DIGITAL SIGNAL PROCESSING

 A signal is defined as any physical quantity that varies with time, space or
another independent variable.
 A system is defined as a physical device that performs an operation on a signal.
 System is characterized by the type of operation that performs on the signal.
Such operations are referred to as signal processing.

Advantages of DSP
1. A digital programmable system allows flexibility in reconfiguring the digital
signal processing operations by changing the program. In analog redesign of
hardware is required.
2. In digital accuracy depends on word length, floating Vs fixed point arithmetic
etc.In analog depends on components.
3. Can be stored on disk.
4. It is very difficult to perform precise mathematical operations on signals in
analog form but these operations can be routinely implemented on a digital
computer using software.
5. Cheaper to implement.
6. Small size.
7. Several filters need several boards in analog, whereas in digital same DSP
processor is used for many filters.

Disadvantages of DSP
1. When analog signal is changing very fast, it is difficult to convert digital form
(beyond 100KHz range)
2. w=1/2 Sampling rate.
3. Finite word length problems.
4. When the signal is weak, within a few tenths of millivolts, we cannot amplify
the signal after it is digitized.
5. DSP hardware is more expensive than general purpose microprocessors & micro
controllers.
6. Dedicated DSP can do better than general purpose DSP.

Applications of DSP
1. Filtering.
2. Speech synthesis in which white noise (all frequency components present to the
same level) is filtered on a selective frequency basis in order to get an audio signal.
3. Speech compression and expansion for use in radio voice communication.
4. Speech recognition.
5. Signal analysis.
6. Image processing: filtering, edge effects, enhancement.
7. PCM used in telephone communication.
8. High speed MODEM data communication using pulse modulation systems such
as FSK, QAM etc. MODEM transmits high speed (1200-19200 bits per second)
over a band limited (3-4 KHz) analog telephone wire line.
9. Wave form generation.

SIGNAL
A signal is a function of one or more independently variables that usually
represent time and/ or space, A signal contains some kind of information that can
be conveyed, displayed, or manipulated.

a. One dimensional- when the function depends on a single variable.

e.g. speech signal

b. Multidimensional- when the function depends on two or more variable.

e.g. Image signal

A. SYSTEM

It is formally defined as an entity that manipulates one or more

signals to accomplish a function, thereby yielding new signal.

Application

1. Communication system
2. Control system
3. Microelectromechanical system
4. Remote sensing
5. Biomedical signal processing
6. Auditory system
 Four Types Of Data:

1. Text- set of characters

2. Audio or speech
3. Image or photos
4. Video- combination of audio and images

Advantages of digitization

1. Flexibility- you can use digital signals to different platforms. For example,
you can transfer and use your data from one computer to another
computer.
2. Repeatability – easy to retrieve information

Steps in converting Analog to Digital Signal

1. Sampling
– is the acquisition of a continuous signal at discrete time intervals.

-how many samples needed at a certain time

2. Quantizing

– process of putting samples

3. Coding

- assigning binary equivalent values

NYQUIST THEOREM

-It provides a prescription for the nominal sampling interval required to avoid
aliasing. The sampling frequency could be at least twice the highest frequency
contained in the signal

fs≥2fc
where in:
fs is the sampling frequency (unit : time/space) and fc is the highest frequency
contained in the signal.

CLASSIFICATION OF SIGNALS

1. Continuous and Discrete Signal

Continuous Time signal

–a signal x(t) is said to be continuous time signal if it is defined for all time t.

Discrete Time Signal

-is defined only at discrete

- usually obtained from continuous-time signal by sampling it on a uniform

rate

Assume: Ts- sampling period

n- integer that may assume to be positive and negative vales

x[n]= x(nTs), n=0,±1,±2,±3...

2. EVEN AND ODD SIGNALS

Even Signals
- x(-t)=x(t) for all t
- symmetric about the vertical axis

Odd Signals

- x(-t)= -x(t) for all t

- Anti-symmetrical about the time origin
 3. Periodic and Non-Periodic Signals

For continious time signal

A periodic signal x(t) is a function of time that satisfies the condition
x(t)=x(t+T) for all t.
Where T= positive constant

For discrete time signal

A discrete time signal x(n) is said to be periodic if x(n)= x(n+N) for
integer n.

Deterministic and Random Signal

Deterministic Signal

It is a signal ab out which there is no uncertainty with respect to its value at

any time.

Random Signal

It is a signal about which there is uncertainty before it occurs.

Discrete-Time Signals and Discrete-Time Systems

 Discrete-Time Signals
A discrete-time signal x ( n ) is a function of an independent variable that is
an integer. It is important to note that a discrete-time signal is not defined at
instants between two successive samples.
 Figure 1 :Graphical representation of a discrete-time signal.

