Unit-II

FOURIER TRANSFORM

2.1.1.Introduction:
We can represent any periodic function x (t), over the entire interval (-∞<t<∞) as
a discrete sum of exponential functions. We can also represent any non-periodic
function in terms of exponential functions over any finite interval ( . For
the purpose of frequency analysis, we need to represent every type of driving function
in terms of exponential functions not over the finite interval but over the entire interval
(-∞, ∞).

2.1.2.Deriving the Fourier Transform from Fourier series:

Let us consider a function x(t) as shown in fig (1). We derive to represent this function
as a sum of exponential functions over the entire interval (-∞, ∞). For the purpose we
construct a new periodic function with period T as shown in fig (2). The period T
is made large enough so that there is no overlap between the pulses of the shape of x(t).
This new function is a periodic function and consequently can be represented with
an exponential fourier series.

Fig: 1 Fig: 2

In the limit if we let T becomes infinite, then the pulses in the periodic function repeat
after an infinite interval. Hence in the limit T→∞, and x(t) are identical i.e.

Thus the fourier series representing over the entire interval will also represent x(t)
over the entire interval if we let T = ∞ in this series.

Exponential fourier series for can be represented as

Where

And dt --------------------> (1)

The term represents the amplitude of the component of frequency . We shall now
assume that T becomes very large. As T becomes larger, (the fundamental frequency)
becomes smaller and the spectrum becomes denser.

As seen from equation (1) when T→∞, the magnitude of each component becomes
infinite similarly small but now there are also an infinite number of frequency
components. The spectrum exists for every value of ω, and is no longer discrete but a
continuous function of ω.

To illustrate this,

Let ----------------------> (2)

Then is a function of , and we shall denote by . Further let

T ----------------------> (3)

Then

= ---------------------> (4)

From equation (1) and (3) we have

∴ ------------------------> (5)

Substituting the value T = in equation (4) we get

 ---------------------> (6)

Equation (6) shows that can be expressed as a sum of exponential signals of

frequencies ω1, ω2, ω3, _ _ _ _ _ _, , _ _ _ _ _ etc. The amplitude of the component of
frequency is (this is equal to .

∴ The amplitude content in of frequency. is not but is proportional to

Let us try a graphical interpretation of equation (6), which represents a discrete sum or
a sum of discrete frequency components. Actually the quantity is complex
in general. However we shall assume that the quantity is real. The figure (3)
shows such a plot of this quantity as a function of ω. This function only at discrete
values of ω at ω1, ω2, ω3, _ _ _ _ _ _, , _ _ _ _ _ etc. Where

Fig: 3

Each frequency component is separated by distance . Therefore the area of the

shaded rectangle in fig (3) is evidently Equation (6) represents the sum
of areas under all such rectangular structure corresponds to n = -∞ to ∞. The sum of
rectangular areas represents approximately the area under the dotted curve. The
approximation becomes better as becomes smaller. In the limit when T→∞,
becomes infinite similarly small and may be represented by dω. The discrete sum in
equation (6) becomes the integral or the area under this curve.

The curve now is a continuous function of ω and is given by . Also at T→∞ the
function and equation (6) and (5) becomes.

x(t) = -----------------> (7)

Where X(ω) = -----------------------> (8)

∴ We can represent a non-periodic function x(t) in terms of exponential functions over

an entire interval (-∞, ∞). Equation(7) represents x(t) as a continuous sum of exponential
functions with frequencies lying in the interval (-∞, ∞). The amplitude of the component
of any frequencies ω is proportional to X(ω). Therefore, X(ω) represents the frequency
spectrum of x(t) and is called the spectral density function.

Equation (7) and (8) are usually referred as the Fourier Transform pair. Equation (8) is
known as the direct Fourier Transform of x(t) and equation (7) is known as Inverse
Fourier transform of X(ω).

Symbolically these transforms are also written as

X(ω) = FT [x(t)] = -----------------> (9)

IFT [X(ω)] = x(t) = = ------------------> (10)

2.1.3.The Fourier Transform of Standard Signals:

2.1.3.1. Single Sided Exponential Function

= t>0
x (t) =
0 t<0

1 t≥0

Because u(t) =
0 t<0

∴FT [x(t)] = X(ω) =

=
=

= =

∴FT [x(t)] = X(ω) = for a>0------------->(11)

F[ =

It follows that the inverse Fourier transform of is

x-1 = -----------------> (12)

Now = =

X(ω) = =j

And the phase function

θ (ω) =

=-

∴X(ω) = ----------------> (13)

∴It is evident from the above equation that magnitude function is an even function and
phase function is an odd function. The signal and its spectrum is shown in the fig 4 (a)
and 4(b) respectively.

