138 A TEXTBOOK OF ENGINEERING MATHEMATICS

2.6 COMPLEX FOURIER TRANSFORM

If a function F(x) defined on the interval (– ∞, ∞) is piecewise continuous in each finite partial
interval and absolutely integrable in (– ∞, ∞), then

f(p) = F{F(x)} = z
−∞
∞
F(x) . eipx dx
is defined as the Fourier transform of F(x) and is denoted by f(p). The function F(x) is called
the inverse Fourier transform of f(p).
The inverse formula for Fourier transform is given by

F–1 {f(p)} = F(x) =

Note. Some authors write the formulae as
1
2π z ∞

−∞
f(p) . e–ipx dp.

f(p) =
1
2π z−∞
∞
eipx . F(x) dx and F(x) =
1
2π z
−∞
∞
e–ipx . f(p) dp.

2.7 FOURIER SINE TRANSFORM

The infinite Fourier sine transform of the function F(x), 0 < x < ∞ is denoted by Fs {F(x)} or fs(p)
and defined by

fs(p) = z
0
∞
F(x) . sin px dx
and the function F(x) is called the inverse Fourier sine transform of Fs {F(x)}.
The inverse formula is given by

F(x) = Fs–1 {fs (p)} =

2
π z ∞

0
fs (p) . sin px dp.

2.8 FOURIER COSINE TRANSFORM

The infinite Fourier cosine transform of the function F(x). 0 < x < ∞ is denoted by Fc{F(x)} or
fc(p) and defined by

fc(p) = Fc{F(x)} = z ∞

0
and the function F(x) is called inverse Fourier cosine transform of fc(p).
F(x) . cos px dx

The inverse formula for infinite Fourier cosine transform is given by

F(x) = Fc–1 [fc(p)] =

2
π z
0
∞
fc (p) . cos px dp.

2.9 SOME IMPORTANT RESULTS

1. z0
∞ sin mx
x
π
dx = , (m > 0)
2
2. z
0
∞ 2
e − x dx =
2
π

3. z
e +e∞ ax

0 e πx + e − πx
− ax
1 a
dx = sec , (– π < a < π)
2 2
4. z
0
∞ e ax − e − ax
e πx − e − πx
1 a
dx = tan , (– π < a < π)
2 2

5.
∞
z sin rx
−∞ ( x − b) 2 + a 2
π
dx = e − ar sin br, (r > 0).
a
INTEGRAL TRANSFORMS 139

2.10 PROPERTIES OF FOURIER TRANSFORMS

2.10.1. Linear Property

If f1(p) and f2(p) are the Fourier transforms of F1(x) and F2(x) respectively then

F [c1F1(x) + c2F2(x)] = c1f1(p) + c2f2(p)

where c1 and c2 are constants.

Proof. We have,

F [c1F1(x) + c2F2(x)] = z ∞

−∞
[ c1F1 ( x) + c2 F2 ( x)] . e ipx dx

= c1 z−∞
∞
F1 ( x) . e ipx dx + c2 z
−∞
= c1 F [F1(x)] + c2F[F2(x)] = c1 f1(p) + c2 f2(p).
∞
F2 ( x) eipx dx

2.10.2. Change of Scale Property (Similarity Theorem)

If f(p) is the complex Fourier transform of F(x) then

1 FG pIJ
F [F(ax)] =
a
f
H aK , a ≠ 0.

Proof. F[F(ax)] = z ∞

−∞
e ipx . F(ax) dx
Put ax = t
⇒ dx =
dt
a

z
F pI
=
−∞
ez ∞
F(t)
dt = 1 ∞ e i GH a JK t F(t) dt = 1 f p .
a
ip

a −∞
t
a
a a
From this property, it is evident that if the width of a function is decreased while its
FG IJ
H K
height is kept constant then its Fourier transform becomes wider and shorter. If its width is
increased, its transform becomes narrower and taller.
Remark. If fs(p) and fc(p) are Fourier sine and cosine transforms of F(x) respectively then
1 FG p IJ 1
fc
FG p IJ .
Fs [F(ax)] =
a
fs
H aK and Fc [F(ax)] =
a H aK
2.10.3. Shifting Property
If f (p) is the complex Fourier transform of F(x) then

F [F(x – a)] = eiap f(p)

Proof. F [F(x – a)] = z−∞

∞
F( x − a) . e ipx dx
Put x − a = u
∴ dx = du

= z∞

−∞
F(u) e ip(u + a) du = eiap
z ∞

−∞
F(u) e ipu du = eiap f(p).

2.10.4. Modulation Theorem

If f(p) is the complex Fourier transform of F(x) then

1
F [F(x) cos ax] = [f(p + a) + f(p – a)]
2
140 A TEXTBOOK OF ENGINEERING MATHEMATICS

Proof. F [F(x) cos ax] = z −∞

∞
e ipx . F( x) cos ax dx

= z ∞

−∞
e ipx . F( x) .
Fe
GH
iax
+ e − iax
2
dx
I
JK
1 ∞ i ( p + a) x
2 −∞
1
=e
LM
z
. F( x) dx +
N
∞

−∞
e i( p− a) x . F( x) dx z OP
Q
= [f(p + a) + f(p – a)].
2
This result has application in radio and television where the harmonic carrier wave is
modulated by an envelope.
Note. If fs(p) and fc(p) are Fourier sine and cosine transforms of F(x) respectively then
1
(i) Fs [F(x) cos ax] = [f (p + a) + fs(p – a)]
2 s
1
(ii) Fs [F(x) sin ax] = [f (p – a) – fc(p + a)]
2 c
1
(iii) Fc [F(x) sin ax] = [f (p + a) – fs(p – a)].
2 s
2.10.5. Convolution Theorem
The convolution of two functions F(x) and G(x) over the interval (– ∞, ∞) is defined as

F∗G= z
−∞
∞
F(u) . G(x – u) du = H(x).
Statement. The Fourier transform of the convolution of F(x) and G(x) is the product of their
Fourier transform i.e.,

F{F(x) ∗ G(x)} = F{F(x)} . F{G(x)}

Proof. We have,

F{F(x) ∗ G(x)} = F
RS
T z ∞

−∞
F(u) . G ( x − u) du
UV
W
= z z ∞

−∞
RS
T −∞
∞
F(u) . G ( x − u) du e ipx dx
UV
W
= z RSTz
∞

−∞
F(u)
∞

−∞
G ( x − u) . e ipx dx du
UV
W
on changing the order
of integration

= z RSTz
∞

−∞
F(u) .
∞

−∞
UV
e ip( x − u) . G ( x − u) d( x − u) e ipu du
W
= z RSTz
∞

−∞
e ipu F(u)
∞

−∞
e ipt G (t) dt du
UV
W | where x – u = t (say)

= z ∞

−∞
e ipu F(u) du . F{G(t)}

= z ∞

−∞
e ipx . F( x) dx . F{G( x)} = F{F(x)} . F{G(x)}.
INTEGRAL TRANSFORMS 141

2.10.6. If Fs [F(x)] = fs(p) and Fc [F(x)] = fc(p), then

d d
(i) Fs [x F(x)] = – [f (p)] (ii) Fc [x F(x)] = [f (p)].
dp c dp s

Proof. (i) fc(p) = z 0

∞
F( x) cos px dx
d
{f (p)} = –
dp c z 0
∞
F( x) . x sin px dx = – Fs [x F(x)]

d
⇒ Fs [x F(x)] = – [f (p)]
dp c

(ii) fs(p) = z0
∞
F( x) . sin px dx
d
{f (p)} =
dp s z 0
∞
F( x) . x cos px dx = Fc [x F(x)]

d
⇒ Fc [x F(x)] = [f (p)]
dp s

EXAMPLES

Example 1. Find the Fourier transform of following functions:

e iωx , for a < x < b U| 1 U|
(i) F(x) = V| (ii) F(x) = , |x|≤ ε
V|
0, otherwise W 2ε
0 , x>ε W
(iii) F(x) = e–|x| (iv) F(x) = e–a|x|, a > 0

(v) F(t) =
t, for|t|< a
.
UV
0, for|t|> a W
Sol. (i) f(p) = z−∞
∞
F( x) . e ipx dx = z b

a
e iωx . e ipx dx

L e OP
= M
i(ω + p) x b

=
e i(ω + p)b − e i(ω + p) a
.
N i(ω + p) Q a
i(ω + p)

(ii) f(p) = z ∞

−∞
F( x) . e ipx dx

= z−ε
ε 1 ipx
2ε
e dx =
1
2∈ z −ε
ε
(cos px + i sin px) dx

=
1
2ε
.2 z
0
ε
cos px dx =
1 sin px
ε p
FG
H
IJ
K
ε

0
=
sin pε
pε
.

(iii) f(p) = z ∞

−∞
F( x) e ipx dx = z−∞
∞
e–|x| eipx dx

= z0

−∞
e x e ipx dx + z 0
∞
e–x eipx dx
142 A TEXTBOOK OF ENGINEERING MATHEMATICS

L e OP
=M
(1+ ip) x 0 Le
+M
OP
− (1− ip) x ∞

=
1
+
1
=
2
.
N 1 + ip Q −∞ N − (1 − ip) Q 0
1 + ip 1 − ip 1 + p2
(iv) We have just proved that
2
F{e–|x|} =
1 + p2
using change of scale property, we get

F{e–a|x|} =
1 LM 2 OP = 2a
.
a MM 1 + FG pIJ 2
PP p2 + a 2
N H aK Q
(v) f(p) = z −∞
∞
F(t) . eipt dt = z a

−a
t eipt dt

= zL a

−a
t (cos pt + i sin pt) dt = 2i z 0
a
t sin pt dt

R F cos pt IJ UV − 1 . FG − cos pt IJ dtOP

z
a

= 2i MSt . G −
a

MNT H p K W H p K PQ 0
0

L a 1 F sin pt I O
a
= 2i M− cos ap + G JP
MN p pH p K P
Q 0

L a
= 2i M− cos ap +
1 O
sin apP =
2i
(sin ap – ap cos ap).
N p p Q p 2 2

Example 2. Find the Fourier transform of the function

sin ax
(i) F(x) = xe–a|x|, a > 0 (ii) F(x) = ,a>0
x
R|1 + x , for − a < x < 0
U|
| a
(iii) F(x) = S1 − x ,
|V .
|| a for 0 < x < a
||
T 0, otherwise W
Sol. (i) f(p) = z ∞
x e–a| x | eipx dx

z z
−∞

= z 0

−∞
x eax eipx dx +
0
∞
x e–ax eipx dx = z
−∞
0
x e(a+ip)x dx +
∞

0
x e–(a–ip)x dx

R x e UV − 1 . e dx + RSx . e
z UV − 1 . e
z
( a + ip) x 0 ( a + ip) x − ( a − ip) x ∞ − ( a − ip) x
0 ∞
=S dx
|T a + ip |W a + ip
−∞
|T − (a − ip) |W
−∞ − (a − ip) 0
0

=– M
Le OP + LM e
( a + ip) x 0 OP = − 1 + 1 = 4 iap − ( a − ip) x ∞

N (a + ip) Q N − (a − ip) Q (a + ip) (a − ip) (a + p )

