Signals and Systems (ESC501)

IEM Kolkata - CSE 3A/3B/3C/3AI

Module - III
Laplace Transform - Introduction

Dr. Ardhendu Kundu

 J2t

7
Signal and System Analysis
using the Laplace Transform

7.1 Introduction
In Chapter 6 we have studied that Fourier transform of a function x(t) can be rep
resented by continuous sum of exponential functions of the form ej, In addition
the Fourier transform technique is applied for solving differential equation that
convert
relates input and output of a system. First, the Fourier transform is used to
the differential equation into algebraic equation and then the Fourier transform
Hj2) and
of the output Y(G2) is obtained as a function of frequency response
time by
Fourier transform of inputX((2). The output is obtained as a function of
taking inverse Fourier transform of Y(G2). But this technique is rarely used be
[Link] the Fourier
cause Fourier transform does not exist for many
transform cannot be used to analyse unstable systems.
include an exponentia!
An altermate approach is to redefine the transform and
is represented as
convergence factor o along with j . That is, the signal x( complex frequency
where s is
asum of complex exponentials of the form e
denoted by s = + js.
x(1) can be represented as
The Fourier transform of a continuous time signal

X(G2) = x()e-dt (7.1)

wherc

x(t)= (7.2)
2T

Replacing j2 by o + j2 in Eq. (7.1) yields

(7.3a)
X(o+j2)=x(1)e (o+jD" dt
 7.2 Signals and Systems

and x(r) can be expressed as

1
x(1) = 21 |X(o+j2)elot/2yde. (7.3b)

The term o+jR is denoted by s. Then d2 =)ds.

Substituting these values in Eq. (7.3), we get

X(s) = x(1)e-dt (7.4a)

-OO

and
G+joo
1
x(t) = X(s)eds (7.4b)
27j.G-joo
Eq. (7.4a)and Eq. (7.4b) constitute complex Fourier transform pair or the bilateral
Laplace transform pair.
Torepresent the Laplace and inverse Laplace transform we use following no
tations

x(1) X(s)
L()]= X(s)
and
Lx)]=x(),
where X(s) is Laplace transform of x(1) and x(1) is inverse Laplace transform of
X(s).
/These equations are called Laplace transform pair;

e Convergence of the Laplace Transfrom

The bilateral Laplace transform of a signal x() exists if

X(s) = | x(t)e d1 (7.5)

is finite.
 Signal and System Analysis using the Laplace Transtorm 7.3

Substituting s=o+j2 in Eq. (7.5) yields

X(s) = | x(t)e-(ota)Ndt

(7.6)

The above equation indicates that Laplace transform is the Fourier transform of
x(t)e . Therefore, a necessary condition for existence of the Laplace transtrom
isabsolute integrability of x(r)e-g, That is, X(s) exists if

x(t)e or dt < oo. (7.7)

|The range of o for which the

Laplace transform converges is x(t) -ot
x(t) e
known as Region Of Convergence
(ROC)| Consider a signal x(1)= 2
e u(t)
eu()as shown in Fig. 7.1. The
(2- o)t
u(t) for o>2
Fourier transform for x(1) does not
in
exist because it is not absolutely
t
(a) (b)
Now
tegrable (Example 6.1 iv). ea,
multiply the signal x() with Fig. 7.1 (a) Exponential sequence eu(t)
Then the resultant signal x()e o = ex
el?-ou() is absolutely integrable for which Fourier transform does isnotabso
ist (b) The signal x(t)e-ot
which
Laplace
for o > 2. Therefore the lutely integrable for o > 2.
is Fourier
transform of x(), which exists for
transformn of x()e
concept is
Some values of o. This
illustrated in Fig. 7.1.
that the Laplace transform exist for signals for which
Thus we can conclude absolutely integrable, then the Fourier
x() is
Fourier transform does not exist. If Laplace transform using the relation
obtained from its
transform of x(t) can be
X(j2) =X(s)=0 (7.8)
 7.4 Signals and Systems