Fig: 4(a) Fig: 4(b)

2.1.3.2.Double Sided Exponential Function :

t>0
x(t) =

t<0

FT [x(t)] = X(ω) =

= t

= t

= +

X(ω) =

∴ and θ(ω) = 0
The signal and its spectrum is shown in the fig 5 (a) and 5(b) respectively.

Fig: 5(a) Fig: 5(b)

2.1.3.3.

∴ FT [x(t)] = X(ω) =

= t

Integrating by parts yields.

X(ω) =

a > 0. X [t

∴ =t

The spectral density function X(ω) may be expressed in terms of magnitude and phase-
density functions.

X(ω) = =
2.1.3.4. A Gate function

A gate function is a rectangular pulse and is defined by as shown in figure 6.

1 |t|<

0 |t|>

Fig: 6(a)

The Fourier Transform of this function is given by

X(ω) =

=A

=A -------------> (16)

∴ X( ) is a real function and can be represented graphically as shown in figure 6(b)

Fig: 6(b)
2.1.4.Existence of Fourier Transform:

The Fourier Transform pair is given by

x(t) =

and X( ) =

For the function x(t) to be Fourier transformable, x(t) should satisfy Dirichlet‟s
conditions given below…

( i.)The function x(t) should be a single valued with a finite number of maxima and
minima and a finite number of discontinuities in any finite time interval

( ii.) The function x(t) is absolutely integrable i.e.

2.1.4.1.Merits of Fourier Transform:

Fourier Transform is most useful for analyzing signals involved in communication

systems. Some of the main advantages of Fourier Transforms are:

a.) The original time function can be uniquely recovered from it.

b.) It has a property analogous to common logarithm that helps in evaluating the
convolution integral.

c.) The phase and amplitude characteristics are readily known.

2.1.5.Singularity Function:

Consider a unit step voltage applied to a capacitor as shown in figure 7. The current i
through the capacitor is given by i = C

Fig: 7(a) Fig: 7(b)

It is easy to say that is zero for all values of t except at t=0 where it is undefined. The
derivation of t=0 does not exist because the function u(t) is discontinuous at this point.

If the same voltage in ramp signal as shown in figure 8. The current through the
capacitor would be a pulse as shown in figure 9.

Fig: 8 Fig: 9

The solution to an ideal unit step voltage does not exist but it is possible to obtain a
solution in the limit by assuming a non ideal source as shown in figure side by and then
letting a go to zero in the limit. Let us designate the un idealized voltage function by

ua (t). In the limit when a goes to zero, this voltage function becomes a unit step
function. The derivation of the function ua(t) is a rectangular pulse of height and
width a. As „a‟ varies, the pulse shape varies but the area of the pulse remains constant.
In the limit when a goes to zero, the height of the pulse goes to infinity and the width of
the pulse is zero. The area of the pulse however, remains unity. The figure 10 shows
that sequence of such pulses as „a‟ varies.

Fig: 10(a) Fig: 10(b)

As „a‟ tends to zero the resultant pulse or function is called as unit impulse function.
Therefore we define the unit impulse function as the derivative of unit step function.
The unit impulse function is denoted by δ(t).Therefore we have

-------(17)
But the function is a rectangular pulse of height and width a. This can be
described as

δ (t) = ---------------> (18)

In the limit as „a‟ goes to zero, the function δ (t) assumes the form of a pulse of infinite
height and zero width. The area of the pulse however remains unity. The function δ (t)
exists only at t =0 and has a zero value of all t ≠ 0. This fact can be expressed as.

δ (t) = 0 t ≠ 0

and -----------------> (19)

Equation (19) may be taken as a definition of the unit impulse function. This description
of an impulse function is more general and represents the impulse function as a limiting
process without any regard to the actual shape of the pulse; i.e. it is not necessary to
define the impulse function as a limiting form of Gaussian pulse, a triangular pulse, an
doubled or single sided exponential pulse, or any other pulse forms as shown in fig 11.

Fig: 11

2.1.6.Sampling or Shifting property of the Impulse Function: -

The function δ (t- is a unit impulse function which exists only at t= and is zero
everywhere else.