Some alternative representations of a discrete-time signal:

1. Functional representation, such as

2. Tabular representation, such as

3. Sequence representation

An infinite-duration signal or sequence with the time origin (n = 0) indicated

by the symbol f is represented as

 A sequence x ( n ) , which is zero for n < 0. can be represented as

A finite-duration sequence can be represented as

 Some Elementary Discrete-Time Signals

1. The unit sample sequence: is denoted as ( n ) and is defined as

Note: Unit Impulse in analog signals

Figure 2: Graphical representation of unit sample signal.

2. The unit step signal is denoted as u(n) and is defined as

Figure 3: Graphical representation of unit step signal.

3. The unit ramp signal is denoted as ur(in) and is defined as

Figure 4: Grapical represntation of unit ramp signal.

4. The exponential signal is a sequence of the form

If the parameter a is real, then x ( n ) is a real signal.

When the parameter a is complex valued. it can be expressed as

where r and  are now the parameters. Hence we can express x ( n ) as

Figure 5: Grapical represntation of exponential signal.
 Classification of Discrete-Time Signals

Energy signals and power signals. The energy E of a signal x ( n ) is defined as

Many signals that possess infinite energy, have a finite average power. The average
power of a discrete-time signal x ( n) is defined as

If we define the signal energy of x( n ) over the finite interval -N ≤ n ≤ N as

then we can express the signal energy E as

and the average power of the signal x( n ) as

Periodic signals and aperiodic signals.

A signal x( n ) is periodic with period N ( N > 0) if and only if

The smallest value of N is called the (fundamental) period. If there is no value of N

that satisfies equation above, the signal is called non periodic or aperiodic.

Symmetric (even) and antisymmetric (odd) signals. A real-valued signal

x( n ) is called symmetric (even) if

On the other hand, a signal x(n) is called antisymmetric (odd) if

The even signal component is formed by adding x( n ) to x(-n) and dividing

by 2. that is.
Figure 7: Example of even and odd signals.

Similarly, we form an odd signal component xo(n) according to the relation

Now, if we add the two signal components, we obtain x ( n ) , that is,

 Simple Manipulations of Discrete-Time Signals

1. Transformation of the independent variable (time).

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 a positive integer, the time shift results in a delay of
the signal by k units of time. If k is a negative integer, the time shift results in an
advance of the signal by |k| units in time.

Example .A signal x ( n ) is graphically illustrated in Fig.a. Show a graphical

representation of the signals x(n - 3) and x(n + 2).

2. Addition, Multiplication, and Scaling of Sequences

Addition: The sum of two signals

is formed by the pointwise addition of the signal values.

Multiplication The multiplication of two signals

is formed by the pointwise product of the signal values.

Amplitude scaling of a signal x(n) by a constant c is accomplished by multiplying

every signal value by c:

This operation may also be considered to be the product of two signals, x(n) and f
(n) = c.

 DISCRETE-TIME SYSTEMS

A discrete-time system is a device or algorithm that operates on a discrete-time

signal, called the input or excitation, according to some well-defined rule, to
produce another discrete-time signal called the output or response of the system.
In general, we view a system as an operation or a set of operations performed on
the input signal x(n) to produce the output signal y ( n ) . We say that the input
signal x ( n ) is transformed by the system into a signal y ( n ), and express the
general relationship between x( n ) and y(n) as

where the symbol T denotes the transformation (also called an operator), or

processing performed by the system on x( n ) to produce y ( n ) .
 Figure 1: Block diagram representation of discrete time system.

 Input-Output Description of Systems

The input-output description of a discrete-time system consists of a
mathematical expression or a rule, which explicitly defines the relation between
the input and output signals (input-output relationship).

Example.

1. Determine the response of the following systems to the input signal

2. The accumulator described by (A) is excited by the sequence x(n)=nu(n).

Determine its output under the condition that:

a.) It is initially relaxed [i.e., y(-1)=0]

b.) Initially y(-1)=1

 Block Diagram Representation of Discrete-Time Systems

 An adder
An adder performs the addition of two signal sequences to form another (the sum)
sequence, which we denote as y(n).

Figure 2: Graphical representation of an adder.

 A constant multiplier
This operation is depicted by Fig. below, and simply represents applying a scale
factor on the input x ( n ) .

Figure 3: Graphical representation of a constant mutiplier.

 A signal multiplier
Figure below illustrates the multiplication of two signal sequences to form another
(the product) sequence, denoted in the figure as y ( n ) .
Figure 4: Graphical representation of a signal mutiplier.