2 2 2 2 2 2 2

z
−∞ 0

(ii) f(p) =
−∞
sin ax ipx
∞

x
e dx =
∞ sin ax

x
(cos px + i sin px) dx z
z RST
−∞

=2
0 z
∞ sin ax

x
cos px dx =
0
∞ sin ( a + p) x

x
+
sin (a – p) x
x
dx
UV
W ...(1)
INTEGRAL TRANSFORMS 143

Case I. If |p| < a then a + p and a – p both are positive and then (1) gives
π π
+ = π. f(p) =
2 2
Case II. If |p| > a then for positive values of p, (a + p) is positive whereas (a – p) is
negative and for negative values of p, (a + p) is negative while (a – p) is positive. Consequently,
we get
π π π π
f(p) = − or – + = 0
2 2 2 2

Hence, f(p) =
RSπ, | p|< a UV
T0, | p|> a W
(iii) f(p) = z ∞

−∞
F(x) . eipx dx = z−a
0 FG 1 + x IJ e
H aK ipx dx + z FGH
0
a
1−
IJ
x ipx
aKe dx

= z a

−a
eipx dx +
1
a z−a
0
x eipx dx –
1
a z0
a
x eipx dx ...(1)

1 LMF e I OP LMF e I OP
F e I
z z
ipx a ipx 0 a
0 e ipx ipx
e ipx
dx – 1
a
=G
H ip JK a MNGH x . ip JK
+
−a −a
−
−a
1.
ip a PQ MNGH x . ip JK 0
−
0
1.
ip
dx
PQ
F I 1 LM e F I OP − 1 LMa e F I OP
ipa − ipa − ipa 0 a
e −e 1 e ipx ipa
1 e ipx
=G
H ip JK a MNa ip+ − GH JK P a M ip
ip ip
Q N −a
−
ip
.
ip GH JK P
Q 0

1L 1 1 O
= M − 1)P
− ipa ipa
(1 − e )− (e
a Np 2
p 2
Q
2 1 2
= 2
− 2 [e–ipa + eipa] = (1 – cos ap) ; p ≠ 0
ap ap ap2
when p = 0,

f(p) = z−a
a
1 . dx +
1
a z 0

−a
x dx −
1
a z a

0
x dx | From (1)

1 x2 F I 0
1 x2 F I a
a a
= 2a +
a 2 GH JK −a
−
a 2 GH JK 0
= 2a – − = 2a – a = a
2 2
R| 2 (1 − cos ap) ; p≠0
U|
Hence, f(p) = S ap 2
V| .
|Ta; p = 0W
Example 3. Find the complex Fourier transform of dirac delta function δ(x – a).

Sol. F{δ(x – a)} = z−∞

∞
δ(x – a) . eipx dx

z F I a+ h
a+ h 1 ipx 1 e ipx
= Lim
h→0 a h
e dx = Lim
h→0 h ip GH JK a
F e − 1I = e
iph F e − 1I = 1
iph
= Lim e ipa
h→0
GH iph JK ipa ∵ Lt
iph → 0
GH iph JK
Remark. For the function δ(t), F[δ(t)] = 1.
144 A TEXTBOOK OF ENGINEERING MATHEMATICS

2
Example 4. Find the Fourier transform of e − x . Hence find the Fourier transform of
2
− ax 2
(i) F(x) = e , (a > 0) (ii) F(x) = e − x /2

2 2
(iii) F(x) = e −4(x − 3) (iv) F(x) = e − x cos 2x.

Sol. f(p) = z−∞

∞ 2
e − x . e ipx dx =
R|FG x − ip IJ U|
z ∞

−∞
e−( x
2
− ipx)
dx

z
2
p2
∞ − S|H 2 K + V|
= e T 4 W dx
−∞

z
FG ip IJ 2

Put x −
ip
=z
=e
2
− ( p / 4)
∞
e
H
− x−
2 K dx 2
−∞ ⇒ dx = dz
= e −( p
2
/ 4)
z ∞
e − z dz
2

z
−∞
∞
− x2 − ( p2 / 4 ) 2 2
⇒ F( e ) = 2e e − z dz = π e − ( p / 4)
...(1)
0
(i) By change of scale property,
−
1 FG p IJ 2

F(e − ax ) =
2 1
πe
4 H aK =
π − ( p 2 / 4 a)
e
a a
1
(ii) Comparing with the result of deduction (i), we get a =
2
2 2
Hence, F(e − x / 2 ) = 2π e − ( p / 2)

(iii) We have, from (1),

2 2
F (e − x ) = π e − ( p / 4)

1 F pI 2

2 π − 4 GH 2 JK
2 π −( p2 / 16)
∴ F ( e −4 x ) = F { e –( 2 x ) } = e = e | By change of scale property
2 2
2 π 3ip –( p2 / 16) π [3ip− ( p2 / 16)]
Hence, F { e −4 ( x − 3) } = e e = e | By shifting property
2 2
(iv) We have, from (1),
2 2
F(e − x ) = π e − ( p / 4)

2 π 1
− ( p + 2) 2
1
− ( p − 2) 2 LM OP
∴ F( e − x cos 2x) =
2 e 4 +e 4 .
N | By Modulation theorem
Q
Example 5. Find the inverse Fourier transform of f(p) = e–|p|y.

Sol. F(x) =
1 ∞ −| p|y − ipx
2π −∞
e .e dp z
=
1
2π −∞
0
e py e − ipx dp +
LM
N
L
0
∞
z
e − py e − ipx dp

O
OP
Q z
1 MR e UV + RS e UV P
p( y − ix ) 0 − ( y + ix) p ∞

= S
2π M|T y − ix |W |T − ( y + ix) |W PQ
N −∞ 0

1 L 1 1 O y
= M +
2π N y − ix y + ix QP =
π( y + x )
. 2 2
INTEGRAL TRANSFORMS 145

Example 6. Find the Fourier transform of

F(x) =
1, |x|< a
.
UV
0, |x|> a W
Hence evaluate

(i) z
−∞
∞ sin ap cos px
p
dp (ii) z0
∞ sin p
p
dp .

Sol. f(p) = z ∞

−∞
2 sin ap
F( x) . e ipx dx = z a

−a
1 . e ipx dx = 2 z0
a
cos px dx

= ,p≠0
p
For p = 0, we find f(p) = 2a
Taking Inverse Fourier transform of f(p), we get

F(x) =
1
2π z ∞

−∞
f ( p) . e − ipx dp =
1
2π z ∞

−∞
2 sin ap − ipx
p
e dp

=
1
π z ∞

−∞
sin ap (cos px − i sin px)
p
dp

⇒
RS1,
T0,
| x| < a
| x| > a
=
1
π
UV
W z ∞

−∞
sin ap cos px
p
dp | Second integral vanishes

⇒
−∞ z
sin ap cos px
p
∞
dp =

Hence the first result.

π, | x|< a
0, | x| > a
RS
T
UV
W ...(i)

Again, from (i), 2 z0

∞ sin ap cos px
p
dp =
RS
π, | x|< a
T
0, | x|> a
UV
W
⇒
Putting x = 0 and a = 1, we get
z0
∞ sin ap cos px
p
Rπ/2,
dp = S
T 0,
| x| < a
| x| > a
UV
W
z0
∞ sin p
p
π
dp = .
2
...(ii)

Example 7. Find the Fourier transform of F(x) =

RS1 − x , 2
if |x|< 1
T 0, if |x|> 1

and use it to evaluate

x3 z FGH
0
∞ x x cos x − sin x
cos dx .
2
Sol. Fourier transform of F(x) is given by
IJ
K
f(p) = z∞

–∞
eipx . F(x) dx = z
−1
1
eipx (1 – x2) dx

=
z1

−1
cos px (1 – x2) dx + i
z 1

−1
sin px (1 – x2) dx

=2 z 1

0
cos px (1 − x 2 ) dx | Using property of Definite Integral
146 A TEXTBOOK OF ENGINEERING MATHEMATICS

LMR O
=2
N
2

Q
1

0
z
SMT(1 − x ) sinppx UVW − (− 2 x) . sinppx dxPP = 4p x sin px dx0
1
z 0
1

4 LR F − cos px I U F − cos px IJ dxOP = 4 LM − cos p + 1 FG sin px IJ OP

z
1 1

= M S xG
1

p MT H
J
KWV − 1. G
H p K PQ p MN p p H p K PQ
N p 0
0
0

4
= (sin p − p cos p)
p3
By inversion formula for Fourier transform

F( x) =
1
2π z ∞

−∞
4
p3
(sin p − p cos p) . e − ipx dp

=
2
π z∞

−∞
(cos px − i sin px) .
F sin p − p cos pI dp
GH p
JK
3

=
4
π z FGH
∞

0
sin p − p cos p
p3
I
JK
. cos px dp | Using prop. of definite integral

Put x =
1
2
, F
FG 1IJ = 4
H 2K π z
0
∞ (sin p − p cos p)
p 3
p
. cos dp
2

⇒
3
4 GH
=−
4
π p z
F p cos p − sin p I . cos p dp
JK 20
∞
3 ...(2)

where F 1I F 1I
F G J = 1− G J = 1− =
1 3
2

H 2K H 2K 4 4

From eqn. (2), z

( p cos p − sin p)
∞

0 p 3
p
. cos dp = −
2
3π
16
Replacing p by x, we get

0 z
( x cos x − sin x)
x 3
∞ x
cos dx = −
2
3π
16
Example 8. Find the Fourier transform of the function
.

F(t)
shown in the adjoining figure.
2

R|
2 , for − 1 < t < 1 U|
Sol. Here, S|
F(t) = 1, for − 2 < t < − 1 V| 1

T
1, for 1 < t < 2 W
–2 –1 0 1 2 t
Fourier transform is given by,

f(p) = z
−2
−1
1 . e ipt dt + z−1
1
2 . e ipt dt + z
1
2
1 . e ipt dt

Fe I +2Fe I +Fe I
ipt −1 ipt 1 ipt 2

=GH ip JK GH ip JK GH ip JK
−2 −1 1

=G
Fe − e I +2Fe − e I +Fe
− ip −2 ip ip − ip 2 ip
− e ip I
H ip JK GH ip JK GH ip JK
INTEGRAL TRANSFORMS 147

Fe ip
− e − ip I F
e 2ip − e −2ip I
= GH ip
+ JK GHip JK
1 1 2
= (2 sin p) + 2 sin 2 p = sin p (1 + 2 cos p).
p p p

Example 9. Find the Fourier sine transform of e–| x |. Hence evaluate z

0
∞ x sin mx
1 + x2
dx.

Sol. In the interval (0, ∞), x is positive so that e–| x | = e–x.