∴ For a given function x(t)

x(t) δ (t- = x( ) δ(t-

Hence multiplication of any function x(t) by an impulse δ (t- also yields an impulse
of strength x( ) at t= .

i.e.
This also follows

2.1.7.The Fourier Transform of an Impulse Function:-

Here x(t) =

∴F [x(t)] = F [ =

From the sampling property of impulse function

F[ =

Therefore the Fourier Transform of a unit impulse function is unity.

Fig: 12(a) Fig: 12(b)

2.1.8.The Fourier Transform of a constant: -

Let x(t) = A

In a gate function if z→∞, the gate function tends to a constant function A. The Fourier
Transform of a constant A is therefore the Fourier Transform of a gate function as
z→∞.

We have F

∴ F [A] =

= 2πA

= 2πAδ ( )

Similarly F [1] = 2πδ ( )

Thus, when x(t) equals a constant it contains only a frequency component of = 0. This
is a logical result, since a constant function is a d.c signal ( = 0), and does not have any
other frequency components.

Fig: 13(a) Fig: 13(b)

2.1.9.Fourier Transform of and :-

The functions and also do not satisfy the condition of absolute

integrability, yet their Fourier Transform exists. Consider a function . We shall
first consider this function to exist only over the interval to and zero outside
the interval. In the limit, may be made infinity.

∴ F [x(t)] = F [ =

=
=

=π

=π

∴ Similarly it can be shown that

Fig: 14(a) Fig: 14(b)

Fig: 15(a) Fig: 15(b)

2.1.10.Fourier Transform of Signum Function: -

The signum function is given by

1 t>0
sgn (t) =
-1 t<0

=
F [sgn (t)] =

= =

F [sgn (t)] =

∴ sgn (t) ↔

Fig: 16(a) Fig: 16(b)

2.1.11.Fourier Transform of unit step function: -

Unit step function is defined as

1 t≥0
u (t) =
0 t<0

The relation between Signum function and u (t) is

sgn (t) = 2u (t) – 1

∴ 2 u (t) = sgn (t) +1

u (t) =
∴F [u (t)] =

F [u (t)] =

Fig: 17(a) Fig: 17(b)

The spectral density function contains an impulse at = 0. Thus the function u(t)
contains a large d.c component as expected. In addition also has other frequency
component. The function u(t) is not a constant. Since it is zero for t<0, and there is an
abrupt discontinuity at t=0 giving rise to other frequency components.

For a true d.c signal, x(t) is constant for -∞<t<∞ . Such a signal indeed has no other
components except dc (i.e. = 0).

2.1.12.Fourier Transform of Eternal Exponential :-

Here we have x(t) =

= +j

∴F [ =F[ +j

=π

F[ =2π
The Fourier Transform of is therefore a single impulse of strength 2 π at = .
The signal is not a real function of time, and hence it has a spectrum which exists at
= alone.
2.1.13.The Fourier Transform of a Periodic Function: -

Strictly speaking, the Fourier Transform of a periodic function does not exist, since it
fails to satisfy the condition of absolute integrability. For a periodic function x(t):

But the transform does exist in the limit.

Alternatively, we may express a periodic function by is Fourier series. The Fourier

Transform of a periodic function is then the sum of Fourier Transform of its individual
components. We can express a periodic function x(t) with period T as

x(t) = where

Taking Fourier transform of both sides we get

F [x(t)] = F

F [x(t)] = 2π

(∵

The above relation states that the spectral density function or the Fourier Transform of a
periodic signal consists of impulses located at the harmonic frequencies of the signal
and the strength of each impulse is the same as times the value of the corresponding
coefficient in the exponential Fourier series.

2.1.14.Fourier Transform of a sequence of Equidistance impulses of unit strength and

separated by T seconds: -

Let us denote this function by a special symbol .

∴ = δ (t) + δ (t-T) + δ (t-2T) + _ _ _ _ _ _ _ + δ (t - nT) +_ _ _ _ _ _ _ +

δ (t + T)+ δ (t + 2T)+_ _ _ _ _ _ _ + δ (t + nT)+_ _ _ _ _ _ _

=
This is a periodic function with periodic T as shown in figure below.

Fig: 18

Fourier series of this function is given by

Where

Function in the interval is simply δ (t), hence

From the sampling (shifting) property of impulse function, the above equation becomes.

∴F

F
 The above relation is very significant. It states that the Fourier Transform of a
unit impulse train of period T is also a train of impulses of strength and separated by
radians ( Therefore, the impulse train function is its own transform.

