 x[n - 3]

Solution:-
Impulse response is obtained by setting x[n] = [n] resulting in

h[n] = a1[n]+a2 [n-1] +a3[n - 2] +a4 [n -3]

Thus in this case, the impulse response is a finite-length

sequence of length 4 given by

{h[ [n ]} {a1, a2, a3, a4}

 IMPULSE RESPONSE
Example
The impulse response of the discrete-time accumulator
n
y[n] x[l ]
l

is obtained by setting x[n] = [n] resulting in

n
y[ n] [l ]
l

The impulse response {h[n]} of the factor-of-2 interpolator

1
y[n] xu [n] xu [n 1] xu [n 1]
2

{h[n]} ={-0.5, 1, 0.5}

TIME-DOMAIN CHARACTERIZATION
Input-Output Relationship -
A consequence of the linear, time invariance property is that
an LTI discrete time system is completely characterized by its
impulse response

Knowing the impulse response one can compute the output of

the system for any arbitrary input

Let h[n] denote the impulse response of a LTI discrete-time

system

We compute its output y[n] for the separate inputs [n]:

As the system is linear, we can compute its outputs for each

member of the input separately and add the individual outputs
to determine y[n]
 INPUT-OUTPUT RELATIONSHIP
Let h(n) be a impulse response [n] h[n]

Recall that x(n) can be written x[n] x[ k ] [ n k ]
as k

[n k ] h[n k ]
Since system is time invariant

x[k ] [n - k ] x[k ]h[n - k ]

Likewise, as system is linear
Note that, x[k] is considered as constant here

x[k ] [n - k ] x[k ]h[n - k ]

By of linearity property of k k

y[n] x[k ]h[n - k ] h[k ]x[n - k ]
Hence I-O relationship of an LTI k k
system can be written as
 LINEAR CONVOLUTION SUM
The above operation is called Linear Convolution
Sum and can be represented compactly as

y[n]=x[n]*h[n]

The linear convolution sum has

Commutative property

x[n] * h[n] = h[n] * x[n]

Associative property

(x[n] * h[n]) * y[n] = x[n] * (h[n] * y[n])

Distributive property

h[n] * (x[n] + y[n]) = h[n] * x[n] + h[n] * y[n]

 RESPONSE OF A LSI SYSTEM
Passing a signal through a LSI
system is equivalent to
convolving it with the systems
impulse response
x [n] h[n] y[n]= x[n]*h[n]

y[n] x[k ]h[n - k ] h[k ]x[n - k ]
k k
 LINEAR CONVOLUTION SUM

y[ n] x[ k ]h[n - k ] h[ k ]x[n - k ]
k k
Solution
Time-reverse h(k) to form h( k)
Shift h( k) to right by n sampling periods if n > 0 or shift
h( k) to the left by n sampling periods if n <0 to form h(n
k)
Form the product v(k)=x(k)h(n k)
Sum all samples of v(k) to develop nth sample of y(n)

If the lengths of x(n) and h(n) are N1 and N2 respectively,

the length of the convolution sum y(n) will be N1+N2 1
 INTERPRETING CONVOLUTION
Example:- Let
h[n] = {3 2 1} & x[n]={0 3 1 2 -1} y[n] x[k ]h[n - k ]
k

0 1 2 3 4 5
0 1 2 3 4 5

Time-reverse h,

-5 -4 -3 -2 -1 0
= 0x3 = 00 1 2 3 4 5
shift by n,
9

Take inner product against fixed x

-2 -1 -0
= 0x2 + 3x3 = 00 1 2 3 4 5
-4 -3 1

9 9

-3 -2 -1 0 1 2
= 0x1+ 3x2+1x3 = 00 1 2 3 4 5
 INTERPRETING CONVOLUTION
Example:- Let
h[n] = {3 2 1} & x[n]={0 3 1 2 -1}

y[n] x[k ]h[n - k ]
0 1 2 3 4 5
0 1 2 3 4 5 k 11

9 9

-2 -1 0 1 2 3
= 3x1+1x2+2x3 = 0
0 1 2 3 4 5
11

9 9

1 2 3
= 1x1 + 2x2+-1x3
= 0
0 1 2 3 4 5
-1 0 4
11

9 9

0 1 2 3 4 5
= 2x1+ -1x2
=0 0
1 2 3 4 5
0
 INTERPRETING CONVOLUTION
Example:- Let
h[n] = {3 2 1} & x[n]={0 3 1 2 -1}

y[n] x[k ]h[n - k ] {0,9,9,11,2,0, 1}
k

In practice the convolution sum can be employed to

compute the output only if either the impulse response
sequence and / or the input sequence is of finite length
 MATRIX INTERPRETATION

y[n] x[n-k]h[k]
k
y[0] x[0] x[ 1] x[ 2] . . h[0]
y[1] x[1] x[0] x[ 1] . . h[1]

y[2] x[2] x[1] x[0] . . h[2]