 A unit delay element

The unit delay is a special system that simply delays the signal passing through it by
one sample.

Figure 5: Graphical representation of a unit delay element.

 A unit advance element

In contrast to the unit delay, a unit advance moves the input x( n ) ahead by one
sample in time to yield x(n + 1).

Figure 2: Graphical representation of a unit advance element.

Example:

Using basic building blocks introduced above. Sketch the block diagram
representation of the discrete-time system described by the input-output
relation.

where x ( n ) is the input and y ( n ) is the output of the system.

Solution: The output y ( n ) is obtained by multiplying the Input x ( n ) by 0.5,

multiplying the previous input x( n - 1 ) by 0.5. Adding the two products, and then
adding the previous output y(n - 1) multiplied by 1/4. Figure a, illustrates this block
diagram realization of the system.

leads to the block diagram realization shown in Fig. b. Note that if we treat "the
system" from the "viewpoint" of an input-output or an external description.
 Classification of Discrete-Time Systems
 Static versus dynamic systems.
A discrete-time system is called static or memoryless if its output at any instant n
depends at most on the input sample at the same time, but not on past or future
samples of the input. In any other case, the system is said to be dynamic or to have
memory.

The systems described by the following input-output equations

are both static or memoryless. On the other hand, the systems described by the
following input-output relations

are dynamic systems or systems with memory.

  Time-invariant versus time-variant systems.
A system is called time-invariant if its input-output characteristics do not
change with time.

A relaxed system T is Time invariant or shift invariant if and only if

implies that

for every input signal x ( n ) and every time shift k. In general, we can write the
output as

Now if this output y(n, k) = y(n - k), for all possible values of &, the system is time
invariant. On the other hand, if the output y(n, k ) # y(n - k), even for one value of
k, the system is time variant.

Example:

Determine if the systems shown in Fig. below are time invariant or time variant.
Solution

(a) This system is described by the input-output equations

Now if the input is delayed by k units in time and applied to the system, it is clear
from the block diagram that the output will be

On the other hand if we delay y(n) by k units in time. We obtain

v ( n . k ) = y(n - k ) . Therefore, the system is time invariant.

 Linear versus nonlinear systems.

A linear system is one that satisfies the superposition principle. Simply
stated, the principle of superposition requires that the response of the system to a
weighted sum of signals be equal to the corresponding weighted sum of the
responses (outputs) of the system to each of the individual input signals.

A relaxed T system is linear if and only if

for any arbitrary input sequences xl(n) and xz(n), and any arbitrary constants a1
and a2.
Figure 7: Graphical representation of the superposition principle. T is linear if and
only if y(n)=y’(n).

Determine if the systems described by the following input-output equations arc

linear or nonlinear.

Solution

(a) For two input sequences x1( n ) and a2(n), the corresponding outputs are

A linear combination of the two input sequences results in the output

On the other hand. a linear combination f the two outputs result in the output

Since they are identical, the system is linear.

 Causal versus Non Causal Systems.

A system is said to be causal if the output of the system at any time depends only
on present and past outputs, but does not depend on future inputs. In
mathematical terms, the output of a causal system satisfies an equation of the form

Where F[.] is some arbitrary function.

If a system does not satisfy this definition, it is called noncausal.

Example:

Determine if the systems described by the following input-output equations are

causal or noncausal.

(f)
  Stable versus unstable systems.
An arbitrary relaxed system is said to be bounded input-bounded output
(BIBO) stable if and only if every bounded input produces a bounded output.
The condition that the input sequence x(n) and the output sequence y(n) are
bounded is translated mathematically to mean that there exist some infinite
numbers, say MX and My , such that

[x(n)] ≤ MX ≤ ∞, [y(n)] ≤ My ≤ ∞,

For all values of n. If for some bounded input sequence x(n), the output is
unbounded (infinite). The system is classified as unstable.

Example:

Consider the nonlinear system described by the input-output equation

y(n)= y2(n-1) + x(n)

Interconnection of Discrete-Time Systems

Discrete-time systems can be interconnected to form larger systems. There are two
basic ways in which systems can be interconnected: in cascade (series) or in
parallel.
In the cascade interconnection the output of the first system is

Y1(n)=T1[x(n)]

and the output of the second system is

We observe that systems T1 and T2 can be combined or consolidated into a single
overall system

Consequently, we can express the output of the combined system as

In general, the order in which the operations T1 and T 2 are performed is important.
That is,

for arbitrary systems.