Fig: 19(a) Fig: 19(b)

2.1.15.Properties of Fourier Transform:

The Fourier Transform is a tool for expressing a function in terms of its

exponential components of various frequencies. Fourier Transform of a function is just
another way of specifying the function. We therefore have two descriptions of the same
function: the time domain and frequency domain descriptions.

The conversion formulae for time domain to frequency domain and vice versa
are known as Fourier Transform pair.

∴ X( ) =

And x(t) =

For convenience, the correspondence between two domains will be denoted by a

double arrow. This the notation x(t) ↔ X( ) denotes that X ( ) is the direct Fourier
Transform of x(t) is the inverse FT of X( ) related by above equation.

1. Symmetry Property: -

If x(t) ↔ X( ) then x(t) ↔ 2πX(- )

Proof: - We have x(t) =

∴2πx(t) =

2πx(-t) =

Since is a dummy variable in the above integral it may be replaced by another

variable τ.
∴ 2πx(-t) =

Replacing t by .

2πX(- ) =

Replacing dummy variable τ by another variable t,

We get 2πX(- ) =

Hence X(t) ↔ 2πx(- )

This property holds good if x(t) is even function.

2. Linearity Property: -

If

then for any arbitrary constants and

3. Scaling Property: -

If x(t) ↔ X( ) then for a real constant a x(at) ↔

Proof: - For a positive real constant a

Let x = at, then for positive constant a.

∴F

Similarly if a < 0 i.e. for negative constant

∴

Significance of the Scaling Property: -

The function x(at) compressed in the time scale by factor a. Similarly a function
represents a function X( ) expanded in the frequency scale by the same factor „a‟. The
scaling property therefore states that compression in the time domain is equivalent to
expansion in the frequency domain and vice versa.

4. Time – Shifting Property: -

If x(t) ↔ X( ) then x ↔ X( )

Proof: F

Let then dt = dx

∴F

This theorem states that if a function is shifted in the time domain by the seconds, then
its magnitude spectrum X( ) remains unchanged out the phase spectrum is changed by
an amount -

We may states that a shift of in the time domain is equivalent to multiplication by

in the frequency domain.

5. Frequency Shifting Property: -

If x(t) ↔ X( ) then x(t)

Proof: F
=

=X

This theorem states that a shift of in the frequency domain is equivalent to

multiplication by in the time domain. It is evident that multiplication by factor
translates the whole frequency spectrum X( ) by an amount 0. Hence this
theorem is also known as “FREQUENCY TRANSLATION THEOREM”.
In communication systems it is often desirable to translate the frequency spectrum. This
is usually accomplished by multiplying a signal
x(t) by a sinusoidal signal. This process is known as modulation. Since a sinusoidal
signal of frequency 0 can be expressed as the sum of exponentials it is evident that
multiplication of a signal x(t) by a sinusoidal signal (modulation) will translate the
whole frequency spectrum.

∴ Fourier Transform of will be

Similarly

Thus the process of modulation translates the frequency spectrum by amount this
result is also known as “modulation theorem”.

6. Time Differentiation and Integration:-

If x(t) ↔ X( ) then ↔ X( ) and provided that is

bounded at = 0. This is equivalent to saying that X(0) = 0 or

Proof: We know that x(t) =

∴

Changing the order of differentiation and integration, we get

∴ It is evident from above equation that

In a similar way the result can be extended to the nth derivation.

Now consider the function ∅ (t) =

Hence if ∅ (t) ↔ ∅ ( ) then x(t) =

i.e. X( ) =

=> =

The above result are valid only if exists, i.e. if is absolutely integrable. This is
possible only if

i.e.

This is equivalent to the condition that X(0) = 0 since

Thus we conclude that differentiation in the time domain is equivalent to multiplication

by j in the frequency domain and the integration in the time domain is equivalent to
division by j in the frequency domain.
7. Frequency Differentiation and Integration: -

If x (t) ↔ X( ) then = -jtx(t) ↔ and

X( ) = F [x(t)] =

Changing the order of differentiation and integration

∴ It is evident from above equations

The extension of this result to higher derivation of X( ) yields

Similarly

∴ We conclude that derivation in frequency domain is equivalent to

multiplication by –jt in the time domain and integration in frequency domain is
equivalent to division by –jt in the time domain.

8. The Convolution Theorem:-

Given two functions form the integral

x(t) =

This integral define the convolution of functions The

convolution integral is also expressed symbolically as

x(t) =

Again as usual we have two theorems: Time convolution and Frequency

convolution.
9.Time Convolution Theorem: -

If and then

i.e.