. . . . . . .
. . . . . . .
 APPLICATION OF CONVOLUTION
Example:- Let
h[n] = {3 2 1} & x[n]={0 3 1 2 -1}y[n] x[k ]h[n - k ]
k

0 1 2 3 4 5

11

99 9

2
0 0
0 1 2 3 4 5
 APPLICATION OF CONVOLUTION
Using Convolution to calculate the Correlations of sequence

Correlation of Signals
Definition
A measure of similarity between a pair of energy signals, x(n) and
y(n), is given by the cross-correlation sequence rxy(l) defined by

rxy [l ] x[n] y[n l ], for l 0,1,2...
n

If x(n)=y(n), we obtain the definition of the autocorrelation of x(n)

rxx [l ] x[n]x[n l ], for l 0,1,2...
n
Note that rxx(0) = x2[n] , is the energy of the signal x(n)
 SIMPLE INTERCONNECTION
Cascade Connection

Impulse response h[n] of the cascade of two systems with

impulse responses h1[n] and h2[n] is h(n)=h1(n)h2(n)

By commutativity,

h(n)=h1(n)h2(n)=h2(n)h1(n)
An application is in the development of an inverse system

Inverse system:-System which recovers input x[n] from

output y[n] (e.g. to undo effects of transmission channel)

i.e If impulse response of channel is known, then x(n) can be

recovered by designing an inverse system of channel
 INVERSE SYSTEMS
[n] is identity for convolution i.e.
x[n] * [n] = x[n]

If cascade connection satisfies the relation

h1(n)h2(n)= (n)
Then LTI system h1(n) is said to be inverse of h2(n) and vice-versa

Consider

z[n] = h2[n] * y[n] = h2[n] * h1[n] * x[n]

If z[n] = x[n]

then h2[n] *h1[n] =[n]

h2[n] is the inverse system of h1[n]

INVERSE SYSTEM EXAMPLE
 SIMPLE INTERCONNECTION
Parallel Connection
Impulse response of two parallel systems added together i.e:
h(n)=h1(n)+h2(n)

The parallel connection of two stable systems is stable

However, the parallel connection of two passive systems

may or may or may not be passive
 STABILITY CONDITION OF AN LTI
BIBO Stability Condition
A discrete time system is BIBO stable if the output sequence {y[n]}
remains bounded for all bounded input sequence {x[n]}
An LTI discrete-time system is BIBO stable if and only if its
impulse response sequence {h[n]} is absolutely summable, i.e.

S h[n]
n
Proof:
Assume h[n] is a real sequence

Since the input sequence x[n] is bounded we have |x[n]| Bx <

Therefore

y[n] h[k ]x[n k ] h[k ] x[n k ] Bx h[ k ] Bx S
k k k
Thus, S < implies, |y[n]|By < indicating y[n] is also bounded
 CAUSALITY CONDITION OF LTI
SYSTEM
Let x1[n] and x2[n] be two input sequences with
x1[n] = x2[n] for n no
The corresponding output samples at no of an LTI system with an
impulse response {h[n]} are then given by
1
y1[n0 ] h[k ]x1[n0 k ] h[k ]x1[n0 k ] h[k ]x1[n0 k ]
k k0 k

y1 [ n ] for n n0 y1 [ n ] for n n0
1
y2 [n0 ] h[k ]x 2 [n0 k ] h[k ]x 2 [n0 k ] h[k ]x 2 [n0 k ]
k k 0 k