In the parallel interconnection, the output of the system ;T1; is y1( n ) and the
output of the system T2 is y2(n). Hence the output of the parallel interconnection
is
ANALYSIS OF DISCRETE-TIME LINEAR TIME-INVARIANT SYSTEMS
TECHNIQUES FOR THE ANALYSIS OF LINEAR SYSTEMS

There are two basic methods for analyzing the behavior or response of a
linear system to a given input signal. One method is based on the direct solution
of the input-output equation for the systems, which in general, has the form:

y(n)=F[y(n-1),y(n-z),...,y(n-N),x(n),x(n-1),...,x(n-M)]

where f [] denotes some function of the quantities in the brackets. Specifically, for
an LTI system, the general form of the input-output relationship is:

y(n)=− ∑𝑵 𝑴
𝒌=𝟏 𝒂𝒌 𝒚(𝒏 − 𝒌) + ∑𝒌=𝟎 𝒃𝒌 𝒙(𝒏 − 𝒌) ........1

where {Ak} and (Bk} are constant parameters that specify the system, and are
independent of an x(n) and y(n). The input-output relationship in (1) is called
difference equation.

The second method for analyzing the behavior of a linear system to a given
input signal is first to decompose or resolve the input signal into a sum of
elementary signals.
Now suppose that the signal x(n) is resolved into a weighted sum of elementary
signal components {Xk(n)} so that:

𝒙(𝒏) = ∑𝒌 𝑪𝒌 𝑿𝒌 (𝒏)
where {Ck) are the set of amplitudes (weighting coefficients) in the decomposition
of the signal x(n).

Now suppose that the response of the system to the elementary signal
component 𝑋𝑘 (𝑛) is 𝑦𝑘 (𝑛) . Thus

𝒚𝒌 (𝒏) = 𝝉[𝑿𝒌 (𝒏)]

Assuming that the system is relaxed and that the response to (n) is 𝐶𝑘 𝑋𝑘 (𝑛), as a
consequence of the scaling property of the linear system.

Finally, the total response to the input 𝑋(𝑛) is

𝒚(𝒏) = 𝝉[𝑿(𝒏)] = 𝝉[∑𝒌 𝑪𝒌 𝑿𝒌 (𝒏)]

=∑𝒌 𝑪𝒌 𝝉[𝑿𝒌 (𝒏)]

= ∑𝒌 𝑪𝒌 𝑿𝒌 (𝒏)

RESOLUTION OF A DISCRETE-TIME SIGNAL INTO IMPULSES

Suppose we have an arbitrary signal 𝑥(𝑛) that we wish to resolve into a sum
of unit sample sequences. We now select the elementary signal 𝑋𝑘 (𝑛) to be

𝑿𝒌 (𝒏) = 𝜹(𝒏 − 𝒌)
where k represents the delay of the unit sample sequences. If we multiply the two
sequences 𝑥(𝑛) and 𝛿(𝑛 − 𝑘), we have

𝒙(𝒏) 𝜹(𝒏 − 𝒌) = 𝒙(𝒌)𝜹(𝒏 − 𝒌)

Which is a sequence that is zero everywhere. Except @ 𝑛 = 𝑘, where its value is
𝑥(𝑘)
FIGURE 1: Multiplication of a signal 𝑥(𝑛) with a shifted unit sample sequence.

Consequently, if we repeat the multiplication of the signal 𝑥(𝑛) by a shifted unit

impulse over all possible delays, −∞ < 𝑘 < ∞, and sum of all the product
sequences, the result will be a sequence equal to the sequence 𝑥(𝑛), that is,

𝒙(𝒏) = ∑∞
𝒌=−∞ 𝒙(𝒌)𝜹(𝒏 − 𝒌)
where the right-hand side of the equation gives the resolution or decomposition
of any arbitrary signal 𝑥(𝑛) into a weighted (scaled) sum of shifted unit sample
sequence.

Example:

Consider the special case of a finite-duration sequence given as:

Resolve the sequence 𝑥(𝑛) into a sum of weighted impulse sequence.