Review Questions

1. What are the limitations of Fourier transform?

2. What is the relation betweenFourier transform and Laplace transform?

3. Define Laplace transform. What is meant by ROC?

S-Plane

The Laplace variable s is a complex variable defined by s = +j2, where o is

the real part of s denoted as Re(s) and the variable S2 is the imaginary part of s
denoted as Im(s). The range of variation of s can be represented schematically
as a s-plane. The horizontal axis of this plane is known as real axis, since it is
defined by
S=+ j0.
The vertical axis is known as imaginary Im(s)
axis and is represented by
LHP RHP

S=0+j2.
left
s-Plane Re(s)
The origin is represented by the
complex zero s =0+ j0. The j2
axis (imaginary axis) divides the s
plene in two halves. The region to the Fig. 7.1 s-plane.
of the imaginary axis (for -co <o <0) is termed the left half plane
(LHP) and
the region to the right of imaginary axis (for 0<o<oo) is termed the right
half
plane (RHP). The whole complex plane is known as s-plane.
The most general form of the Laplace transform is a ratio of two
polynomials
bysh +by-1s-1 +...+bys+ho
X(s) = N(s) (7.9)
D(s) SN+ ay-1e+...+ajst ao

The numerator polynomial of X(s) is N(s) = bys +

with the degree M. The zeros of X(s) are he values ofsby-1sM+...+ bys +bo
for which the transform
is zero. They are found by setting the numerator
the resulting equation. That is
polynomial to zero and solving
 Signal and Sy sten Analysis using the Laplace Transtorm 7.5

Ns)= bys+bY +...+ bs t b0 =0 (7.10)

bo
bM bart
= by(s -*)=0. (7.11)
k=1

The (-A} are the roots of the numerator polynomial and are zeros of X(s). The
zeros are represented in the s-plane with the 'o' symbol.
The denominator polynomial of X(s) is

D(s) =s+ay-s+...+ajs t ao (7.12)

with degree N. Parameter N isalso called the order of the transform X(s).
The poles of the transform X(s) are the values of s for which the transform is
infinite. The poles are found by setting the denominator polynomial to zero and
solving the resulting equation. That is

D(s) =s+ay-1st...+ajs + ao =0
N

=Jl(s-P:) =0. (7.13)

k=1

The {pk are the roots of the denominator polynomial and are poles of X(s). The
poles are representcd in the s-plane with the 'x' symbol.

Review Questions

1. What is s-plane?
2. What are poles and zeros? How they are represented in s-plane?

Solved Problem 7.1 Find the Laplace transform and ROCof thefollowing sig
nals

(i) e-au(t)
(i) -eau(-)
7.6 Signals and Systems

Solution:
K) Given
x(1) =eu(t)

X(s) =x(t)e dt
u(t)= 1 for >0
eMu(t)edt =0 for t<0

Re(s + a)>0
J
e-(ats)"dt
Re(s) > - a
We know that s = +j2, therefore the above integral converges if
Re(s),o>-a. Now
- 1
X(s) =
1
sta
1
L<e u()]=Sta ROC: o>-a (7.14)

(ii) Given

x(t) =-e u-t)

X(s)=x(()edt
u(-t)=0 for t>0
|-e"u-)e dt =1 for t <0

=-etariMdr
Re(s +a) <0
=’ Re(s) <-a
 Signal and System Analysis using the Laplace Transtorm 7.7

Since.s = o+ jS2. the integral converges if Re(s). that is o<-u and

1
X(s) = ROC: o<-a (7.15)
Sta

From solved problem 7.1 we can note that the Laplace transforms of both signals
"u(t)is ¡> -u and for
are jdentical, but the ROCs are different. The ROC of eat
eMu-) is G< -a. The ROCs of both signals are shown in Fig. 7.3.

S-plane s-plane
ROC
ROC

0 -a

(a) (b)

Fig. 7.3 (a) ROC of e u(t) (b) ROC of -e u-t).

points in the s- plane for

The shaded region in Fig. 7.3a represents the set of shaded region
Similarly the
which the Laplace ransform of eu() converges.
The signal e au(t) is acausal signal
in Fig.7.3b indicates the ROC of -eu-t).
non-causal signal. Ingeneral, the ROCof a causal
and the signal -eul-) isa ROC of anoncausal signal
where a is some coDstant and the
signal is Re fs}>a,
is Re s}<a.