Since system is causal y1[n] = y2[n] for n no

Hence for y1[n0] = y2[n0] second term on RHS of above two
euation should also be equal.
But x1[n] may not be equal to x2[n] for n > no Hence the only
way to satisfy these two equations is if
h[k] = 0 for k <0
An LTI discrete-time system is causal if and only if its impulse
response {h[n]} is a causal sequence
 CAUSALITY CONDITION OF LTI
SYSTEM
Example 1
The discrete-time system defined by
y[n] = a1x[n]+ a2x[n -1] + a3x[n - 2] + a4x[n- 3]
is a causal system as it has a causal impulse response is
{h[n]} ={a1 a2 a3 a4}
Example 2
The discrete-time
n accumulator defined by
y[n] [l ] [ n]
l
is a causal
n
system as it has a causal impulse response given by
h[n] [l ] [n]
l

Example 3
The factor-of-2 interpolator defined by
1 1
y[n] xu [n] xu [ n 1] xu [n 1]
2 2
is non-causal as it has a non-causal impulse response given by

h[n] 0.5, 1, 0.5


FINITE-DIMENSIONAL LTI SYSTEMS
An important subclass of LTI discrete-time systems is
characterized by a Linear Constant Coefficient Difference
Equation
N
of Mthe form
Where
d k y[ n k ] Pk x[n k ]
k 0 k 0 x[n] & y[n] are input & output of system

And d k & Pk are constants characterizing system

Order of system is max(N,M), which is order of difference

equation
Ify[we assume
N d the system
M p to be causal, then the output y[n]
n] k y[n k ] k x[n k ] Provided d 0 0
can be krecursively
1 d 0 computed
k1 d0 using

y[n] can be computed for all n n0 , knowing x[n] and the

initial conditions y[n0-1], y[n0-2], , y[n0-N],
SOLUTION OF LCCDE APPROACH 1
It is possible to implement an LTI system characterized by a
LCCDE
Even though such systems have an impulse response of
infinite length, computation of output response involves two
finite sums of products
These components are independently computed and added to
yield a total solution
y[n]
Thus total solution yc[n]
can be written as:- y [ n]
p

Complementary Solution Particular Solution
N
satisfies d k y[ n k ]0 for given forcing functin x[n]
k 0

i.e
Complementary solution is solution with x[n]=0
&
Particular solution is solution with x[n]0
 COMPLEMENTARY SOLUTION
General form of unforced oscillation i.e. systems natural
modes
Assume y has form
c
yc [n] n
N n k
dk 0
k 0
n N
d d .... d
0
N
1
N1
N 1
d 0
N
N
This is of the form d k N k
k 0

Thus
N N k
dk 0 Is known as Characteristic polynomialof system
k 0

Let 1, 2, ., N be N distinct roots Then the general form of

complementary solution is given by
yc[n]= 11n+ 22n+ + NNn
Where 1, 2, N are constants determined from specified initial
condition
 PARTICULAR SOLUTION
Particular solution reflects input
If x[n] is a constant then yp[n] is also assumed to be constant
 LCCDE
Example
Determine the total solution for n 0 of DT system characterized
by
y[n]+ y[n-1]-6y[n-2] = x[n] ---------Eq 1
For step input: x[n] = 8[n] and initial conditions y[-1]=1, y[-2]=
-1
 LCCDE
Solution
First determine complementary solution
by setting x[n]=0 & y[n]=n
we arrive at n + n-1 - 6n-2 = n-2 (2 + - 6) = n-2 ( + 3)( - 2)=0
Hence roots of Ch Polynomial are 1=-3, 2=2
Hence complementary solution is of form yc[n]= 1(-3)n+ 2(2)n

For particular solution

we assume yp[n]=
Substituting in Eq 1 + -6 =8[n] i.e n 0 = -2
Hence Total solution y[n]= 1(-3)n+ 2(2)n - 2
By using initial conditions 1 & 2 can be found out
At n=0 At n=1
y[0]+ y[-1] -6y[-2] = x[0] y[1]+ y[0] -6y[-1] = x[0]
1 + 2 2 +1 + 6 = 8 1(-3)+2(2)2+1+2 6 =8
1 + 2 = 3 -21 + 32 = 18
1 = -1.8, 2 = 4.8
SOLUTION OF LCCDE APPROACH 2
Zero Input Response / Zero State Response

LCCDEs is to solve separately for following twos conditions:

yzi[n], response with zero input (just ICs)
yzs[n], response with zero state (just x[n] & all initial conditions as
0)

Because of linearity, y[n] = yzi[n]+yzs[n]