Answer:

RESPONSE OF LTI SYSTEM TO AN ARBITRARY INPUTS: THE CONVOLUTION

SUM

To determine the response of any relaxed linear system to any input signal, we
denote the response 𝑦(𝑛, 𝑘) of the system to the input unit sample sequence at
𝑛 = 𝑘 by the special symbol ℎ(𝑛, 𝑘), −∞ < 𝑘 < ∞. That is

𝒚(𝒏, 𝒌) = 𝒉(𝒏, 𝒌) = 𝝉[𝜹(𝒏 − 𝒌)] (A)

In (A) we note that n is the time index and k is a parameter showing the
location of the input impulse. If the impulse at the input is scaled by an amount
𝐶𝑘 = 𝑥(𝑘), the response of the system is the correspondingly scaled output, that
is,

𝑪𝒌 𝒉(𝒏, 𝒌) = 𝒙(𝒌)𝒉(𝒏, 𝒌)
Now, if the input is the arbitrary signal 𝑥(𝑛) that is expressed as a sum of
weighted impulses, that is,

𝒙(𝒏) = ∑∞
𝒌=𝟎 𝒙(𝒌)𝜹(𝒏 − 𝒌)
then the response of the system to 𝑥(𝑛) is the corresponding sum of weighted
outputs, that is,

𝒚(𝒏) = 𝝉[∑∞
𝒌=−∞ 𝒙(𝒌)𝜹(𝒏 − 𝒌)]

= ∑∞
𝒌=−∞ 𝒙(𝒌)𝝉[𝜹(𝒏 − 𝒌)] (B)

= ∑∞
𝒌=−∞ 𝒙(𝒌)𝒉(𝒏, 𝒌)

Clearly, (B) follows from the superposition property of linear systems, and is
known as the superposition summation. Thus (B) applies to any relaxed linear
(time-variant) system.

If the system is time invariant, the formula in (B) simplifies considerably. In fact,
if the response of the LTI system to the unit sample sequence 𝛿(𝑛) is denoted as
ℎ(𝑛), that is

𝒉(𝒏) = 𝝉[𝜹(𝒏)]
then by the time-invariance property, the response of the system to the delayed
unit sample sequence 𝛿(𝑛 − 𝑘) is

𝒉(𝒏 − 𝒌) = 𝝉[𝜹(𝒏 − 𝒌)]

Consequently, the formula in (B) reduces to

𝒚 (𝒏) = ∑ ∞
𝒌=−∞ 𝒙(𝒌)𝒉(𝒏 − 𝒌) (C)

The formula in (C) that gives the response 𝑦(𝑛) of the LTI system as a function of
the input signal 𝑥(𝑛) and the unit sample (impulse) response ℎ(𝑛) is called a
convolution sum.

Suppose that we wish to complete the output of the system at same time instant,
say 𝑛 = 𝑛0 . According to (c), the response at 𝑛 = 𝑛0 is given as

𝒚(𝑵𝒐 ) = ∑∞
𝒌=−∞ 𝒙(𝒌)𝒉(𝑵𝒐 − 𝒌)
 To summarize, the process of computing the convolution between 𝑥(𝑘) and
ℎ(𝑘) involves the following four step:

1. Folding. Fold ℎ(𝑘) about 𝑘 = 0 to obtain ℎ(−𝑘).

2. Shifting. Shift ℎ(−𝑘) by 𝑁𝑜 to the right(left) if 𝑁𝑜 is positive(negative), to

obtain ℎ(𝑁𝑜 − 𝑘)

3. Multiplication. Multiply 𝑥(𝑘) by ℎ(𝑁𝑜 − 𝑘) to obtain the product sequence

𝑉𝑁𝑜 (𝑘) = 𝑥(𝑘)ℎ(𝑁𝑜 − 𝑘)

4. Summation. Sum all the values of the product sequence 𝑉𝑁𝑜 (𝑘) to obtain the
value of the output at the time 𝑛 = 𝑁𝑜

Examples:

The impulses response of a linear time-invariant system is

Determine the response of the system to the input signal

Answer:

PROPERTIES OF CONVOLUTION AND THE INTERCONNECTION OF LTI

SYSTEMS
It is convenient to simplify the notation by using an asterisk to denote the
convolution operation. Thus,
 IDENTITY AND SHIFTING PROPERTIES
The unit sample sequence 𝛿(𝑛) is the identity element for convolution,
that is

𝒚(𝒏) = 𝒙(𝒏) ∗ 𝜹(𝒏) = 𝒙(𝒏)

If we shift 𝛿(𝑛) by k, the convolution sequence is shifted also by k, that is

𝒙(𝒏) ∗ 𝜹(𝒏 − 𝒌) = 𝒚(𝒏 − 𝒌) = 𝒙(𝒏 − 𝒌)

 COMMUTATIVE LAW

𝐱(𝐧) ∗ 𝐡(𝐧) = 𝐡(𝐧) ∗ 𝐱(𝐧)

(Interpretation of the commutative property of convolution)

 ASSOCIATIVE LAW

[𝒙(𝒏) ∗ 𝒉(𝒏)] ∗ 𝒉𝟐 (𝒏) = 𝒙(𝒏) ∗ [𝒉𝟏 (𝒏) ∗ 𝒉𝟐 (𝒏)]

𝒚(𝒏) = 𝒚𝟏 (𝒏) ∗ 𝒉𝟐 (𝒏)

 = [𝒙(𝒏) ∗ 𝒉𝟏 (𝒏)] ∗ 𝒉𝟐 (𝒏)

If 𝒉(𝒏) = 𝒉𝟏 (𝒏) ∗ 𝒉𝟐 (𝒏)

Then 𝒚(𝒏) = 𝒙(𝒏) ∗ 𝒉(𝒏)

Implications of the associative (a) and the associative and commutative (b)
properties of convolution.