Fourier sine transform of f(x) = e–x is given by

Fs{f(x)} = z ∞

0
f(x) sin px dx = z0
∞
e–x sin px dx

Le −x O ∞
p
=M (− sin px − p cos px)P =
N1 + p 2
Q 0
1 + p2
Using inversion formula for Fourier sine transform, we get

f(x) =
2 ∞
z
F {f(x)} sin px dp or e–x =
π 0 s
2
π z
0
∞ p
1 + p2
sin px dp

Replacing x by m, we have
1+ p
e–m =
2
π z
0
∞ p
2
sin mp dp =
2
π z
0
∞ x sin mx
1 + x2
dx

Hence, z ∞

0
x sin mx
1 + x2
dx =
π
2
e–m.

e − ax
Example 10. Find Fourier sine transform of , a > 0. Hence find Fourier sine
x
1
transform of .
x

Sol. fs (p) = z 0
∞
F(x) sin px dx = z
0
∞ e −ax
x
. sin px dx = I (say) ...(1)

∴
dI
dp
=
d
dp
F
GH z
0
∞ e − ax
x
. sin px dx
I
JK | Differentiating (1) w.r.t. p

= z0
∞ e − ax ∂
x ∂p
(sin px) dx = z 0
∞ e − ax
x
. x cos px dx = z ∞

0
e–ax cos px dx

LM e (− a cos px + p sin px)OP

=
− ax ∞
=
a
Na + p 2
Q 2
0
a 2 + p2
Integration w.r.t. p yields,
F pI
I = tan GH JK + c –1 ...(2)
a
Initially when p = 0, I = 0 ∴ c=0 | From (2)
FG pIJ
∴ From (2), I = tan–1 H aK
∴ z0
∞ e − ax
x
. sin px dx = tan–1
FG pIJ
H aK
148 A TEXTBOOK OF ENGINEERING MATHEMATICS

Take limit as a → 0
1
x z ∞
sin px dx =
0
π
2
Example 11. Find Fourier cosine transform of the following functions:
R| x, 0<x<
1

| 1
2
(i) F(x) = S 1 − x, < x<1
|| 0, 2
x>1
|T
(ii) F(x) = S
R|cos x, 0<x<a
(iii) F(x) =
e ax + e − ax
or
cosh ax
,–π<a<π
|T 0, x>a e πx + e − πx cosh πx

x2
(iv) F(x) = e–2x + 4e–3x (v) F(x) = sin .
2
Sol. (i) fc(p) = z
0
∞
F(x) cos px dx = z
0
1/2
x cos px dx + z1/2
1
(1 – x) cos px dx

F x sin px IJ
z RS UV
z
1/2 1
1/2 sin px sin px 1 sin px
=G – dx + (1 − x) – (– 1) dx
H p K 0
0 p p T W 1/2
1/2 p

1 p cos px F
I – 1 sin p – 1 FG cos pxIJ
1/2 1

=
2p
sin
2
+
p2
GH
JK 2 p 2 p H K 0
2
1/2

1 FG cos p − 1IJ − 1 FG cos p − cos pIJ

=
p2 H 2 K p H 22K
1 FG 2 cos p − 1 − cos pIJ
=
p2 H 2 K
(ii) fc(p) = z ∞
F(x) cos px dx

z z
0
a 1 a
= cos x cos px dx = [cos (1 + p) x + cos (1 – p)x] dx
0 2 0

=
LM
1 sin (1 + p) x sin (1 − p) x
+
OP a

=
LM
1 sin (1 + p) a sin (1 − p) a
+
OP
2 1+ p N 1− p Q 0
2 N
1+ p 1− p Q
I FG e I
(iii) fc(p) = z
0
∞ Fe
GH e
ax

πx
+ e − ax
+e − πx
I cos px dx = F e
JK GH e z ∞

0
ax

πx
+ e − ax
+e − πx JK H
ipx
+ e − ipx
2 JK dx
=
1
2
LM e
N z
e +e
0
+e
∞ ( a + ip) x
dx + πx
e − ( a + ip) x
+e
− πx
e +e z ∞

0
( a − ip) x

πx
− ( a − ip) x

− πx
dxP
O
Q
L OP
1 L1 F a + ip I 1 F a − ip I O 1M
H 2 JK + 2 sec GH 2 JK PQ = 4 MM cos FG a + ipIJ + cos FG a − ipIJ PP
1 1
sec G
2 MN 2
=
MN H 2 K H 2 K PQ
INTEGRAL TRANSFORMS 149

a ip a p
2 cos
cos cos cosh
1 2 2 2 2
= . = ∵ cos ip = cosh p
4 FG
a + ip a − ipIJ FG IJ cos a + cosh p
cos
2 H cos
2 K H K
(iv) fc(p) = z0
∞
(e–2x + 4e–3x) cos px dx

= z0
∞
e–2x cos px dx + 4 z ∞

0
e–3x cos px dx = 2
p +4
2
+ 2
p +9
12

R|sin F x + pxI + sin F x I U|V dx

(v) fc (p) = z 0
∞
sin
x2
2
cos px dx =
1
2 z 0
∞
S| GH 2 JK GH 2
T
2 2
− px JK W|
=
1
2 z 0
∞
sin
Fx
GH 2
2
+ px dx –
I
JK
1
2 z –∞

0
F x + pxI dx = 1
sin G
H 2 JK 2
2
z ∞

−∞
sin G
F x + pxI dx
H 2 JK
2

Fx I
z z
2

1 ∞ i GH 2 + px JK 1 ∞ i ( x 2 + 2 px )
= Im. part of e dx = Im. part of e2 dx
2 −∞ 2 −∞

= Im. part of
1
2 z ∞

−∞
e
i
2 {( x + p) 2 − p2 }
dx = Im. part of
1 − ip2 /2
2
e z −∞
∞
e
−
( x + p) 2
2i dx

1 −(ip2 / 2) π
= Im. part of e
2 1
2i

π F cos p − i sin p I FG cos π + i sin π IJ

2 2
= Im. part of
2 GH 2 2 JK H 4 4K

π F cos p 2
p I 1 π F 2
p p I 2 2
=
2 GH 2 − sin J
2 K
.
2
=
2 G cos
H 2 − sin
2 JK
.

1
Example 12. Find Fourier cosine transform of and hence find Fourier sine
1 + x2
x
transform of .
1 + x2

Sol. fc(p) = z
0
∞
1
1 + x2
cos px dx = I (say) ...(1)

∴
dI
dp
= z
∞

0
− x sin px
1 + x2
dx = – z 0
∞ (1 + x 2 − 1) sin px
x(1 + x 2 )
dx

=– z
0
∞ sin px
x
dx + z0
∞ sin px
x(1 + x 2 )
dx

dI
dp
=–
π
2
+ z ∞

0
sin px
x(1 + x 2 )
dx ...(2)
150 A TEXTBOOK OF ENGINEERING MATHEMATICS

Again,
d 2I
dp2
= z 0
∞ x cos px
x (1 + x ) 2
dx = z0
∞ cos px
1 + x2
dx = I | From (1)

d 2I
⇒ –I=0 ...(3)
dp2
Solution of (3) is, I = c1 ep + c2 e–p ...(4)
dI
∴ = c1 ep – c2 e–p ...(5) | From (4)
dp
When p = 0,
dI
I=
∞ 1
0 1+ x
π
z2 dx =
π
2
| From (1)

and =– | From (2)

dp 2
Applying to (4) and (5), we get
c1 + c2 = π/2 and c1 – c2 = – π/2
so that, c1 = 0, c2 = π/2
π –p
∴ From (4), eI=
2

⇒ z
∞ cos px

0 1+ x 2
dx =
π –p
2
e

Differentiating w.r.t. p, we get

z
0
∞ − x sin px
1+ x 2 dx = –
π –p
2
e ⇒ z 0
∞ x sin px
1+ x
2
2
dx =
π –p
2
e .

Example 13. Find the Fourier cosine transform of e − x .

2
Sol. Fourier cosine transform of e − x is given by
2
Fc { e − x } =
Differentiating w.r.t. p, we have
z ∞

0
2
e − x cos px dx = I (say) ...(1)

dI
dp
=–
0
∞
zL 2
x e − x sin px dx =
1 ∞
2 0
2
(sin px) (– 2x e − x ) dx z
=
1
2 MN{
sin px e − x
2 ∞

0
−p
∞

0
2
}
cos px e − x dx z
(Integrating by parts)
OP
Q
dI
p ∞ − x2
=–
2 0
p
e z
cos px dx = –
p
2
I

=– dp
dp 2
p2
p2 −
4
Integrating, we have log I = – + log A or I=A e ...(2)
4

Now when p = 0, from (1), I = z

0
∞ 2
e − x dx =
2
π

π
∴ From (2), =A
2
INTEGRAL TRANSFORMS 151

p2
− x2 π − 4
Hence, I = Fc { e }= e .
2
1 F I 2
Example 14. If f c (p) =
2 GH JK
tan −1 2 , then find F(x).
p
F 2 I = tan LM 2 OP = tan F 1 I − tan F 1 I
Sol. tan −1 GH p JK
2
−1

N ( p − 1) + 1Q 2 GH p − 1JK GH p + 1JK −1 −1

F( x) =
2
π 2
1
z∞
tan G
0
F 2 I cos px dp
H p JK
−1
2

=
1 R
π z
STtan FGH p 1− 1IJK − tan FGH p 1+ 1IJK UVW cos px dp
0
∞
−1 −1

1L
= M tan G
πN
F 1 IJ cos px dp − tan FG 1 IJ cos px dpOP
z ∞

H p − 1K
0
−1
H p + 1K Q z0
∞
−1

= I1 + I2 (say) ...(1)
LMR F 1 I sin px U OP
where I1 =
1
π MNSTtan GH p − 1JK . x VW
−1
∞

0
− z
0
∞ −1
( p − 1) 2
.
1
RS1 + 1 UV .
sin px
x
dp
Q
T ( p − 1) W 2

=
1
πx z0
∞ sin px
( p − 1) 2 + 1
dp

Similarly, I2 =
−1
πx z 0
∞ sin px
( p + 1) 2 + 1
dp

From (1), I=
1
πx z RST
0
∞ sin px
( p − 1) 2 + 1
−
sin px
( p + 1) 2 + 1
UV dp
W
1 π −x π
e sin x − e − x
=
LM O
sin (− x)P = e
−x
sin x
2πx 1 1 N Q x
.
Example 15. Solve the integral equations:

R| 1,
z z
0≤t<1
∞ ∞
(i) f(x) cos λx dx = e −λ
(ii) F(x) sin tx dx = S 2, 1 ≤ t < 2.
0 0
|T 0, t≥2

Sol. (i) f ( x) =
2
π z0
∞
e – λ cos λx dλ =
2
.
1
π 1 + x2
∵ f c (λ ) = e – λ

(ii) F( x) =
2
π z∞

0
fs (t) sin tx dt =
2
π
LM
N z
0
1
sin tx dt + z 2

1
2 sin tx dt
OP
Q
LMFG − cos tx IJ − 2 FG cos tx IJ OP
2
1 2

MNH x K H x K PQ
=
π 0 1

2 L 1 cos x F cos 2x − cos x IJ OP = 2 (1 + cos x − 2 cos 2x) .

= M − −2G
π Nx x H x x K Q πx
152 A TEXTBOOK OF ENGINEERING MATHEMATICS

Example 16. Solve the integral equation

z 0
∞
F(x) cos px dx =
RS1 − p,
T 0,
0≤ p≤1
p>1

Hence prove that z 0

∞ sin 2 t
t 2
dt =
π
2
.

Sol. Let z0
∞
F( x) cos px dx = fc ( p) , then fc ( p) =
RS1 − p,
T 0,
0≤ p≤1
p>1
∴ By inversion formula for Fourier cosine transform, we have

F( x) =
2
π z ∞

0
fc ( p) cos px dp =
2
π
LM
N z
0
1
(1 − p) cos px dp + z1
∞
0 . cos px dp
OP
Q
=
2 LM
(1 − p) .
sin px
− (– 1) .
− cos px OP 1

=
2LM
−
cos x 1
+ 2 =
OP
2(1 − cos x)
π N x x2 Q 0 π N x2 x Q
πx 2
.