Proof: - F

From the time shifting theorem we get

∴F

10.Frequency Convolution Theorem: -

If then

i.e.
 LAPLACE TRANSFORM: (LT)

2.2 INTRODUCTION:

► If we generalize the complex sinusoidal representation of a continuous-time signals

offered by the Fourier Transform to a representation in terms of complex exponential
signals, which is termed the Laplace Transform.

► The Laplace Transform provides a broader characterization of system and their

interaction with signals.

► Laplace Transform can be used to analyze a large class of continuous signals that are
not absolutely integrable, for these signals Fourier Series and Fourier Tranform are not
possible.

► The Laplace Transform classified as two types:

I. Uni-lateral Laplace Transform

II. Bi-Lateral Laplace Transform

►The prime rate of the Laplace Transform in engineering is transient and stability
analysis of casual LTI systems described by differential equations.

2.2.1 Concept of Laplace Transform:

Let eSt be a complex exponential with complex frequency

S = σ + jω

eSt = e(σ+jω)t = eσt ejωt

= eσt (cosωt+jsinωt)

= +j ----------------------> (1)
Real Imaginary
Consider an LTI system

Input x(t)=eSt

= h(t)*x(t)

Substitute x(t) = eSt

=

= eSt

Define the transfer function H(S) =

y(t) = H(S)eSt ---------------------------------→ This indicates the action of

the system on an input eSt
is multiplicated by H(S).

 Expressing the H(S) in polar form:

H(S) = |H(S) |ej∅(S) ↕ y (t) = |H(S) |ej∅(S)eSt

Substitute S = σ + jω

y (t) = |H (σ + jω) |eσt . e(jωt+∅(σ+ω))

= |H (σ + jω) |eσt cos (ωt +∅ (σ+jω)) + j |H (σ + jω) |eσt sin (ωt +∅ (σ+jω)) --------> (2)

Comparing the equation (1) and (2)

The system changes the amplitude of the input by |H (σ + jω) | and shifts the phase of
the sinusoidal components by ∅ (σ + jω). The system does not change the damping
factor σ of the input or sinusoidal frequency ω.

H (S) =

S = σ + jω and using t as variable of integration for arbitrary signal x (t).

H(S) = ------------------> (3) X(S) =

 Relation between Laplace Transform and Fourier Transform:

From eq (3) or from Laplace Trasform

H (σ + jω) =

= -----------------> (4)

From the definition of Fourier transform…

H (jω) =

H (σ + jω) = FT[e-σth(t)] ------------------> (5)

Laplace transform of h(t) is equal to the Fourier transform of e-σth(t)

=> σ=0 LT(h(t)) = FT(h(t))-------------- > (6)

Finding the Inverse Fourier Transform

h (t)e-σt = ejωt dω

We may recover h(t) by multiplying with eσt on both sides

h (t) = ejωt..eσtdω

= e(σ+jω)tdω

Substitute S = σ + jω => ds =

ω→∞
S = σ + j∞ h (t) = * =
ω → -∞
S = σ - j∞
h (t) → is inverse Laplace Transform
For arbitrary signal x (t)
=> X (S) = x (t) =

Definition for Definition for

Laplace Transform Inverse Laplace
Transform

2.2.2 Region of Convergence (ROC) for Laplace Transforms: -

Laplace transform of x (t) is the Fourier transform of x(t)e -σt. Hence the Laplace
Transform is guaranteed to converge if x(t)e-σt is absolutely integrable.
Means
|dt < ∞
Then |X(S)| < ∞
The range of real part of {S} = σ for which the Laplace Transform converges is termed
the region of convergence (ROC). That is, the ROC consists of those values of σ for
which the Fourier transform of x(t)e-σt converges. ROC provide information whether
x(t) is Fourier transformable or not.
S-plane: It is convenient to represent the complex frequency S graphically in terms of a
complex plane called the S-plane.

The horizontal axis represents the

real parts of S and the vertical axis
represents the imaginary part of S

Fig: 20

Poles and zeros: Laplace Transform x(s) is a ratio of two polynomials.

X(s) =
=> We can write x(s) in terms of factors.
X(s) = =K
=> When s=Zi the numerator N(S) = 0 and transfer function vanishes
=> Zi ‘s are the roots of the equation N (S) = 0
are defined as zeros.