 DISTRIBUTIVE LAW
𝒙(𝒏) ∗ [𝒉𝟏 (𝒏) + 𝒉𝟐 (𝒏)] = 𝒙(𝒏) ∗ 𝒉𝟏 (𝒏) + 𝒙(𝒏) ∗ 𝒉𝟐 (𝒏)

Interpretation of the distributive property of convolution: Two LTI systems

connected in parallel can be replaced by a single system with ℎ(𝑛) =
ℎ1 (𝑛) + ℎ2 (𝑛).
 CAUSAL LINEAR TIME-INVARIANT SYSTEMS
For a linear time-invariant system, causality can be translated to a
condition on the impulse response. To determine this relationship, let us
consider a linear time invariant system having an output at time 𝑛 = 𝑛𝑜
given by the convolution formula

𝒚 (𝒏𝒐 ) = ∑ ∞
𝒌=−∞ 𝒉(𝒌)𝒙(𝒏𝒐 − 𝒌)
= ∑∞ −𝟏
𝒌=𝟎 𝒉(𝒌)𝒙(𝒏𝒐 − 𝒌) + ∑𝒌=−∞ 𝒉(𝒌)𝒙(𝒏𝒐 − 𝒌)

= [𝒉(𝒐)𝒙(𝒏𝒐 ) + 𝒉(𝟏)𝒙(𝒏𝒐 − 𝟏) + 𝒉(𝟐)𝒙(𝒏𝒐 − 𝟐)

+ 𝒉(𝟑)𝒙(𝒏𝒐 − 𝟑) + … . . ] + [𝒉(−𝟏)𝒙(𝒏𝒐 + 𝟏)
+ 𝒉(−𝟐)𝒙(𝒏𝒐 + 𝟐) + ⋯ ]

If the output at time 𝑛 = 𝑛𝑜 is to depend only on the present and past

inputs, the, clearly, the impulse response of the system must satisfy the
condition
𝒉(𝒏) = 𝟎, 𝒏<𝟎

Hence an LTI System is causal if an only if its impulse response is zero for
negative values of 𝑛 .Thus for a causal system, ℎ(𝑛) = 0 for 𝑛 < 0, we have
the two equivalent forms

𝒏
𝒚 (𝒏) = ∑ ∞
𝒌=𝟎 𝒉(𝒌)𝒙(𝒏 − 𝒌) = ∑𝒌=−∞ 𝒙𝟗𝒏)𝒉(𝒏 − 𝒌)

It is sometimes convenient to call a sequence that is zero for < 0 , a

causal sequence, and one that is nonzero for 𝑛 < 0 and 𝑛 > 0, a non causal
sequence .
If the input to a causal LTI System is a causal sequence [𝑥(𝑛) =
0 𝑓𝑜𝑟 𝑛 < 0], the two equivalent forms of the convolution formula become

𝒚(𝒏) = ∑𝒏𝒌=𝟎 𝒉(𝒌)𝒙(𝒏 − 𝒌)

 = ∑𝒏𝒌=𝟎 𝒙(𝒌)𝒉(𝒏 − 𝒌)

Clearly, the response of a causal system to a causal input sequence is

causal, since 𝑦(𝑛) = 0 for 𝑛 < 0.

 STABILITY OF LINEAR TIME-INVARIANT SYTEMS

If 𝑥(𝑛) is bounded, there exist a constant Mx such that
I 𝒙(𝒏)I≤ 𝑴𝒙 < ∞

Similarly, if the output is bounded, there exist a constant My such that

I 𝒚(𝒏)I< 𝑴𝒚 < ∞ for all n

Let us consider the convolution formula

𝒚 (𝒏) = ∑ ∞
𝒌=−∞ 𝒉(𝒌)𝒙(𝒏 − 𝒌)

If we take the absolute value of both sides of this equation, we obtain

I 𝒚 (𝒏) I = I ∑ ∞
𝒌=−∞ 𝒉(𝒌)𝒙(𝒏 − 𝒌) I

Now, the absolute value of the sum of terms is always less than or equal
to the sum of the absolute values of the terms. Hence