Deduction. Since
z 0
∞
F( x) cos px dx =
RS1 − p,
T 0,
0≤ p≤1
p>1
where F(x) =
2(1 − cos x)
πx 2

∴
2
π z ∞

0
1 − cos x
x 2 . cos px dx =
1 − p, 0 ≤ p ≤ 1
0, p > 1
RS
T
x
When p = 0, we have
2
π z
0
∞ 1 − cos x
x2
dx = 1 or
z
0
∞ 2 sin 2
x2
2 dx = π
2

Putting x = 2t so that dx = 2dt, we get z0

∞ sin 2 t
t2
π
dt = .
2

Example 17. Taking the function F(x) = RS1, 0<x<π

T0, x>π

FG 1 − cos pπ IJ . sin px dp = R|S π2 ,

show that z ∞

H p K
0
|T 0,
0<x<π
x>π
.

Sol. Consider F( x) = S
R1, 0 < x < π
T0 , x > π
Taking sine transform,

fs ( p) = z
0
∞
F( x) . sin px dx =
z π

0
1 . sin px dx =
1 − cos pπ
p
By inversion formula for Fourier sine transform.
2
π z FGH
0
∞

p
IJ
1 − cos pπ
K sin px dp =
RS
1, 0 < x < π
T
0, x > π
R| π , 0 < x < π
⇒ z FGH
0
∞ 1 − cos pπ I
p
JK sin px dp = S| 2
T 0, x > π.
INTEGRAL TRANSFORMS 153

Example 18. Find Fourier sine and cosine transform of F(x) = xn e–ax ; a > 0, n > – 1
Hence find Fourier sine and cosine transforms of
(i) xm–1 (ii) x–m.
Sol. Here F(x) = xn e–ax

fs(p) = z0
∞
xn e–ax . sin px dx = z 0
∞
xn e–ax .
Fe
GH
ipx
− e − ipx
2i
I dx
JK
=
1
2i z
LM x e
N0
∞
n − ( a − ip) x
dx − z 0
∞
x n e − ( a + ip) x dx
OP
Q
=
1
2i
LM Γ(n + 1)
N (a − ip) n+ 1
−
Γ (n + 1)
(a + ip) n + 1
OP
Q
∵ z0
∞
e − zx x n − 1 dx =
Γ(n)
zn
Put a = r cos θ, p = r sin θ
so that, a + ip = r(cos θ + i sin θ)
∴ (a + ip)n+1 = rn+1 {cos (n + 1) θ + i sin (n + 1) θ}
and a – ip = r (cos θ – i sin θ)
∴ (a – ip)n+1 = rn+1 {cos (n + 1) θ – i sin (n + 1) θ}
p
Also, r2 = a2 + p2 and tan θ =
a
Γ(n + 1) LM 2i r n+1
sin (n + 1) θ OP
Now, fs(p) =
2i N r 2(n + 1) Q
Γ(n + 1) sin (n + 1) θ Γ(n + 1) sin (n + 1)θ p
= n+1 = where tan θ =
r 2
(a + p ) 2 ( n + 1)/ 2 a
Γ(n + 1) cos (n + 1)θ p
Similarly, fc (p) = 2 2 ( n + 1)/ 2 where tan θ =
(a + p ) a
∴ We have the relations,

z0
∞
e–ax . xn sin px dx =
Γ(n + 1) sin (n + 1)θ
(a 2 + p2 ) ( n + 1)/ 2
...(1)

and z ∞

0
e–ax xn cos px dx =
Γ(n + 1) cos (n + 1)θ
(a 2 + p2 ) ( n + 1)/ 2
...(2)

p
where tan θ =
a
Put a = 0 and replace n by m – 1 in (1) and (2), we get

z
∞

0
xm–1 sin px dx =
pm
Γ(m) sin m π / 2
...(3)

and
∞

0 z
xm–1 cos px dx =
Γ(m) cos m π / 2
pm
Replace m by m + 1 in (3) and (4), we get
...(4)

z
0
∞
xm sin px dx =
Γ(m + 1) sin (m + 1) π/2
p m+ 1 =
Γ(m + 1) cos m π / 2
pm +1
...(5)
154 A TEXTBOOK OF ENGINEERING MATHEMATICS

and
0 z ∞ − Γ(m + 1) sin m π / 2
xm cos px dx =
pm+ 1
Replacing m by – m in (5) and (6), we get
...(6)

0 z
x–m sin px dx =
Γ(1 − m) cos m π / 2
p1 − m
π
π 1 cos m π / 2 ∵ Γ (m) Γ (1 − m) = ;
= . . sin mπ
sin m π Γ(m ) p1−m 0< m<1
π
π cos m
p m−1 2 π p m−1 mπ
= . = . cosec
Γ(m) 2 sin m π cos m π 2 Γ( m ) 2
2 2

Similarly, z0
∞
x–m cos px dx =
π p m−1
.
2 Γ( m )
sec
mπ
2

ASSIGNMENT

R| 0, for x < a U|
1. Find Fourier transform of F(x) = S| 1, for a < x < b . V|
T 0, for x > b W
2. (i) Show that the Fourier transform of

F(x) =
RSa −| x|,
T 0,
for | x | < a
for | x | > a > 0
(ii) Find the Fourier transform of
2 UV
is 2 (1 – cos ap). Hence show that
p W z0
∞ sin2 t
t2
dt =
π
2
.

3.
F(x) =
| x|< 1
| x|> 1
RS1,
T0,
UV
hence evaluate
W
∞ sin x

0 x
dx. z
If f(p) is the Fourier transform of F(x), prove that F[eiax F(x)] = f(p + a). F(t)
1

4. Find the Fourier transform of the single gate function (rectangular pulse)
shown in the adjoining figure.
– T/2 T/2 t

F(t)
5. Find the Fourier transform of the function F(t)as shown in the figure:
LM U|OP
R| At , 0<t<T
MNHint. F(t) = S|A,T
V
T < t < 2T |WPQ
A
M N
T
Find Fourier transform of F(x) = S0,
Rx , 2
| x |< a UV
6.
T | x|> a . W O T 2T t
7. Find Fourier sine and cosine transform of
Fig. of Q. 5
(i) F(x) = e–x, x ≥ 0 (ii) F(x) = 2e–5x + 5e–2x

(iii) F(x) = cosh x – sinh x (iv) F(x) =

RS1, 0≤x<a UV
T0 , x>a W
(v) x e–ax, a > 0.
INTEGRAL TRANSFORMS 155

x, R| for 0<x<1 U|
8. Obtain Fourier cosine transform of F(x) = 2 − x, S| for 1< x < 2 V|
0, T for x>2 W
Also find its Fourier sine transform.
9. Find Fourier sine transform of
R|
0, 0 < x < a U| 1
S|
(i) F(x) = x, a ≤ x ≤ b V| (ii) F(x) =
x ( x 2 + a2 )
T
0, x > b W
2 eax + e− ax cosh ax
(iii) F(x) = x e− x /2
(iv) F(x) = or ,–π<a<π
eπx − e− πx sinh πx

e − ax − e −bx
(v)
x
1 1
10. Find Fourier sine transform of and deduce that fs (cosec h πx) = tanh (p/2)
e −e πx − πx 2
11. Find Fourier cosine transform of F(x) = sech πx.
Fe I
GH p JK and hence evaluate F FGH 1p IJK .
− ap
12. Find Fs–1 s
–1

13. By taking e–ax for F(x); a > 0, x > 0; show that

14.
z
0 a + p
∞ cos px
2
π –ax
2a
e 2 dp =
and
Find F(x) if its Fourier sine transform is
z ∞

0
p sin px
a2 + p2
dp =
π –ax
2
e .

(i)
p
(ii)
π
(iii) (2πp)1/2 (iv)
RSsin p, 0< p< π
.
UV
1 + p2 2
2 2
T 0, p≥π W
2 2
15. Find Fourier cosine transform of e− a x
and hence evaluate Fourier sine transform of x e− a x
.
1
16. Find Fourier sine and cosine transform of .
x

dn
17. If f(p) is the Fourier transform of F(x), prove that F[xn F(x)] = (– i)n {f(p)}.
dpn

1 2
/( p2 + 1)}
18. A certain function of time F(t) has the following Fourier transform f(p) = e− { 2 p
p2 + 1
Using the properties of the Fourier transform, obtain the Fourier transforms of
(i) F(2t) (ii) F(t – 2)
(iii) F(t) cos 2t (iv) e2it F(t).
19. State and prove the convolution theorem for the Fourier transform. Verify this theorem for the
functions f(t) = e–t and g(t) = sin t.

Answers
( eibp − eiap ) 2 sin p π
1. 2. (ii) ,p≠0;
ip p 2
FG pT IJ FG where sin c θ = sin θ IJ A
( eipT − 1) +
A 2ipT
e
4. T sin c H 2 KH θ K
5.
Tp 2 ip
156 A TEXTBOOK OF ENGINEERING MATHEMATICS

F 2a 2
4 I
6. GH p − J sin ap + 4pa cos ap
p K
3 2

p
;
1 F 2 5 I ; 10 F 1 1 I
7. (i)
1 + p2 1 + p2
(ii) p GH p 2
+ 25
+
p 2 J GH p
+ 4K 2
+ 25
+ 2
p
J
+ 4K

p
;
1 FG 1 − cos ap IJ ; sin ap 2ap
;
a2 − p2
(iii)
1+ p 2
1+ p 2 (iv)
H p K p (v) 2
(a + p ) 2 2
(a2 + p2 )2
2 2
8. 2 cos p (1 – cos p) ; sin p (1 – cos p)
p p2

( a cos ap − b cos bp) (sin bp − sin ap) π

9. (i) + (ii) (1 – e–ap)
p p2 2a2
π 2 sinh p FG p IJ − tan FG p IJ −1
(iii)
2
p e−( p / 2) (iv)
2 (cos a + cosh p)
(v) tan–1
H aK H bK
10.
1
tanh
p
11.
1
sech
p
12.
2 F xI
tan GH JK ; 1 –1
4 2 2 2 π a
1 1 2 sin πx
14. (i) e–x (ii) (iii) (iv) π .
x x x (1 − x 2 )
π − ( p2 / 4 a 2 ) p π − ( p2 / 4 a 2 ) π π
15. e ; e 16. ,
2a 4 a3 2p 2p
F 2p I 2

2
− GH p + 4 JK
2 1 2
/ ( p2 + 1)}
18. (i) e (ii) e2ip . 2
e− {2 p
p2 + 4 p +1

LM |RS 2( p + 2) |UV2
|RS 2( p − 2) |UV O 2
|RS 2( p + 2) U|V
2

T| ( p − 2) + 1 W| PP
− −
(iii)
1
MM ( p + 21) e T| ( p + 2) + 1 W| +
2 1
e
2
(iv)
1
e
−
T| ( p + 2) + 1 W| .
2

2
MN
2
+1 ( p − 2)2 + 1 PPQ ( p + 2)2 + 1

2.11 FOURIER TRANSFORMS OF THE DERIVATIVES OF A FUNCTION

Let u ( p, t ) be the Fourier transform of the function u(x, t).