=> S=Pi then D(S) = 0

Pi „s are roots of D(S)=0 are defined as poles

2.2.3 Properties of ROC

For a given X(s), there may be more than one inverse transform . Laplace Transforms
are identical and ROC‟s are different.
There is no one to one correspondence between X(s) and x(t).
1. The ROC of X(s) consists of strips parallel to the jω-axis(Imaginary) in the s-plane.
2. If x(S) is rational, then the ROC must not contain any poles.
3. If x(t) is of finite duration and is absolutely integrable (i.e )then the
ROC is the entire S-plane.
4. If x(t) is right-sided and of infinite duration (i.e. x(t)=0 for all t<T 1 for some finite T1).
Then the ROC is the region in the S-plane to the right of the right most pole.
5. If x(t) is left-sided and of infinite duration (i.e x(t)=0 for all t>T 2 for some finite T2),
then the ROC is the region in the S-plane to the left of the left most pole.
6. If x(t) is two-sided and of infinite duration ( the signal is of infinite extent for both t<0
and t>0), then the ROC will consist of a strip in the S-plane.
Examples:
EX-1 Determine the Laplace transform of the signal
x(t) = e-atu(t) a>0

X(s) =

1 t≥0
Since u(t)= => x(s) =
0 t<0

X(s) = e-(S+a)∞ + e-(S+a)0

In the first term t→∞ the term will become
Zero when s+a>0
s>-a
σ+jω>-a
σ>-a => Re (s)>-a. ----- Condition for convergence
If this condition satisfied the term –I becomes zero.
x(s) = ROC: Re(s)>-a

Fig: 21(a) Fig: 21(b)

EX: 2 Determine the Laplace Transform of signal

x(t) = -e-atu(t) a>0

X(s) = =

X(s) = -

X(s) = - X(s) = =
X(s) = +

To converge x(s) the II term becomes zero. For that (s+a)<0 => => 0

Re(s) < -a ----ROC condition.

Fig: 22

EX: 3 Determine the Laplace Transform of x(t) = e-2tu(t) – e-3tu(t)

Solution: LT[x(t)] = LT[e-2tu(t)] – LT[e-3tu(t)]

LT
e-σtu(t) ↔ Re(s)>-9

Similarly
LT
e-2tu(t) ↔ Re(s)>-2

LT
e-3tu(t)↔ Re(s)>-3

X(s) = - = ROC: Re(s)>-2

Pole-zero plot:

Fig: 23(a) Fig: 23(b)

As per ROC: property-4 (right of the right most pole)

EX: 4 .Determine the Laplace Transform of x(t) = - e-2tu(t) + e-3tu(t)

LT[x(t)] = LT[- e-2tu(t)] + LT[e-3tu(t)]

LT
-e-at u(-t) ↔ ROC: Re(S)<-a

LT
-e-2tu(-t) ↔ Re(s)<-2

LT
e-3tu(-t)↔ Re(s)<-3

X(s) = - = ROC:Re(s)<-3

As per ROC property-5

The ROC is the region in s-plane

to the left of the left most pole.
Pole at s=-3 left most pole.

Fig: 24

EX: 5 Determine the Laplace Transform of the two sided signal x(t) = e -a|t| a>0.

Solution:
eat t<0
x(t) = e-a|t|=
e-at t≥0

Fig: 25(a) Fig: 25(b)

x(t) = eatu(-t) + e-atu(t)

LT[x(t)] = LT[eatu(-t)] + LT[e-atu(t)]

e-atu(t) ↔ Re(s)>-a

eatu(-t) ↔ Re(s)<+a

X(s) = - =

e-a|t| ↔

ROC: -a<Re(s)<a

EX: 6 Determine the Laplace of the finite duration signal

eat 0<t<T
x(t) =
0 otherwise

X(s) = = =

= X(s) = [1- ]

The ROC: entire s-plane for a finite duration signal

At =>s =-a x(s) = 0/0

x(-a) = =T

EX: 7 Determine the Laplace transform of standard signals

(a.) x(t) = δ(t) → unit impulse function

(b.) x(t) = u(t) → unit step

(c.) x(t) = r(t) → unit ramp

Solution:

(a.) x(t) = δ(t)

X(s) = =
= = =1

δ(t)↔1 ROC: entire s-plane

(b.) x(t) = u(t)

X(s) = = =

X(s) =
s>0 =>term I becomes zero.