I 𝒚 (𝒏) I ≤ ∑ ∞
𝒌=−∞|𝒉(𝒌)||𝒙(𝒏 − 𝒌)|

If the input is bounded, there exist a finite number Mx such that |𝑥(𝑛)| ≤
𝑀𝑥 .
 |𝒚(𝒏)| ≤ 𝑴𝒙 ∑∞
𝒌=−∞|𝒉(𝒌)|

From this expression, the output is bounded if the impulse response of

the system satisfies the

𝑺𝒏= ∑∞
𝒌=−∞|𝒉(𝒌)| < ∞

That is , a linear time-invariant system is stable if its response is

absolutely summable.

 SYSTEMS WITH FINITE-DURATION AND INFINITE-DURATION IMPULSE

RESPONSE
A Finite-Duration Impulse Response (FIR) system has an impulse
response that is zero outside of some finite time interval. Without loss of
generality, we focus on causal FIR systems, so that

𝒏(𝒏) = 𝟎, 𝒏 < 𝟎 𝒂𝒏𝒅 𝒏 ≥ 𝑴

The convolution formula for such a system reduces to

𝒚(𝒏) = ∑𝑴−𝟏
𝒌=𝟎 𝒏(𝒌)𝒙(𝒏 − 𝒌)

For this expression, the output at any time n is simply a weighted linear
combination of the input signal samples x(n), x(n-1),.....,x(n-M+1). FIR
system has a finite memory f length -M samples.
In contrast, an IIR linear time-invariant system has an infinite-duration
impulse response. Its output, based on the convolution formula, is

𝒚 (𝒏) = ∑ ∞
𝒌=𝟎 𝒉(𝒌)𝒙(𝒏 − 𝒌)
Examples:
1. If 𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛), show that ∑ 𝑦 = ∑ 𝑥 ∑ ℎ, where ∑ 𝑥 =
∑∞𝑛=−∞ 𝑥(𝑛).
Answer:
∑𝑛 𝑦(𝑛) = ∑𝑘 ℎ(𝑛) ∑ℎ 𝑥(𝑛)
2. Compute the convolution 𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛) of the following signals
and check the correctness of the results by using the test in (1).

a) 𝑥(𝑛) = {1,2,4} , ℎ(𝑛) = {1,1,1,1,1}

↑ ↑
Answer: 𝑦(𝑛) = {… . ,0,1,3,7,7,7,6,4,0, … }
↑
∑𝑛 𝑦(𝑛) = 35
b) 𝑥(𝑛) = {1,2, −1} , ℎ(𝑛) = 𝑥(𝑛)
↑
Answer:𝑦(𝑛) = {… ,0,1,4,2, −4,1,0, … }
∑𝑛 𝑦(𝑛) = (4)
c) 𝑥(𝑛) = {1,1,2} , ℎ(𝑛) = 𝑢(𝑛)
↑

Answer: 𝑦(𝑛) = 𝑢(𝑛) + 4(𝑛 − 1) + 2𝑢(𝑛 − 2)

3. Determine and sketch the convolution of the signals

1 2 4 5
Answer: 𝑥(𝑛) = {0, , , 1, , , 2}
3 3 3 3
↑
ℎ(𝑛) = {1,1,1,1,1}
↑
𝑦(𝑛) = 𝑥(𝑛) ∗ ℎ(𝑛)
 1 10 20 11
={ , 1,2, , 5, , 6,5, , 2}
3 3 3 3

CORRELATION OF DISCRETE-TIME SYSTEMS

 CROSSCORRELATION AND AUTOCORRELATION SEQUENCES
Suppose that we have two real signal sequences 𝑥(𝑛) and 𝑦(𝑛) each
of which has finite energy. The cross correlation of 𝑥(𝑛) and 𝑦(𝑛) is a
sequence 𝑟𝑥𝑦 (𝑙), which is defined as

𝒓𝒙𝒚 (𝒍) = ∑∞
𝒏=−∞ 𝒙(𝒏)𝒚(𝒏 − 𝟏) , 𝒍 = 𝟎. ±𝟏, ±𝟐 …

or equivalently as,

𝒓𝒙𝒚 (𝒍) = ∑∞
𝒏=−∞ 𝒙(𝒏 + 𝟐)𝒚(𝒏) , 𝒍 = 𝟎. ±𝟏, ±𝟐 …

The index 𝑙 is the (time) shift (or lag) parameter and subscripts 𝑥𝑦 on
the cross correlation sequence 𝑟𝑥𝑦 (𝑙) indicate the sequence being
correlated.
Now if we reverse the order of the indices 𝑥𝑦, we obtain the cross
correlation sequence