Then

∂u
u ( p, t) = z
−∞
∞
u( x, t) e ipx dx

∂ 2u
Suppose u and both vanish as x → ± ∞. Then the Fourier transform of is given
∂x ∂x 2
by

F
RS ∂ u UV =
2

T ∂x W 2 z −∞
∞ ∂ 2u ipx
∂x 2
e dx

LMRe OP
=
MNST
ipx ∂u
∂x
− ipe ipx . u
UV
W
∞

−∞
+ z−∞
∞
(ip) 2 e ipx . u dx = − p2
PQ z
−∞
∞
ue ipx dx = − p2 u

F ∂ uI 2
Hence, F GH ∂x JK = – p
2
2
u where u = F(u)
INTEGRAL TRANSFORMS 157

If us ( p, t) and uc ( p, t) be the Fourier sine and cosine transforms of u(x, t), then

us ( p, t) = z
0
∞
u sin px dx and

∂ 2u
u ( p, t) = z ∞

0
u cos px dx

The Fourier sine transform of is given by

∂x 2

Fs
RS ∂ u UV =
2

T ∂x W2 z
0
∞ ∂ 2u
∂x 2
sin px dx

LMRsin px ∂u − p cos px . uU − p sin px . u dxOP

MNST
=
∂x
VW PQ
∞

0
z0
∞
2

L R U − p u sin px dxOP R U
= M pSuV
MN T W x=0
z
0
∞

PQ = pSTuVW − p u
2

x=0
2
s

F ∂ uI 2
F G
H ∂x JK = p(u) – p u
Hence, 2
where u = F (u)
s 2 x=0 s s s

Similarly, we have,

F ∂ u I = – FG ∂uIJ
2
Fc GH ∂x JK H ∂x K
2
x=0
– p2 u c where uc = Fc(u)

In general, the Fourier transform of the nth derivative of F(x) is given by

F d FI n
F GH dx JK = (– ip)
n
n F{F(x)}

Note. s may also be used as a parameter in place of p.

2.12 CHOICE OF INFINITE FOURIER SINE OR COSINE TRANSFORM

∂ 2u
For exclusion of from a differential equation, we require
∂x 2
∂u FG IJ
(i) (u)x = 0 in sine transform. (ii) ∂x
x=0
in cosine transform. H K
2.13 APPLICATIONS OF FOURIER TRANSFORMS TO HEAT CONDUCTION
(TRANSFER) EQUATIONS

In one dimensional heat transfer eqns, the partial differential equation can easily be trans-
formed into an ordinary differential equation by applying Fourier transforms. The required
solution is then obtained by solving this equation and inverting by means of the complex inversion
formula. This is illustrated through the following examples.
158 A TEXTBOOK OF ENGINEERING MATHEMATICS

EXAMPLES

∂u ∂ 2 u
Example 1. Solve the equation = , x > 0 , t > 0 subject to the conditions
∂ t ∂x 2

(i) u = 0 when x = 0, t > 0 (ii) u =

RS
1, 0 < x < 1
when t = 0
0, x ≥ 1 T
and (iii) u(x, t) is bounded. (G.B.T.U. 2011; U.P.T.U. 2009, 2015)
Sol. Since (u)x = 0 is given, taking Fourier sine transform of both sides of the given
equation, we have

z
0
∞ ∂u
∂t
sin px dx = z ∞

0
∂ 2u
∂x 2
sin px dx

Hz
d FG ∞
u sin px dx = p(u)
IJ
⇒
dt 0 K2
x=0– p u

⇒
du
dt
+ p2 u = 0
2
...(1) where u = z0
∞
u sin px dx

Solution to (1) is u = c1 e − p t , ...(2)

where c1 is a constant.
Now, when t = 0, Fourier sine transform of u(x, t)

( u )t = 0 = z0
∞
u(x, 0) sin px dx = z
0
1
1. sin px dx + z
1
∞
0 . sin px dx

F − cos px IJ
=G
1
=
1 − cos p
H p K 0
p
∴ From (2), ( u ) t = 0 = c1
1 − cos p
⇒ c1 =
p

∴ From (2), u =
FG 1 − cos pIJ e− p
2
t
H p K
Applying inverse Fourier sine transform, we have

u(x, t) =

which is the required solution.

2
π z FGH
∞

0 p
IJ
1 − cos p − p2t
e
Ksin px dp

Example 2. Determine the distribution of temperature in the semi-infinite medium x ≥ 0

when the end x = 0 is maintained at zero temperature and the initial distribution of tempera-
ture is F(x).
Sol. Let u(x, t) be the temperature at point x at any time t. Heat flow equation is
∂u ∂ 2u
= c2 , (x > 0, t > 0) ...(1)
∂t ∂x 2
Subjected to the initial condition u(x, 0) = F(x) ...(2)
and the boundary condition u(0, t) = 0 ...(3)
INTEGRAL TRANSFORMS 159

Taking Fourier sine transform of eqn. (1), we get

z0
∞ ∂u
∂t
sin px dx = c2 z ∞

0
∂ 2u
∂x 2
sin px dx

du
⇒ = c2 [p (u)x = 0 – p2 u ] = – c2p2 u ∵ (u) x = 0 = 0
dt

⇒
du
dt
+ c 2 p2 u = 0
2 2
...(4) where u = z
0
∞
u sin px dx

−c p t
Solution to (4) is u = c1 e ...(5)
Taking Fourier sine transform of (2), we get

From (5),
( u )t = 0 = z
0
( u ) t = 0 = c1
∞
F( x) sin px dx = fs(p)
⇒ c1 = fs(p)
| say

2 2
−c p t
∴ From (5), u = fs(p) e
Now Taking its inverse Fourier sine transform, we get

u(x, t) =
2
π z ∞

0
fs ( p) e − c
2
p2 t
sin px dp

∂u ∂ 2u
Example 3. Use Fourier sine transform to solve the equation =2 under the
∂t ∂x 2
conditions
(i) u(0, t) = 0 (ii) u(x, 0) = e–x (iii) u(x, t) is bounded.
Sol. The given equation is
∂u ∂ 2u
=2 ...(1)
∂t ∂x 2
Taking Fourier sine transform on both sides of eqn. (1), we get

z 0
∞ ∂u
∂t
sin px dx = 2
du
∞ ∂ 2u

0 ∂x 2 zsin px dx

⇒ = 2 [p(u)x = 0 – p2 u ]
dt
⇒
du
dt
+ 2p2 u = 0
2
where u = z
0
∞
u sin px dx

Its solution is u = c1 e − 2 p t
...(2)
where c1 is a constant.

At t = 0,

From (2),
( u )t = 0 = z
0

( u ) t = 0 = c1
∞
(u) t = 0 sin px dx = z0
∞
e − x sin px dx =
p
1 + p2
...(3)

...(4)
p
∴ From (3) and (4), c1 =
1 + p2
p 2
From (2), u = 2
e− 2 p t
...(5)
1+ p
Taking inverse Fourier sine transform, we get

u(x, t) =
2
π z ∞

0 1+ p
p
2
2
e −2 p t sin px dp.
160 A TEXTBOOK OF ENGINEERING MATHEMATICS

Example 4. The temperature u in the semi-infinite rod 0 ≤ x < ∞ is determined by the

∂u ∂ 2u
differential equation =k subject to conditions
∂t ∂x 2
∂u
(i) u = 0 when t = 0, x ≥ 0 (ii) = – μ (a constant) when x = 0 and t > 0
∂x

Making use of cosine transform, show that u(x, t) =

2μ
π z
0
∞ cos px
p 2
2
(1 − e − kp t ) dp.

Sol. Taking Fourier cosine transform on both sides of given equation, we get

z ∞

0
∂u
∂t
. cos px dx = k zL 0
∞ ∂ 2u
∂x 2
cos px dx

OP
⇒
du
dt
=k −
∂u
∂x
MN FGH IJK x =0
− p2 u
Q
du
= kμ – kp2 u where u = z0
∞
u cos px dx

⇒ + kp2 u = kμ ...(1)
dt
2
kp t
I.F. = e
Solution to (1) is

u . e kp2 t
= zμ
2
kμ . e kp t dt + c1 =

2
μ
p 2
e kp2 t
+ c1

⇒ u = 2
+ c1 e − kp t
...(2)
p
At t = 0, ( u )t = 0 = z ∞

0
μ
(u) t = 0 cos px dx = 0 ...(3)

From (2), ( u )t = 0 =
+ c1 ...(4)
p2
μ
⇒ c1 = – 2 | From (3) and (4)
p
μ − kp2 t
∴ From (2), u = 2 (1 – e )
p
Taking inverse Fourier cosine transform, we get

u=
2 ∞ μ
π 0 p2
2

z
(1 − e – kp t ) cos px dp.

Example 5. (i) If the initial temperature of an infinite bar is given by

θ(x) =
RSθ , for| x|< a
0
T0, for| x|> a
determine the temperature at any point x and at any instant t.
(ii) If the initial temperature of an infinite bar is given by
RS1,
μ(x, 0) = 0, otherwise
for − c < x < c UV
T W
determine the temperature of an infinite bar at any point x and at any time t > 0.
INTEGRAL TRANSFORMS 161

Sol. (i) To determine the temperature θ(x, t), we have to solve the heat-flow equation
∂θ ∂ 2θ
= c2 2 , t>0 ...(1)
∂t ∂x

subject to the initial condition θ( x, 0) =

RSθ , for|x|< a
0 ...(2)
T0, for|x|> a
Taking Fourier transform of (1), we get

z ∞∂θ ipx
−∞ ∂t
e dx = c 2 z−∞
∞ ∂ 2θ
∂x 2
e ipx dx

or
d
dt z ∞

−∞
θ e ipx dx = − c 2 p2 θ

or
dθ
dt
= − c 2 p2 θ
Now taking the Fourier transform of (2), we get
...(3) where θ = z
–∞
∞
θ e ipx dx

z z L e OP ipx a
∞ a
θ ( p, 0) = θ( x, 0) e ipx
dx = θ e dx = θ M
ipx
−∞ −a
0
N ip Q 0
−a

LM e ipa
−e − ipa OP = 2θ 0
LM e − e OP = 2θ sin pa
ipa − ipa
0
= θ0 ...(4)
N ip Q p N 2i Q p
dθ
From (3), = − c 2 p2 dt
θ
2
p2 t
Integrating, log θ = − c 2 p2 t + log A or θ = A e−c

2θ 0 sin pa 2θ 0 sin pa
Since θ = when t = 0, from (4), we get A=
p p
2θ 0 sin pa − c2 p2 t
∴ e θ=
p
Taking its inverse Fourier transform, we get

θ( x, t) =
1
2π z ∞

−∞
2θ 0 sin ap − c2 p2t − ipx
p
.e .e dp

=
θ0
π z ∞

−∞
sin ap − c2 p2 t
p
.e (cos xp − i sin xp) dp

Nz z
=
θ0 LM ∞ sin ap − c2 p2t
.e cos xp dp − i
∞ sin ap − c2 p2 t
.e sin xp dp
OP
π −∞ p −∞ p Q
=
θ0
π z −∞
∞sin ap − c 2 p2t
p
.e cos xp dp

(The second integral vanishes since its integrand is an odd function)

=
2θ 0
π z ∞

0
sin ap − c 2 p2 t
p
.e
θ
cos xp dp = 0
π z 0
∞

p
2
e−c p t
2

. 2 sin ap cos xp dp

=
θ0
π z ∞

0
e−c
2
p2 t FG sin (a + x) p + sin (a − x) pIJ dp
H p K
162 A TEXTBOOK OF ENGINEERING MATHEMATICS