X(s) = Re(s)>0 LT
u(t) ↔ ; Re(s)>0

(c.) x(t) = r(t) = tu(t)

X(s) = = =

X(s) = = -

= + + = ; ROC: Re(s) > 0

EX: 8 Determine the Laplace transform of the following signals:

(a.) x(t) = sin(ω0t)u(t)

(b.) x(t) = cos(ω0t)u(t)

Solution: x(t) = sin(ω0t)u(t)

LT[x(t)] =

e-atu(t) ↔ Re(s)>-a

↔ Re(s)>0

LT
 ↔ Re(s)>0

X(s) = =

X(s) = =

Both the terms converges. R(s)>0.

(b.) x(t) = cos(ω0t)u(t)

x(t) =

e-atu(t)↔ ; Re(s)>-a

↔ Re(s)>0

↔ Re(s)>0

X(s) = =

2.2.3.INVERSE LAPLACE TRANSFORMS

ILT evaluating methods:

1- Knowledge of ROC can be used to determine a unique ILT. The most common form
of other knowledge is causality, stability or existence of the Fourier Transform.

Causal → right sided inverse transform of each side.

Stable → signal is absolutely integrable and thus has Fourier Transform.

Stability and existence of Fourier Transform are equivalent conditions.

→ The ROC includes the jω axis in the s-plane or Re(s) = 0

ILT → is obtained by comparing the pole locations to the jω axis. If pole is right of jω
axis the right sided ILT. If pole is lies on the left of jω axis, then the left – sided Inverse
transform is chosen.
Properties of the Laplace Transform:

The properties of the bilateral Laplace Transform are similar to those of the Fourier
transform. The derivations of many of these properties are analogous to those of the
corresponding properties of the Fourier transform.

I Linearity: If x1(t) ↔ X1(s) with ROC = R1

x2(t) LT
↔ X2(s) with ROC = R2
LT
ax1(t) + bx2(t) ↔ aX1(s) + bX2(s)

Proof: LT [ax1 (t) + bx2 (t)] =

=a

= aX1(s) + bX2(s)

II Time shifting: If x (t) ↔ X(s) with ROC = R

Then x (t-t0) ↔ X(s) with ROC = R.

Proof: LT [x (t – t0)] =

A change of variables is performed by letting τ = t – t0

dτ = dt t = τ + t0

τ → ∞ as t → -∞ and τ → ∞ as t → ∞

III Shifting in the S-Domain:

x (t) ↔ X(s) with ROC = R.

then x(t) ↔ X(s - with ROC = R + Re

The Laplace Transform of x (t) is given by

LT [x (t) ]= e-st dt

= dt

= X(s – s0).

The ROC associated with X(s-s0) is that of X(s), shifted by R . Note that if X(s) has a
pole or zero at s=a, then X(s-s0) has a pole or zero at (s-s0) = a i.e. s = (a + s0).

IV Time Scaling

If x(t) ↔ X(s) with ROC = R

Then x(at) ↔ X with ROC = aR.

Proof: Case I: For a positive real constant „a‟ Case II:

L{x (at)} = For a negative real constant „-a‟

Let τ = at => dτ = adt L{x (-at)} =

τ → -∞ as t = -∞ and Let τ = -at => dτ = -adt

τ → ∞ as t = ∞ =

=
= X
=

Combining the two cases we obtain

L{x (at)} = X

V Differentiation in the Time Domain:

x (t) ↔ X(s) with ROC = R.

then ↔ S X(s) with ROC containing R.

Proof: x (t) = L-1[X(s)] = [X(s)] .

Differentiating both sides we get

 = [S X (s)]

=L-1[S X (s)]

∴ ↔ S X(s)

↔ Sn X (s)

VI Differentiation in the S-domain:

x (t) ↔ X (s) with ROC = R

t x (t) ↔ with ROC = R.

Proof: X (s) =

Differentiating both sides with respect to S we get

L [-t x (t)] =

t x (t) ↔

x (t) ↔ (-1)n

VII Convolution property:

If x1(t) ↔ X1(s) with ROC = R1

x2(t) ↔ X2(s) with ROC = R2

then x1(t) * x2(t) ↔ X1 (s) X2(s) with ROC containing R1∩R2.

The convolution property states that convolution in the time domain corresponds to
multiplication in the S-Domain.

Proof: x1(t) * x2(t) = dτ.

L [x1(t) * x2(t)] = dt

=
Inter changing the order of Integration.

= .

Using the time shift Property the term in the bracket becomes X2(s) e-sτ

=x2(s) dτ = X1(s)X2(s)

VIII Conjugation property:

x (t) ↔ X (s)

x* (t) ↔ X* (s*)

L [x*(t)] = = = [X(s*)]*

= X*(s*)

IX Multiplication property:

x1 (t) ↔ X1 (s) with ROC = R1

x2 (t) ↔ X2 (s) with ROC = R2

x1(t)x2(t) ↔ [X1(s)*X2(s)] = 1 with ROC containing R1∩R2 the

Laplace transform maps the multiplication of two signals into the convolution of their
transformation.