signal
7.2 Findthe Laplace transform of the
SsProbiem: May 2005, Nov 2005

find ROC.
x(1)=eu(t) +eu(t) and
locations?
What are the pole-zero
u(t)
Solution: Given x(t) =e-ult) +e
X(s) =x()edi
7.8 Signals and Systems

|eut)e"dt+e u(t)e dt
-O0

-2.5
Mdt+e s+2 dt -3 -2 0
0

Converges if Converges if
Re(s) > -3 Re(s) > -2
Fig.7.4 ROC of the signal
converges ifRe(s) >-2 x(t)=eu(t)+e"u(t)
-1 1
X(s) = -(s+2)" lo
1
ROC: Re(s) > -2
S+3s+2
(s+2) + (s+3)
(s+3)(s+2)
26+5
X(s) = s2+ 5s+6 ROC: o >-2
The poles can be obtained by equating the denominator polynomial to zero. That
iss+ 5s +6=0’(s+2)(5+3) =0. The poles are at s =-2 ands= -3 The
zero iss =-5/2. The poles and zeros are located in s-plane, as shown in Fig.7.4

Solved Problem 7.3 Find the Laplace transform of the signal

x(t) =eau(t) +eu-t) and find ROC. What are pole locations?
Solution:
X(s) =x(t)e dt

e "ult)e dt + eu(-t)e"dt

-fe -(atsMdt + | e+rdt

Converges if Converges if
Re(s) > -a Re(s) < -b

converges if - a < Re(s)< -b

 7.9
Signal and System Analysis using the Laplace Transform

-1
e
-(s + a)
1 -b
-a
s+b Sta
ROC: a< Re(s) <-b (7.17)
The pole are at s=-aand s= -b. Fig.7.5 ROC of the signal
u(t) + e u(-t)

Solved Problem 7.4 Find the Laplace transform of the signal

AUNov 2003
x() =et-be|

Solution:
Given *(1)=ebiel

X(s) = x()e"dt
’.00

j2
- =ele"dt

=ee stdt + ebt e -st dt

0 -b

-/edt +e t d1 0

Converges if Converges if Fig.7.6 ROC of the signal e

Re(s) < b Re(s) > -b

1 +
X(s) = -(s-b) -(s +
-1 ROC: --b< Re(s) <b (7.18)
S-b s+b

poles are at s = tb.

Inthis case the
 7.10 Signals and Systems

Properties of R0C
From solved problem (7.3) and (7.4) we can find that two different signals can
have identical expressions for their Laplace transform and the signalare distin
guishable only by ROC. That is the ROC of X(s) depends on the nature of x(1).
In this section we will study the properties of ROC.
/ Property 1 The ROC ofX(s) consists of strips parallelto the j2- axis in the
s-plane. (see Fig.7.5 and Fig.7.6)
Aroperty 2 The ROC does not contain any poles.
Property 3 Ifx(1) is of finite duration and is absolutely integrable, then the
ROCis the entire s-plane.
Property 4 Itx() is a right-sided signal, that is x()=0 for t < tÍ < then the
ROC isof the form Re(s) > amay. Where Cmax equals the maximum
real part of any of the poles of X(s).
Property 5 Ifx() is a left-sided, that is x(r) = 0for ! > > -«,Then the
ROCis of the form Re(s) < anin:
Where amin equals the minimum real part of any of the poles of X(s).
Property 6 Ifx() is a two-sided signal, than the ROC is of the from
a < Re(s) < a.

Review Question
1. What is ROC of Laplace transform? State its properties.
AUNoy 2008

Solved Problem 7.5 Show that for a right-handed signal satisfying x(t) =0 for
t<to,Xs) converges for some value of s, the ROC ofX(s) is Re(s) > amar where
amax equals the maximum real part of any of the poles of X(s).
Solution: Given a right-handed signal satisfying
x(t)=0 for t< to
Let us assume that X(s) converges for Re(s) = Oo. Thei the
satisfies magnitude of X(s)

X(.)|s x()e "| dt

=|k(0)|le o||e-Jlo dt
 7.11
SIgnal and System Analvsis using the Laplace Transform

dt
Since x() =0fort<to

K()|=()je -Oo dt < oo

Let Re(s)= a> Oo. Then

Ix{)le- dt =k()le- e(a-o) dt

to

<e-(a-o)Mox()le-o dt <oo

the ROC
From above equation we find that X(s) converges for Re(s) = a. Since
Cquals
of X(s) cannot include any poles ofX(s), then Re(s) > max, Where Cmax
the maximum real part of any of poles ofX(s).