=
θ0
π z
0
∞ 2
e − v sin
RS
T
(a + x)v
c t
+ sin
(a – x)v dv
c t v
UV
where v2 = c2p2t
W
=
θ0
2
erf
RS F
a+x
T GH
2c t
+ erf
a− x
2c t
IJ ∞
−t 2 FG dt π IJ UV
. ∵ 0 e sin ( ct / x ) . t = 2 erf
K H KW
c
2 x
. z FG
H
IJ
K
(ii) To determine the temperature μ(x, t), we have to solve the heat-flow equation
∂μ ∂ 2μ
= k2 ,t>0 ...(1)
∂t ∂x 2
subject to the initial condition μ(x, 0) = 1, for − c < x < c ...(2) UV
0, otherwise W
Taking Fourier transform of (1), we get

z ∞ ∂μ ipx
− ∞ ∂t
. e dx = k2
∞ ∂2μ

− ∞ ∂x 2
. e ipx dx z
⇒
d ∞
dt − ∞z μ e ipx dx = – k2p2 μ

⇒
dμ
dt
= – k2 p2 μ where μ = z
−∞
∞
μ e ipx dx

dμ
⇒ = – k2p2 dt
μ
Integrating, we get
log μ = – k2p2t + log A | A is a constant
2 2
or, μ = A e −k p t ...(3)
Now, taking Fourier transform of (2), we get

μ (p, 0) = z−∞
∞
μ(x, 0) eipx dx

= zc

−c
eipx dx =
e ipc − e − ipc
ip
=
2
p
sin cp ...(4)

2
From (3), At t = 0, sin cp = A
p
2 2 2
∴ μ = sin cp e − k p t
p
Taking its inverse Fourier transform, we get

μ(x, t) =
which is the required solution.
1 ∞ 2
2π − ∞ p
2 2

z
sin cp e − k p t . e–ipx dp

Example 6. Use Fourier cosine transform to show that the steady temperature u in the
semi-infinite solid y > 0 when the temperature on the surface y = 0 is kept at unity over the strip
| x | < a and at zero outside the strip is
1 LM
tan −1
a+ x FG
+ tan −1
a−x IJ FG IJ OP
π MN y H y K H K PQ
The results z
0
∞
e − px x −1 sin rx dx = tan −1
FG r IJ (r , p > 0) may be assumed.
H pK
INTEGRAL TRANSFORMS 163

∂ 2u ∂ 2u
Sol. Taking Fourier cosine transform of + = 0 , we have
∂x 2 ∂y 2

z0
∞ ∂ 2u
∂x 2
. cos px dx + z ∞

0
∂ 2u
∂y 2
. cos px dx = 0

FG ∂uIJ
–
H ∂x K x=0
− p2 u +
d2
dy 2
(u ) = 0 where u = z0
∞
u cos px dx

d2 u ∂u
⇒ − p2 u = 0 ∵ → 0 as x → ∞ ...(1)
dy 2 ∂x
Its solution is u = c1e py + c2 e − py ...(2)
But u is finite so c1 = 0, otherwise u → ∞ as y → ∞
∴ From (2), u = c2 e − py ...(3)
But u= z ∞
u cos px dx

z
0

From (3),
(u ) y = 0 =

(u ) y = 0 = c2
z
0
∞
(u) y = 0 cos px dx =
0
a
1 . cos px dx =
sin pa
p
...(4)

sin pa
∴ c2 =
p
sin pa − py
∴ From (3), u= e
p
Applying inverse Fourier cosine transform, we get

u=
2
π z ∞

0
sin pa − py
p
e cos px dp =
1
π z 0
∞ e − py
p
(2 sin pa cos px) dp

=
1
π z ∞

0
e − py

p
[sin (a + x) p + sin (a − x) p] dp

=
1 LM
tan −1
a+ x FG
+ tan −1
a−x IJ FG IJ OP .
π N y H y K H KQ
ASSIGNMENT
1. Apply appropriate Fourier transform to solve the partial differential equation
∂V ∂2 V
= ; x > 0, t > 0
∂t ∂x2
subject to the conditions
RS
x, 0 ≤ x ≤ 1
(i) Vx (0, t) = 0
T
(ii) V(x, 0) = 0, x > 1 (iii) V(x, t) is bounded.

∂u ∂ 2u
2. Solve = k 2 for x ≥ 0, t ≥ 0 under the conditions
∂t ∂x
(i) u(0, t) = u0, t > 0 (ii) u(x, 0) = 0, x ≥ 0 (iii) u(x, t) is bounded.
164 A TEXTBOOK OF ENGINEERING MATHEMATICS

3. Using the method of Fourier transform, determine the displacement y(x, t) of an infinite string
given that the string is initially at rest and that the initial displacement is f(x), – ∞ < x < ∞. Show
1
that the solution can also be put in the form y(x, t) = [ f ( x + ct) + f ( x − ct)] .
2
4. Using Fourier transform, solve
∂V ∂2 V
= , − ∞ < x < ∞, t > 0; V ( x, 0) = f ( x)
∂t ∂x2
5. Using Fourier transforms, solve the following initial boundary value problem:

∂u 2
2 ∂ u
1, − 1 < x < 0 U|
∂t
=c
∂x 2
, − ∞ < x < ∞, t > 0 ; u( x, 0) = −1, 0 < x < 1 V|
0, otherwise W
Answers

1. V(x, t) =
2
π z FGH F
∞

0
sin p cos p − 1 −
p
+
p2
e
I
JK p2 t
cos px dp

2. u(x, t) =
2u0
π z GGH
0
∞ 1 − e− kp
p
2
t I
JJ sin px dp.
K
4. V(x, t) =
1
2π z ∞

−∞
2
F( p) e− p t . e− ipx dp where F( p) =
z
−∞
∞
f ( x) eipx dx

5. u(x, t) =
1
iπ z−∞
∞ 1
p
2 2
(1 − cos p) e − c p t e − ipx dp

2.14 FINITE FOURIER TRANSFORMS

The finite Fourier sine transform of F(x), 0 < x < l is defined as

fs ( p) =
0
l
F( x) . sin
p πx
l z
dx ; p ∈ I
Similarly, the finite Fourier cosine transform of F(x), 0 < x < l is defined as

fc ( p) =
0
l
F( x) . cos
p πx
l z
dx ; p ∈ I
Generally, the choice of π as the upper limit of integration in these transforms is found
convenient and can easily be arranged by having suitable substitutions to actual problems,
then

Note. fs(p) is always zero when p = 0.

fs ( p) = z
0
π
F( x) . sin px dx and fc ( p) = z
0
π
F( x) . cos px dx.

2.15 INVERSE FINITE FOURIER TRANSFORMS

Inversion formulae are given as follows:

(1) When upper limit is π
For sine transform:
∞
2
F( x) =
π ∑
p=1
fs ( p) sin px
INTEGRAL TRANSFORMS 165

For cosine transform:

F( x) =
1
π
fc (0) +
2
π
∞

∑ f ( p) cos px where f (0) stands for

p=1
c c z
0
π
F( x) dx .

(2) When upper limit is l

For sine transform:
∞
2 p πx
F( x) =
l ∑ f ( p) sin
p =1
s
l
For cosine transform:

F( x) =
1
l
fc (0) +
2
l p =1
∞

∑ f ( p) cosc
pπ x
l
where fc(0) stands for z 0
l
F( x) dx.

EXAMPLES
x
Example 1. Find finite Fourier sine transform of F(x) = 1 − .
π
Sol. The finite Fourier sine transform of F(x) is given by,

fs ( p) = z FGH
π

0
IJ
K 1−
x
π
sin px dx

R x F cos px IJ UV − FG − 1IJ . FG − cos px IJ dx

= SFG 1 − IJ G −
z
π
π

TH π K H p K W H π K H p K 0
0

1 F sin px I
π
1
= − G
p πp H p K
J = 1p . 0
cos k( π − x)
Example 2. Find finite Fourier cosine transform of F(x) = − .
k sin kπ
Sol. fc ( p) = z
− cos k (π − x)
0
π

k sin kπ
cos px dx

= –
1
2k sin kπ 0
π
z
[cos { k (π – x) + px} + cos { k(π – x) – px}] dx

= –
1 LM
sin (kπ – kx + px) sin (kπ – kx – px)
–
OP π

2k sin kπ p– k N p+k Q 0

=
1 1
−
1
= 2
1 LM
, k ≠ 0, 1, 2, ...
OP
2k p − k p + k p − k2 N Q
Example 3. Find F.F.S.T. and F.F.C.T. of F(x) = 2x, 0 < x < 4.

Sol. (i) fs ( p) = zL
0
4
2 x . sin
pπx
4
dx (Here l = 4)

MM R|| − cos p4πx U||OPP R| − cos pπx U| R 32

4

MM | FG pπ IJ V|PP
= 2x . S − z0
4
2.S
|
F pπ I
4 | dx |
V
|| GH 4 JK || |T
= S pπ
(1 − cos pπ) , p ≠ 0

N |T H 4 K |WQ 0 T W 0, p=0
166 A TEXTBOOK OF ENGINEERING MATHEMATICS

(ii) fc ( p) = z
0
4
2 x . cos
pπx
4
dx

LM R sin pπx UOP R| sin pπx U|

4

|| 4 ||P
= M2 x . S
MM | FG pπ IJ V|PP
− 2.S
| 4 |V dx
z
|| FGH p4π IJK ||
0
4

N |T H 4 K |WQ T
0 W
R| − cos pπx U| 4

32 8 | 4 | = 32 (cos pπ − 1), p ≠ 0
= sin pπ − S
pπ | F pπ I |
V pπ 2 2
pπ
G J
|T H 4 K |W 0

When p = 0, zR
fc ( p) = fc (0) =
4

0
2 x dx = 16 .

U|
Example 4. Find F(x) if fc (p) =
|S sin2ppπ/2 , p = 1, 2 , ...
V| where 0 < x < 2π.
|T π/4, p=0 W
∞
1 π 2 sin pπ/2 pπx
Sol. F( x) = . +
2π 4 2π
∑
p=1
2p
. cos
2π
∞
1 π 2 sin pπ / 2 px
= . +
2π 4 2π ∑
p =1
2p
cos
2
∞
1 1 sin ( pπ/2) px
F( x) = +
8 π ∑
p =1
2p
cos
2
.

ASSIGNMENT
1. Find finite Fourier sine and cosine transforms of
(i) F(x) = x2, 0 < x < π (ii) F(x) = 1, 0 < x < π (iii) F(x) = x, 0 < x < π.

2. Find finite Fourier cosine transform of F( x) = 1 −

FG x IJ 2

H π K .