Proof: L [x1(t) x2(t)] =

= x2(t)e-stdt

Interchanging the order of Integration and noting that X1(s1) does not depend on t1,then
we have

= ds1

Using the shifting in the S-domain property the bracketed term is X2(s-s1).

= X1(s1) X2(s-s1) ds1

= [X1(s)*X2(s)]

LT
x1 (t) x2(t) ↔ [X1(s) * X2(s)]

X Integration in the Time Domain:

x (t) ↔ X(s) with ROC =R.

then dτ ↔ X (s) with ROC containing R ∩ {Re (s) >0}.

Proof: If a signal x (t) is convolved with a unit step function u(t). We get

x (t) * u (t) = dτ

Since =1 t > τ = τ<t

=0 t<τ

x (t) * u (t) =

The convolution of a signal with a unit step function is the same as the cumulative integral
of the signal.

= x (t) * u (t)

Applying Laplace transforms on both sides

L =L

u (t) ↔

= X (s) u (s)

= X (s)

2.2.4.Laplace Transform of Periodic Signals

The Laplace transform of a casual signal can be determined from the knowledge of the
Laplace transform of its first cycle. Consider periodic signal x (t) with period T o.
 Fig:26

Let x1(t), x2(t), x3(t) be the signals representing the 1st, 2nd, 3rd cycles of a periodic signal
x (t),therefore x(t) can be written as,

x (t) = x1(t) + x2(t) + x3(t) + x4(t)_ _ _ _

= x1 (t) + x1(t-t0) + x1(t-2t0) + x1(t-3t0)

LT
If x1(t) ↔ X1(s)

X (s) = X1 (s) +X1 (s) + X1(s) +____

(Using time shifting property)

= X1(s)

=X1(s) = X1(s) )n

X (s) =

∴ The Laplace transform of a periodic signal x (t) is given by

X(s) = Where X1(s) is LT of

one cycle.

Analysis of the system using Laplace Transform:

Fig: 27
y (s) = X (s) H (s)

H (s) =

2.2.5.Impulse Response and Step Response:

The impulse response h (t) is defined as the output of an LTI system due to a unit
impulse signal input x (t) applied at time t=0.

h (t) ↔ H (s)

Where H (s) is the system function or transfer function. The step response s(t) is defined
as the output of an LTI system due to a unit step signal x(t) = u(t)

The step response of an LTI system with Impulse response h(t) is given by

s (t) = h (t) * x (t)

= h (t) *u(t)

=
u(t – τ) = 1 t>τ
s (t) = = 0 t<τ

L [s (t)] = L
= (Using Integration property.)

2.2.6.UNILATERAL LAPLACE TRANSFORMS

In many applications we are considering the causal signals. In causal system we are
interested in the behavior of the system t ≥ 0.

In causal system we do not need to consider the ROC. All are right sided sequences.

The differentiation property of ULT is to analyze the behavior of a causal system

described by differential equations with initial conditions.

X (s) =

The ULT and BLT are same for signals that are zero for t <0.
Differentiation in time domain:

x (t) ↔

Integrating by parts

x (t) ↔ x (t) +s

= x (t)

= sX (s) – x ( )

General form:

↔ x (t) -s x (t) _ _ _ _ _ _ sn-2 - sn-1x (0+)

= - sn-1x (0) - sn-2 |_ _ _ _ _ _ _ x (t)

= ______ x (t)

d3x (t) = s3X(s) – s2x (0) – sx1 (0) – x1 (0)

Initial and final value theorems:

The initial and final value theorem allows us to determine the initial value of x (t), x (0 +),
and the final value from X(s). These theorems are most often used to evaluate either the
initial or final values of a system output without explicitly determining the entire time
response of the system.

x (o+) =

x (∞) =

Initial value theorem:

Proof

LT = = s X(s) – x (0)

lim s → ∞ = lim s → ∞

lim s → ∞ = lim s → ∞ s X(s) – x (0)

x(0) = lim s → ∞ s X (s).

Final value Theorem:

lim s → 0 = s X(s) – x(0)

lim s → 0 = lim s → 0

= lim s → 0

x (∞) – x (0) = lim s → 0

x(∞) = lim s → 0