Practice Problem 7.1

pole-zeroplot with the ROC
Find the Laplace transform X(s) and sketch the
for the following signals
Yx()=eu(1) +eult)
(ii) x() =eu(r) +eu(-)
(iii) x() =eut) +eu(-) 2s +7
X(s) = ROC: Re(s) >-3
(i)Ans: Poles at s=-3;-4 (s+3)(s+4)
-7
X(s)= ROC -4<
Ans: Poles at s= 3;-4 (s+4)(s-3)
(ii)
Re(s) < 3. not
A(S) does not exist since ROCdo
(iii) Ans: Poles at s = -3;4
overlap.

Ihe Unilateral Laplace Transform

74 is applied to the signals that are causal; for the
transform
The unilateral Laplacetime. <0.
Signals that is zero for
 7.12 Signals and Systems

defined by
The unilateral Laplace transform of a signal x()is
(7.19)
X(*)=x0)e s dt
()e ole-Mdt (7.20)

Because le|=1, the integral convergesif

x(1)e ldt < , (7.21)

Solved Problem 7.6 Find the unilateral Laplace transform of the following
signals

()x(1) = 8(t) (ii) x() = u(r)

i) X(s) =x{)e "dt

S(t)edt=1

’L[8()] =1 (7.22)

i) x(t) = u(t)

X(¢)= x()e "dt= |edt

-1
S S

1
|L{u(1) = (7.23)

Solved Problem 7.7 Find the unilateral Laplace transform of

(i) x(t) = cos lot (ii) x() = sinSlot
 7.13
Signal and System Analysis using the Laplace Transform
Solution:
(i) Given x() = cos Qot=e4e o

esol + e
X(s) = x(t)e dt = 2
edt

= 0.5 eole "dt + le-ioledt

==0.5 eb-ja"dt + estjmo) 'dtd

1 1
0.5 +
S-j0 s+jsoJ
2s
-

2+J
(7.24)
L[cos(2o)] =

(ii) Given x(t) = sin lot

1
=

Lx() =L2-ea}
2j 0

e-(s-j0) e-(stjo)

-(s-jo) -(s+ jS2o)

2js20
2js- jSZo s+js0

LÍsin Slo t]=

=2+
7.14 Signals and Systemns

Practice Problem 7.2

Find the unilateral Laplace transform of the signal

x(1) =sin'slo

Ans:
s²+ 4s2;

7,5 Properties of unilateral Laplace transform

7.5.1 Linearity
If x1() and x>() are two signals with Laplace transform Xi (s) and X>(s), then
Llax1 (1)+ bx2()] = aX,(s) + bX,(s) (7.25)

where a and bare constants.

Proof

Llas, () + bxt) =a() +brz(t)]e "dt

ax1()e dt +b x()edt
= aXi(s) + bX(s)

Solved Problem 7.8 Find the Laplace transform of the signal

x(1) = 2eu(1) + 4eu(t)
Solution: Given x(t)=2e-u(t) +4eu(t)
X(s) = Lx()] = L(2e u(t) + 4eul)l
Using linearity property
X(s) = 2L|e ut)] +4L<e "u()
2 4
s+2
ROC : Re(s) > -2
s+4
 Signal and System Analysis using the Laplace Transform 7.15

7.5.2, TraDsform of derivatives

If L(1)] = X(s)
d (7.26)
then
Lx()=sX(s) -x(0 )
Proof

From thc definitionof Laplace transform, we can write

d
(7.27)

This equation may be integrated by parts by letting

-st
and dv= dx(t)
’ du =-seSdt and V=x(t)

udv =vu- vdu

L=ste),-x0)-e*)dt 0

=e +sx(t)edt
0
0

=-x(0)+sX(s)
sX(*) -x(0 ) (7.28)

x()e =0.
Since X(s)exists, lim
the transform of the second derivative, let us write
Tofind
dxlt)
d12 dt dt
 7.16 Signals and Sysiems
dx
[d'x)] dt
(0)
d12
dx
=slsX(s) -x(0 ))-0)
dt
dx (7.29)
=s'X(s) -sx(0 ) (0)