3. Find finite Fourier sine transform of

x
(i) F(x) = (ii) F(x) = sin nx, n ∈ I (iii) F(x) = x(π2 – x2)
π
(iv) F(x) = x (π – x) (v) F(x) = ecx (vi) cos mx.
4. Find finite Fourier cosine transform of

(i) F( x) =
π
−x+
x2
(ii) sin nx, n ∈ I (iii) F( x) =
1, 0 < x < π / 2
.
UV
3 2π − 1, π / 2 < x < π W
R|x, 0 ≤ x ≤ π / 2 U|
5. Find finite Fourier sine transform of F( x) = S|π − x, π ≤ x < πV| .
T 2 W
6. Find inverse finite Fourier sine transform of
2π (− 1) p− 1 1 − cos pπ
(i) fs ( p) = , p = 1 , 2 , ... (0 < x < π) (ii) fs ( p) = , 0 < x < π.
p3 p2 π 2
INTEGRAL TRANSFORMS 167

Answers

R|− π (− 1) + 2 [(− 1)
2 p
U R| 2π(− 1) , p ≠ 0U|
p

p ≠ 0| | |V
p
− 1], 2
1. (i) (a) f ( p) = S p p 3
V (b) f ( p) = S p
s
|T 0, p = 0|W
c
|| , p = 0||
π 3

T 3 W
R| 1 {1 − (− 1) p}, p ≠ 0U| R0, if p ≠ 0UV
(ii) (a) f ( p) = S p V (b) f ( p) = S
s
|T 0, p = 0|W
c
Tπ, if p = 0W
R| {(− 1) p − 1} , p ≠ 0U|
R| π (− 1) p+1, p ≠ 0U| | p2 |V
(iii) (a) f ( p) = S p V (b) f ( p) = S
s
|T 0, p = 0|W
c
|| π , 2
p = 0|
|
T 2 W
R 2 , p ≠ 0U
2.
| 2
f ( p) = S πp
|V
c
|| , p = 0||
π
T3 W
3.
(− 1) p+ 1
(i) (ii) π
0, if p ≠ n
(iii)
UV 6π
(− 1) p+ 1
p / 2, if p = n W p3
2 p p p
(iv) 3 [1 − (− 1) ] (v) [1 − (− 1) p eπc ] (vi) [1 − (− 1) p cos mπ]
p p + c2
2
p − m2
2

1 2n
4. (i) 2 (ii) 2 ; (n − p) is odd and 0 if even
p n − p2
2 pπ 2 pπ
(iii) sin ; p ≠ 0 and 0 if p = 0 5. 2 sin
p 2 p 2
2
∞
2π(− 1) p− 1 2
∞
F 1 − cos pπ I
6. (i)
π ∑
1
p3
sin px (ii)
π3 ∑ GH p JK sin px.
1
2

2.16 PARSEVAL’S IDENTITY FOR FOURIER TRANSFORMS

M.A. Parseval (1755–1836), a French Mathematician, gave the following result.

If the Fourier transforms of F(x) and G(x) are f(p) and g(p) respectively, then

(i)
1 ∞
2π – ∞ z
f ( p) g ( p) dp = ∞ F( x) G ( x) dx
–∞ z
(ii)
1 ∞
2π − ∞ z
| f ( p)|2 dp =
∞
−∞ z
|F( x)|2 dx
where bar implies the complex conjugate.

Proof. z ∞
F( x) G ( x) dx = z ∞
F( x)
RS 1 z ∞ UV
g ( p) e ipx dp dx
Using inversion formula
−∞ −∞ T 2π −∞ W for Fourier transform

=
1
2π z−∞
∞
g ( p) RSz
T−∞
∞
F( x) e ipx dxUV dp
W
|Changing order of integration

=
1
2π z−∞
∞
f ( p) g ( p) dp ...(1) | By definition
168 A TEXTBOOK OF ENGINEERING MATHEMATICS

Now, take G(x) = F(x) in (1), we get

1
2π z −∞
∞
f ( p) f ( p) dp = z ∞

−∞
F( x) F( x) dx

⇒
1 ∞
2π – ∞

Hence the results.

z 2
| f ( p)| dp =
∞
– ∞
| F( x)|2 dx z ...(2)

Corollary 1. Following Parseval’s identities for Fourier cosine and sine transforms can be
proved as above:

(i)
2
π z
0
∞
f c ( p) gc(p) dp =
z
0
∞
F( x) G(x) dx (ii)
2
π z ∞

0
|f c ( p)|2 dp = z0
∞
|F( x)|2 dx

(iii)
2
π z
0
∞
f s ( p) gs(p) dp = z
0
∞
F( x) G ( x) dx (iv)
2
π z ∞

0
|f s ( p)|2 dp = z0
∞
| F( x)|2 dx .

EXAMPLES

Example 1. Using Parseval’s identity, show that z

0
∞ x2
(a + x ) (b 2 + x 2 )
2 2
dx =
π
2 (a + b)

Hence find z0
∞ x2
( x 2 + 1) 2
dx .

p
Sol. If F(x) = e– ax then fs(p) =
a + p2 2

∴ By Parseval’s identity for sine transform,

2
π z
0
∞
f s ( p) g s ( p) dp = z
0
∞
F( x) G ( x) dx

⇒
2
π z ∞
0
p
a 2 + p 2 b2 + p 2
⋅
p
dp =
z
0
∞
e – ax . e – bx dx

LM OP
z
∞
∞ p2 π e
– ( a + b) x
π
⇒ dp = =
0 (a 2 + p2 )(b2 + p2 ) MN
2 – (a + b) PQ 0
2(a + b)
Thus,

z ∞

(a 2 + x 2 ) (b2 + x 2 )
0
x2
dx =
π
2(a + b)
For a = b = 1, we get

z ∞

0 ( x + 1)2
x2
2
dx =
π
4

z
2
∞ sin ax π (1 − e − a )
Example 2. Using Parseval’s identity, show that dx = .
0 x( a 2 + x 2 ) 2 a2
INTEGRAL TRANSFORMS 169

Sol. Let F(x) = e–ax and G(x) =

RS1, 0< x<a UV
T0, x>a W
a sin ap
then fc (p) = 2 2
and gc(p) =
a +p p
Parseval’s identity for Fourier cosine transform is
2
π z ∞

0
f c ( p) gc (p) dp = z
0
∞
F( x) G ( x) dx

z z
2
2 ∞ a sin ap ∞
1 – e–a
⇒ dp = e – ax . 1 dx =
π 0 p ( a 2 + p2 ) 0 a

∴ z ∞

0
sin ax
2
x( a + x )
2
dx =
2a
π
2
(1 – e – a )
2

ASSIGNMENT

1. Using Parseval’s identity, show that z ∞

0
dx
(a2 + x 2 ) (b2 + x 2 )
=
π
2ab (a + b)
.

Hence evaluate z
0
∞ dx
( x 2 + 1) 2

2. If F(x) =
RS1,
T0,
| x| < a
| x| > a
UV
, prove that
W z
0
∞ sin 2 ax
x2
dx =
πa
2
using Parseval’s identity.

3. Evaluate using Parseval’s identity:

(i) z
0
FG 1 – cos x IJ dx
∞
H x K
2
(ii) z
0
∞ sin 4 x
x 2
dx
LMHint: Take F( x) = 1,
N 0,
0≤ x≤1
x≥1
OP
Q
z
FG sin x IJ
∞
4
dx = π/3
LMHint: Take F( x) = RS1 –| x|, | x|< 1 UVOP
4. Prove that:
H x K
0
N T 0, otherwise WQ
5. Use Parseval’s identity to prove that: z
0
∞ ( x cos x – sin x) 2
x 6
dx =
π
15

Answers

1. π/4 3. (i) π/2 (ii) π/2

2.17 THE Z-TRANSFORM

The Z-transform plays an important role in the field of Communication Engineering and Control
Engineering at the stage of analysis and representation of discrete-time linear shift invariance
system. When continuous signals are sampled, discrete-time functions arise. The application
of Z-transform in discrete time systems is similar to that of the Laplace transform in continuous
time systems.
170 A TEXTBOOK OF ENGINEERING MATHEMATICS

2.18 DEFINITIONS

2.18.1. One-Sided Z-transform

Let {f(k)} be a sequence defined for all positive integers ‘k’. Then the Z-transform of f(k) is
defined as
∞
Z{f(k)} = F(z) =
k=0
∑ f (k) z −k
...(1)

where z is an arbitrary complex number and Z is an operator of Z-transform.

This is one-sided Z-transform.

2.18.2. Two-Sided Z-transform

If {f(k)} is a sequence defined for k = 0, ± 1, ± 2, ....
∞
then Z{f(k)} = F(z) =
k=−∞
∑ f ( k) z − k ...(2)

where z is an arbitrary complex number and Z is an operator of Z-transform.

This is two sided Z-transform.
Note 1. If f(k) = 0 for k < 0 then {f(k)} is called a casual sequence.
−1
2. If f(k) is a non-casual sequence, f(k) = 0 for k ≥ 0, then its Z-transform is F(z) = ∑ f (k) z
−∞
−k
and
is also called one sided Z-transform.
3. The curly bracket { } represents sequence. Sequence {f(k)} is an ordered list of real or complex
numbers.
4. The infinite series on R.H.S. of (1) will be convergent only for certain values of z depending on
sequence {f(x)}.
5. The inverse Z-transform of Z{f(k)} = F(z) is defined as Z–1[F(z)] = {f(k)}.

2.19 UNIT STEP AND UNIT IMPULSE SEQUENCES

Unit step sequence is defined as 1

u(k) =
RS1, k≥0 UV
T0, k<0 W
Unit impulse sequence is defined as –2 –1 0 1 2 3 k

δ(k) =
RS1, k=0 UV 1

T0, k≠0 W δ(k)

–2 –1 0 1 2 3 k

2.20 RELATION BETWEEN UNIT IMPULSE SEQUENCE AND UNIT STEP

SEQUENCE

∞
u(k) = ∑ δ(k)
k=−∞
and δ(k) = u(k) – u(k – 1)
INTEGRAL TRANSFORMS 171

We have δ(n – k) =
RS1, k=n UV
T0, k≠n W
∞
Also, f(n) = ∑
k=−∞
f (k) δ (n − k) .

2.21 Z-TRANSFORM OF UNIT IMPULSE FUNCTION

We know that δ(k) =

1, k = 0 UV
0, k ≠ 0 W
∞
∴ Z{δ(k)} = ∑ δ(k) z
k=−∞
−k
= [...... + 0 + 0 + 1 + 0 + 0 + ......] = 1

Hence, Z{δ(k)} = 1

2.22 Z-TRANSFORM OF DISCRETE UNIT STEP FUNCTION

We know that u(k) =

1, k ≥ 0 UV
0, k < 0 W
∞ ∞
∴ Z{u(k)} = ∑ u( k) z − k = ∑z
k=0
−k

k=−∞

1 z
= 1 + z–1 + z–2 + ...... = −1 = ; | z | > 1.
1− z z−1

2.23 Z-TRANSFORM FOR DISCRETE VALUES OF t

If f(t) is a function defined for discrete values of t, where t = nT, n = 0, 1, 2, ......, T being the
sampling period, then Z-transform of f(t) is defined as
∞
Z{f(t)} = ∑ f (nT)z
k=0
−k
= F( z ).

Note. The important element of discrete-time systems is the samples in which a switch close to admit
an input signal in every T seconds. A samples is a conversion device which converts a continuous signal
into a sequence of pulses occurring at sampling instants 0, T, 2T, ...... where T is the sampling period.