Both solutions are real But, have to solve for i twice

 LCCDE
Example
Characterizing difference equation of DT system y[n]+y[n-1]-6y[n-2] =
x[n]
initial conditions y[-1]=1, y[-2]= -1 . y[n] for step input: x[n] = 8[n]
Solution
First determine Zero input response
Complementary solution (is found as per previous method) is of form
yc[n]= 1(-3)n+ 2(2)n.
y[0] = -y[-1] + 6y[-2] = -1 6 = -7 = 1+ 2.
y[1] =-y[0] + 6y[-1] = +7 + 6 = 13= -31 + 22.
Hence 1 = -5.4 & 2 = -1.6 & yzi[n] = -5.4(-3)n+ -1.6 (2)n.
Now for zero state response
Total solution (as per previous method) is
y[n]= 1(-3)n+ 2(2)n - 2
y[0] = x[0] - ???? + ???? = 8 = 1+ 2 - 2
y[1] = x[1] - y[0] + ???? = 0 = -31+ 22n - 2
Hence 1 = 3.6 & 2 = 6.4 & yzs[n] = 3.6(-3)n+ 6.4(2)n 2.

Therefore total solution is y[n] = yzi[n]+yzs[n]

IMPULSE RESPONSE CALCULATION
Impulse response {h[n]} of DT system is systems response when
x[n]= [n]
Which is nothing but a Zero State response with x[n]= [n]
It means x[n]=0 for n > 0 Hence yp[n]=0
Impulse response can be calculated from Complimentary Solution by
determining i for zero initial condition

In previous Example
complementary solution is yc[n]= 1(-3)n+ 2(2)n for n0. Hence IR is
h[n]=1(-3)n+ 2(2)n
If x[n]= [n] then
h[0] = 1 + 2 = 1 These values are obtained by using
& h[1] =-31+22 = 0 y[n]+ y[n-1]-6y[n-2] = x[n]

Therefore 1 = 0.6 & 2 = 0.4

Hence h[n]= 0.6(-3)n+ 0.4 (2)n
 STABILITY BASED ON ROOTS OF
CH EQUATION
Impulse response samples of stable LTI system decays to
zero as n becomes large
However it is not possible to check stability just by
examining only finite samples of impulse response.
BIBO stability of LTI characterized by LCCDE can be inferred
from the values of roots i of its characteristic equation
Establishing Stability criterion :-
Impulse response is same as complementary solution
So for Ncase of distinct roots we can say
h[n] i in
i 1

For system to be stable S h[ n ]
n
N N
n n
h[ n ] i i i i
n0 n 0 i 1 i 1 n 0

n
If i 1 for all i Then i Hence h[ n ]
n 0 n 0
 CLASSIFICATION OF LTI SYSTEMS
Based on Impulse Response Length
Finite impulse response (FIR)
infinite impulse response (IIR)

Based on the Output Calculation Process

Non-recursive System
Recursive System

Based on the Coefficients

Real Discrete-Time System
Complex Discrete-Time System
 CLASSIFICATION BY IR LENGTH
Finite impulse response (FIR)
If the impulse response h[n] is of finite length, then system is
known as a finite impulse response (FIR) DT system
i.e h[n] = 0 for n < N1 and n > N2, N1 < N2
The convolution sum description here is
N2
y[n] h[k ]x[n k ]
k N1
Output y[n] of FIR system can be computed directly from
convolution sum as it is a finite sum of products
Examples :- Moving-average system and linear interpolators

Infinite impulse response (IIR)

If impulse response is of infinite length, then system is known
as infinite impulse response (IIR) DT system
The class of IIR systems we are concerned with in this course
are characterized by linear constant coefficient difference
equations
 CLASSIFICATION BY OUTPUT
CALCULATION
Non-recursive System
Output can be calculated sequentially, knowing only present
and past input samples

Recursive System
Output computation involves past output samples in addition
to present and past input samples
 CLASSIFICATION BY COEFFICIENTS
Real Discrete-Time System
The impulse response samples are real valued

Complex Discrete-Time System

The impulse response samples are complex valued
 PROBLEMS
Problem 2.1
Determine L1-norm, L2-norm, L-norm of following finite length
sequence
{x[n]}={4.50 ,-2.68, -0.14, 3.91, 2.62, -0.43, -4.81, 3.21, -0.55}