In the above expression the quantity (0-) is the derivative of x(1) evaluated at
I=07.
Similarly
L dn
=sX(s)-=1 x(0)--2dx(0)... an-'x(0) (7.30)
dt dtn-1

7.5.3 Transform of the integrals

The Laplace transform of integral of x(1) is given by
0

X(s) (7.31)

Proof.

x(r)dr = x(t)dt +/x()dt 0

(7.32)

L x(t)dt =L x()dt + x()dr (7.33)

esdt

The integration may be carried out by parts udv = vå fvdu

where u= x(r)dt ’ du =x(t)

0

and dy =edt ’ = st

S -e
 7.17
Signal and System Analysis using the Laplace Transform
Hence

L -1 (7.34)
S
0 0

X(s)
the
The first term vanishes since e approaches zero as / tends to infinity and at
lower limit

x(r)dt = 0

X(s)
Then L
S

transform of
When the first term in Eq. (7.32) is a constant then the Laplace
0

0
[x(t)dt)
[x(t)dt can be written as
Therefore 0
[x(t)dr)
X(s) (7.35)
S S

7.5.4 Scaling property

If Lx()] = X($)
1
then Lz(ar)= ) (7.36)

Given x(1) =e u(), hnd inverse Laplace transform of

Solved Problem 7.9
e-3s X(2s) JNTUNov 2004
L<e-'ut)] =
Solution: We know we have
property
Using time scaling
1
7.18 Signals and Systems

If A(s)=
s +I
4
X(2s) = +1 - 2X(2s) = s+2
s+2

L2r(2*)}) =x (.a=)
2

L'X(2)] = )
Using time shifting property we have
Llx(t-to)] =e oX(s)
’Leslo X(s)] =x(t -to)
3
»1e"x2)
=)-e(u()
Practice Problem 7.3

FindLaplace transform of the signal x(t) = e-i/2 cos

(u)
7.5.5 Aime shift
If Lx()]= X()
then L[x(t-to)] =e oX(s) (7.37)
Atime shift of to corresponds to multiplication by complex exponential e-so,
Proof

L[x(t -to)]= |x(t- to)e"dt

0

Let t -to = p, then dt = dp and

Lx(t- t,)] = x(p)e lotpdp

-lo

=e
sto
|x(p)e Pdp
(0

-e sto X(s)
 7.19
Signal and System Analysis using the Laplace Transform

Solved Problem 7.10 Find the Laplace transform of x(1) =e u()- u

5)] and its ROC. JNTUNov 2004
Solution:
Given
x() =e u(t) - u(t -5)]
=eSu(t) -e "u(t -5)

=e -Su(t) -e-25-S-5)(t -5)

1
We know L<e u(1))= s t a ROC: Re(s) > -a
Leu(t)]= ,ROC: Re(s) > -5
Using time shifting property we have
Lk(t- to))=eoX(s)
1
Hence Lle s-ult - 5)] =eSs. Re(s) > -5
s+5
e25e25

’ Lk()] = st5 S+5

1-e-5s+5)
ROC; Re(s) > -5
S+5

Practice Problem 7.4

of the signal x(t) =eult-2).
Find Laplace transform

7.5.6 Frequency sh
If L()] = X()
(7.38)
then Le "x)] = X(s+ a)
Proof Le-"x())= e "al)e ádt
- lx()e(ots"dt =X(s +a)
)
7.20 Signals and Systems

Similarly (7.39)
L|e"x(1)] = X(s-a)
domain introduces a shift in
Multiplication by a complex exponential in time
complex frequency s in frequency domain.
=ea sin S2otu(t)
Solved Problem 7.11 Find Laplace transform of thesignal x() AUMay 2004

Solution: We knowthat the Laplace transform of

sin Slofu(1) =
Using frequency shifting property
L<e sin 2otu(1)] = (s+ a)2+ 2%
Solved Problem 7.12 Find the Laplace transform of
(4e cos 5t - 3e sin 5t]u(t) and its ROC. JNTU2005
Solution:

We have L|cos lotu(t)] =

s+S
and L|sin 20tu()]=
Using frequency shifting property
s+a
L[e cos lotu()] = (s+a)? +% ROC: Re(s) >-a
and Le sin Slgtu()] = (s+a)? + ROC: Re(s) >- a
we obtain L<4e- cos 5t-3e sin5 t]u()
4(s+2) 15
(s+2)2+25 (s+2)2 +25
Practice Problem 7.5
FindLaplace transform of the signal and its ROC. JNTUMay 2004
x() =e' sin(lot + p)u(t)
 7.21
Signal and System Analysis using the Laplace Transform
7.5.7 Diffefentiation in the s-domain

If LËx()] = X(s)
d (7.40)
then L[-tx()]=X(s)
Proof

X(s) =x(t)e-dt
Differentiate with respect to s on both sides

d
dX(s) =| x()(e
ds
dt

x()(-)e"dt

=-x()edt
= L[-tx()
’ L[-tx(1)] = dX(s)
ds

In the same way d"X(s)

LÊ(-)'x(1)]= (7.41)
dsn

Laplace transform of the signal x(1) = -te u(t)

Solved Problem 7.13 F1nd
AUNov 2006
Solution:

We have Lle "u(t)] = s a

’ L<e u()]= s+2

 7.22
Signals and Systemns
Using differentiation ins-domain property
dX(s)
L[-ix(1)] = ds
d1
’ L[-te u(r)] = dss+2
-1
(s+ 2)2

Practice Problenm 7.6

Find the Laplace transform of the signal x(t) =cos Slotu(1)

Solved Problem 7.14 Find Laplace transforin of x(t) =e'cos(S2o+ p)u(t)

and its ROC.
JNTUMay 2004
Solution:
Given x(:) =e'cos(lot + )u(t)
=elcos 2ot cosp -sin lot sin plu(t)
LËx()] = L[(e cos Slotcos - e sin lot sin¢)u())
=cos p L [ecoslotu(1)] sin pL<e sin S2gtu(t)]
cos p(s+ 1). lo sin o
(s+1)+6 (s+1)2 + % ROC: Re(s) >-1
7,5.8 Time convolution
If L()]=Xi(s)
and
LËa()] =X,(s)
then
Lk1(1) *x2)] =X()X, (s) (7.42)
Proof
The signals x1 (1) and x2() are causal.

Therefore xi() *x»()= |x(r)x2lt-r)dr

 7.23
Signal and System Analysis using the LaplaceTransform

Lx10)*x2())=|M()x(- t)u( -r)e dtd

0 0

u(t-t)=1for t>r
=0 fort <t

Note:

Here we used step function to change the limit from (0 to t) to(0 to oo).
Let tT=p, then dt = dp and

Lin() **2()) = /x() /(p)u(p)e- o+td

0

/n()a()-p)e -SPe sdp dt

=Xa(9) x(r)e-dt
= X(s)X, (s)
Thus the Laplace transforn of the convolution of twosignals is the product of
their respective Laplace transforms.
Solved Problem 7.15 Given x1() =eu() and x2(1) =e u(t). Determine
Y(s) where y(t) =x1(t-2)"*2(-t+3) AU Dec 2006
Solution: Given

x,() =e -ult) and x2(1) =eu(t)

X (s) = L[eu()) = S+2 and X2(s) = Lle " u(r)] =s+3
Let x(t) =xlt-2) and h(1) =n{-+3).
y() =x() *h(t)
 7.24 Signals and Systems

Using time convolution property Y(s) =X(s)H(s)

X(s) = L|i( -2))
Using time shifting property
Li(2) = eX(s)

H(s) = Lr2(-t +3)]

The signal z(-1+3) can be obtained by timereversal of x(1)and then shifting it
right by 3Units
Using time reversal property
1
Lí(-1)]= X(-s) = -s +3
Using time shifting property
e-3s
H(s) = Lr2(-t+3)] = eX(-s) =
3-s
-3s
Y(s) = X($)H(s) = ( 3-s

7.5.9 Frequency convolution

C+joo
1
Lz(t)x2(1))= 2rj Xi(u)Xy(s - u)du (7.43)
C-joo

7.5.10 Ipitial value theorem

The ipiíal value theorem allows us to
rectly from the transform X(s) without calculate initial value of x(t) i.e., x(0) di
the need for finding the inverse
of X (s). transform
We know

=sX(s) -x(0)
(7.44)