𝒓𝒚𝒙 (𝒍) = ∑∞
𝒏=−∞ 𝒚(𝒏)𝒙(𝒏 − 𝒍)

or equivalently as,

𝒓𝒚𝒙 (𝒍) = ∑∞
𝒏=−∞ 𝒚(𝒏 + 𝟐)𝒙(𝒏)

By comparison, we conclude that

𝒓𝒙𝒚 (𝒍) = 𝒓𝒚𝒙 (−𝒍)

Therefore, 𝑟𝑦𝑥 is simply the folded version of 𝑟𝑥𝑦 (𝑙) , where the folding
is done with respect to 𝑙 = 0. Hence, 𝑟𝑦𝑥 (𝑙) provides exactly the same
information as 𝑟𝑥𝑦 (𝑙), with respect to the similarity of 𝑥(𝑛) to 𝑦(𝑛).
 For the special case where 𝑥(𝑛)= 𝑦(𝑛) we have the autocorrelation of
𝑥(𝑛), which is defined as the sequence

𝒓𝒙𝒙 (𝒍) = ∑∞
𝒏=−∞ 𝒙(𝒏)𝒙(𝒏 − 𝒍)

or equivalently as,

𝒓𝒙𝒙 (𝒍) = ∑∞
𝒏=−∞ 𝒙(𝒏 + 𝟐)𝒙(𝒏)

In dealing with finite-duration sequences, it is customary to express the

autocorrelation and cross correlation in terms of the finite limits on the
summation. In particular, if 𝑥(𝑛) and 𝑦(𝑛)are causal sequences of length
𝑁[𝑖. 𝑒., 𝑥(𝑛) = 𝑦(𝑛) = 0 𝑓𝑜𝑟 𝑛 < 0 𝑎𝑛𝑑 𝑛 ≥ 𝑁], the crosscorrelation
and autocorrelation sequence maybe expressed as

𝒓𝒙𝒚 (𝒍) = ∑𝑵−|𝒌|−𝟏

𝒏=𝒍 𝒙(𝒏)𝒚(𝒏 − 𝟏)

and

𝒓𝒙𝒙 (𝒍) = ∑𝑵−|𝒌|−𝟏

𝒏=𝒍 𝒙(𝒏)𝒙(𝒏 − 𝟏)

where 𝑖 = 𝑙, 𝑘 = 0 for 𝑙 ≥ 0, and 𝑖 = 0, 𝑘 = 𝑙 for 𝑙 > 0.

Examples:
Determine the autocorrelation sequence of the following signals.
a. 𝑥(𝑛) = {1,2,1,1}
↑
Answer: 𝑟𝑥𝑥 (𝑙) = {1,3,5,7,5,3,1}
↑
b. 𝑦(𝑛) = {1,1,2,1}
↑
Answer: 𝑟𝑥𝑥 (𝑙) = {1,3,5,7,5,3,1}
↑
THE 𝒁-TRANSFORM AND ITS APPLICATION TO THE ANALYSIS
OF LTI SYSTEMS

Transform techniques are an important tool in the analysis of signals and

linear time-invariant (LTI) systems. The 𝑧-transform plays the same role in the
analysis of discrete-time signals and LTI systems as the Laplace transform does in
the analysis of continuous-time signals and LTI systems.

THE DIRECT 𝒛-TRANSFORM

The 𝑧-transform of a discrete-time signal 𝑥(𝑛) is defined as the power series

𝒙(𝒛) = ∑∞
𝒏=−∞ 𝒙(𝒏)𝒛
−𝒏
(1)

where 𝐶 is a complex variable. (1) is sometimes called the direct 𝒛-transform

because it transforms the time-domain signal 𝑥(𝑛) into its complex-plane
representation 𝑋(𝑧). The inverse procedure [i.e.., obtaining 𝑥(𝑛) from 𝑋(𝑧) is
called the inverse 𝒛-transform.
For convenience, the 𝑧-transform of a signal 𝑥(𝑛) is denoted by
𝑿(𝒛) = 𝒁{𝒙(𝒏)}
whereas the relationship between 𝑥(𝑛) and 𝑋(𝑧) is indicated by
z
𝒙(𝒏) ↔ 𝑿(𝒛)
Since the 𝑧 -transform is an infinite power series, it exists only for those values of
𝑧 for which this series converges. The region of convergence (ROC) of X (𝑧) is the
set of all values of 𝑧 for which X (𝑧) attains a finite value. Thus any time we cite a
𝑧-transform we should also indicate its ROC.