Solution
Lp-normn of a sequence
1 {x[n]} is defined by
p
p
x p x[n]
n p is +ve
Where
integer. Typical values of p are 1,2,
x2 x
RMS value of x[ n] & mean absolut value of x[ n]
N N
x
x max peak absolute value of {x[n]}

Norm provides a estimate of the size of a signal

In above example
L1-norm =
 PROBLEMS
Solution 2.1 contd/.

n= 1 2 3 4 5 6 7 8 9 |x[n]|p

x[n] 4.50 -2.68 -0.14 3.91 2.62 -0.43 -4.81 3.21 -0.55

|x[n]| 4.50 2.68 0.14 3.91 2.62 0.43 4.81 3.21 0.55 22.85 L1-norm=22.85

|x[n]|2 20.25 7.18 0.02 15.29 6.86 0.18 23.14 10.30 0.30 83.53 L2-norm=9.14

9.59 1.66 8.30 6.71 5.56 1.05 2.10 9.46 1.13 2.10
|x[n]|1000 L-norm=4.81E+00
E+97 E+64 E-129 E+88 E+62 E-55 E+102 E+75 E-39 E+102
 PROBLEMS
Problem 2.5
Consider following sequence
x[n] = { -4 5 1 -2 -3 0 2 } -3 n 3

y[n] ={ 6 -3 -1 0 8 7 -2 } -1 n 5
Generate following (a) {c [n]}=x[-n+2]; (b) {u[n]}= x[n]+y[n-
2]
Solution 2.5
(a)
x[n] = { -4 5 1 -2 -3 0 2 } -3 n 3
-n+2= 5 4 3 2 1 0 -1
C[n]=x[-n+2] 0 0 2 0 -3 -2 1

C[n]= { 1 -2 -3 0 2 0 0 } -1 n 5

(b) y[n] = { 0 0 6 -3 -1 0 8 } -1 n 5
n= -3 -2 -1 0 1 2 3

x[n]= { -4 5 1 -2 -3 0 2 } -3 n 3
{u[n]}=
x[n]+y[n-2] {
-4 5 7 -5 -4 0 10 } -3 n 3
 PROBLEMS
Problem (~) 2.7
Analyze block diagram of LTI discrete-time system of following figure and
develop the relation between y[n] and x[n].
v[n] v[n-1] v[n-2]

w[n] u[n]

Solution (~) 2.7

v[n]=x[n]w[n], w[n]=d 1v[n1]+d2u[n], & u[n]=v[n2]+x[n].
From these we get
w[n]=d2x[n]+d1x[n1]+d2x[n2]d1w[n1]d2w[n2].
From the figure we also obtain
y[n]=v[n2]+w[n]=x[n2]+w[n]w[n2],
d1y[n1]=d1x[n3]+d1w[n1]d1w[n3],
and d2y[n2]=d2x[n4]+d2w[n2]d2w[n4],
y[n]=d2x[n]+d1x[n1]+x[n2]d1y[n1]d2y[n2].
 PROBLEMS
Problem (+) E1
Let {y[n]} = {1 1 11 3 30 28 48} be a sequence
obtained by a linear convolution of a finite-length sequence {x[n]}
With the sequence {h[n]}= {1 2 3 4 }
First sample in each sequence is time instant n = 0.
Determine x[n].

Solution (+) E1
The length of x[n] should be is 7 4 + 1 = 4.
M N M N
y[n] x[k ]h[n - k ] or h[k ]x[n - k ]
k 0 k 0

M N 1
y[n] x[n]h[0] h[k ]x[n - k ]
k 1
1 M N 1
x[n] y[ n ] h[ k ] x[ n - k ] Thus we get x[n] = {1 3 2 12} 0n
h[0]
k 1

3
 PROBLEMS
Problem 2.64
Determine expression for IR of LTI discrete-time system shown bellow:

Solution 2.64
We can write
v[n] = (h1[n] + h3[n] * h5[n]) * x[n]
and y[n] = h2[n] * v[n] + h3[n] * h4[n] * x[n].
Thus, y[n] = (h2[n] * h1[n] + h2[n] * h3[n] * h5[n] + h3[n] * h4[n])
* x[n].
Hence the impulse response is given by
h[n] = h2[n] * h1[n] + h2[n] * h3[n] * h5[n] + h3[n] * h4[n].