EXAMPLES
Example 1. Find the Z-transform of the following sequences:
(i) f(k) = {15, 10, 7, 4, 1, − 1, 0, 3, 6 } (ii) f(k) = {5, 6, 1, 2, – 1, 0, 8, 4, 3}
A
(iii) f(k) =
RS 1 UV (iv) f(k) =
RS 1 UV , – 3 ≤ k ≤ 3
T4 W
k
T2 Wk

(v) f(k) =
RS5 ,
k
k<0
.
UV
T2 ,
k
k≥0 W
172 A TEXTBOOK OF ENGINEERING MATHEMATICS

Sol. (i) The symbol ↑ is used to denote the term in zeroth position i.e., when k = 0. k is
an index of position of a term in a sequence.
1 1 3 6
Z{f(k)} = 15z3 + 10z2 + 7z + 4 + − 2 + 0 + 4 + 5
z z z z
1 1 3 6
⇒ F(z) = 15z3 + 10z2 + 7z + 4 + − 2 + 4 + 5
z z z z
(ii) In case the symbol ↑ is not given, extreme left term is considered as zeroth term
corresponding to k = 0. Here, the zeroth term is 5.
6 1 2 1 8 4 3
∴ Z{f(k)} = 5 + + 2 + 3 − 4 + 0 + 6 + 7 + 8
z z z z z z z
6 1 2 1 8 4 3
⇒ F(z) = 5 + + 2 + 3 − 4 + 6 + 7 + 8
z z z z z z z
∞ ∞
1
(iii) Z{f(k)} =
k=−∞
∑
f (k) z − k =
k=−∞
4 k
z−k ∑
1 1 1
= ...... + 64z3 + 16z2 + 4z + 1 + + 2
+ + ......
4 z 16 z 64 z 3
3
(iv) Z{f(k)} = ∑ f (k) z
−3
−k
(since – 3 ≤ k ≤ 3)

3
1 1 1 1
= ∑2
−3
k
z − k = 8z3 + 4z2 + 2z + 1 + +
2z 4 z 2
+ 3
8z
−1 ∞
(v) Z{f(k)} =
k=−∞
∑ 5k z−k + ∑
k=0
2k z− k

= [...... + 5–3z3 + 5–2z2 + 5–1z] + 1 +

LM 2
+
4
+
8
+ ......
OP
N z z2 z3 Q
5 −1 z 1 z z
= −1
+ = + ; | z | < 5, | z | > 2.
1 − 5 z 1 − (2 / z) 5 − z z − 2

Example 2. Find the Z-transform of

1 1
(i) f(k) = {a| k |} (ii) f(k) = ,k≥1 (iii) f(k) = ,k≥0
k k(k + 1)

(iv) f(k) = cos

kπ
,k≥0 (v) f(k) =
0, k > 0
.
UV
2 1, k ≤ 0 W
∞
Sol. (i) Z{f(k)} = ∑
k=−∞
f (k) z − k

−1 ∞ F aI
∞

+ az) + ∑ GH z JK
k

=
k=−∞
∑ a −k
z −k
+ ∑
k=0
k
a z −k
= (...... + a3 z 3 + a 2z 2
k=0

az a FG IJ −1
az z ∵ (1 − x) −1 =
∞

∑ xk
=
1 − az
+ 1−
z H K = +
1 − az z − a
.
k=0
INTEGRAL TRANSFORMS 173

F 1I
ZG J=∑
∞
1 −k
(ii)
H kK k=1
k
z

1 1 1 1 1 FG IJ
= + 2 + 3 + ...... = – log 1 −
z 2z 3z z
if
z
<1 H K
zFG IJ
= log
z−1 H K if | z | > 1.

RS 1 UV = Z FG 1 − 1 IJ = Z FG 1IJ − Z FG 1 IJ
T k(k + 1) W H k k + 1K H kK H k + 1K
(iii) Z

F z IJ − ∑ 1 z = log F z I − FG 1 + 1 + 1
= log G
∞
IJ
H z − 1K k+1 k=0
GH z − 1JK H 2z 3z
−k
2
+ ......
K
F z IJ − zR|S 1 + 1 FG 1IJ + 1 FG 1IJ + ......U|V
= log G
2 3

H z − 1K |T z 2 H z K 3 H z K W|
F z I R F I 1 U
H z − 1JK − z ST− log GH 1 − z JK VW
= log G

F z IJ − z log FG z IJ = (1 – z) log FG z IJ .
= log G
H z − 1K H z − 1K H z − 1K
F kπ IJ = ∑ cos kπ z
Z G cos
∞
−k
(iv)
H 2K k=0
2

1 1 FG 1 IJ −1
z2
= 1−
z2
+
z4
− ...... ∞ = 1 +
H z2 K = 2
z +1
if | z | > 1.

∞ ∞
1
(v) Z{f(k)} = ∑
k=−∞
f (k) z − k = ∑
k=0
zk =
1− z
; if | z | < 1.

Example 3. Find the Z-transform of

R| ak U| 1
(i) f(k) = S k ! , k≥0
V (ii) f(k) = , k ≥ 0.
|T0, otherwise|W
(k + 1) !
∞ ∞
ak −k (az −1 ) k
∑ ∑
−1
Sol. (i) Z{f(k)} = z = = e az = ea/z.
k=0
k! k=0
k!

(ii)
L 1 OP = ∑ 1 z = 1 + 1 z + 1 z + ......
ZM
∞
–k –1 –2
N (k + 1) ! Q (k + 1) !
k=0 2! 3!
L 1 z + 1 z + ......OP
= z Mz + −1 −2 −3

N 2! 3! Q
L
= z M1 + z +
1
z +
1 O
z + ...... − 1P
−1 −2 −3

N 2 ! 3 ! Q
−1
= z (ez – 1) = z(e1/z –1).
174 A TEXTBOOK OF ENGINEERING MATHEMATICS

Example 4. Find the Z-transform of

(i) u(k – 1) (ii) 4 k δ(k – 1) ; k ≥ 0
(iii) δ(k – n) ; k ≥ 0 (iv) { nCk} ; 0 ≤ k ≤ n.

∑
∞
1 . z−k =
1 1 1
+ 2 + 3 + ...... =
1 1 1 FG IJ
Sol. (i) Z{f(k)} =
k=1
z z z z
1 + + 2 + ......
z z H K
1 1 1
= if <1
z FG 1 − 1IJ z
H zK
1
= if | z | > 1.
z−1
∞
4
(ii) Z{f(k)} = ∑
k=0
4k δ(k – 1) z–k =
z
.

∞
1
(iii) Z{f(k)} = ∑
k=0
δ(k – n) z–k =
zn
, n is (+)ve integer.

n
(iv) Z{f(k)} = ∑
k=0
nC
k z–k = 1 + nC1 z–1 + nC2 z–2 + ...... + nCn z–n = (1 + z–1)n.

Example 5. Determine the Z-transform of the sequence given by

R| U|
|| 21, k
k<0
||
f(k) = S G J
F I k
k = 0, 2, 4, ...... V
|| H 2 K
,
||
|T FGH 13 IJK
k
, k = 1, 3, 5,...... |
W
What is the region of convergence for the Z-transform F(z)?
∞
Sol. F(z) = ∑
k=−∞
f (k) z–k

∑ FGH 2 IJK ∑ FGH 3 IJK

−1 ∞ k ∞ k
1 1
=
k=−∞
∑ 2k z–k +
k=0
z–k +
k=0
z–k

(k-even) (k-odd)

∑2
∞
−m m
∞
F 1I
+ ∑G J
2p
−2 p F 1I
∞
+ ∑ G J
2q + 1
z − (2 q + 1)
=
m=1
z
H 2K
p=0
z
H 3K
q=0

where m = − k,
k k−1
p= ,q=
2 2
= F1(z) + F2(z) + F3(z)

z/2 z2 z/3
= + +
1 − z / 2 z2 − 1 1
z2 −
4 9
INTEGRAL TRANSFORMS 175

Region of convergence for F1(z): | z | < 2

1
Region of convergence for F2(z): | z | >
2
1
Region of convergence for F3(z): | z | >
3
1
Hence, Region of convergence for F(z): < | z | < 2.
2

ASSIGNMENT

1. Determine the Z-transform of the following sequences :

(i) f(k) = {2, 4, 5, 7, 0, 1, 2} (ii) f(k) = {3, 1, 2, 5, 7, 0, 1}
A A
(iii) f(k) = {0, 0, 1, 2, 5, 4, 0, 1} (iv) f(k) = {1, 2, 5, 4, 0, 1}
(v) f(k) = δ(k + n).
2. Find the Z-transform of
0, R| for k < 0 U|
1, for 0 ≤ k ≤ 5
f(k) = 2, S| for 6 ≤ k ≤ 10 . V|
3, T for k > 10 W
3. Find the Z-transform of
R| F 1I k

(i) u(k – 4) (ii) δ(k – 5)

G J
(iii) S H 3 K
k≥0
.
|T(− 2) k
k≤−1
4. Determine Z-transform for the sequences given below :
f(0) = 1 ; f(1) = 4.7 ; f(2) = 0 ; f(3) = 0
f(4) = 0.75 ; f(5) = 2 ; f(k) = 0 ; k ≥ 6.
5. Express the signals shown below in terms of unit impulse functions and hence find the Z-trans-
form.
(i) f(k) (ii) f(k)

2
2 1 2
1
0.5 1

–1 3
k k
–2 –1 0 1 2 –2 0 1 2

–1 – 0.5

Answers
7 1 2 7 1
1. (i) 2z2 + 4z + 5 + + 3 + 4 (ii) 3z3 + z2 + 2z + 5 + + 3
z z z z z
(iii) z–2 + 2z–3 + 5z–4 + 4z–5 + z–7 (iv) 1 + 2z–1 + 5z–2 + 4z–3 + z–5 (v) zn
2. 1 + z + z + z + z + z + 2(z + z + z + ...... + z ) + 3(z + z–12 + ......)
–1 –2 –3 –4 –5 –6 –7 –8 –10 –11
176 A TEXTBOOK OF ENGINEERING MATHEMATICS

z −4 z
−
z
3. (i) ;|z|>1 (ii) z–5 (iii) 1
1− z −1
z− z+2
3
4. F(z) = 1 + 4.7 z–1 + 0.75 z–4 + 2 z–5
5. (i) f(k) = 0.5 δ(k + 1) + 2δ(k) + 2δ(k – 1) + δ(k – 2); F(z) = 0.5z + 2 + 2z–1 + z–2
(ii) f(k) = δ(k + 2) – δ(k + 1) + 2δ(k) + δ(k – 2) – 0.5 δ(k – 3); F(z) = z2 – z + 2 + z–2 – 0.5 z–3.

2.24 PROPERTIES OF Z-TRANSFORMS

2.24.1. Linearity Property

Z{a f(k) ± bg(k)} = a Z{f(k)} ± b Z {g(k)}

∞
Proof. Z{a f(k) ± b g(k)} = ∑
k=−∞
{ a f (k) ± b g (k)} z − k | by definition

∞ ∞
=a ∑
k=−∞
f (k) z − k ± b ∑
k=−∞
g (k) z − k = a Z{f(k)} ± b Z{g(k)}.

Remark. If Z{f(k)} = F(z) and Z{g(k)} = G(z), then Z–1 [a F(z) ± b G(z)] = a Z–1 {F(z)} ± b Z–1 {G(z)}
where a and b are constants and Z–1 is inverse Z-transform operator.

2.24.2. Change of Scale Property

Z{|ak f(k)|} = F
FG z IJ
If Z{f(k)} = F(z), then
H aK
∞
Proof. F(z) = Z{f(k)} =
k=−∞
∑ f (k) z −k

z
Substituting for z, we get
a
zFG IJ ∞
FG z IJ −k ∞
F
a H K ∑
=
k=−∞
f (k)
H aK = ∑a
k=−∞
k
f (k) z − k = Z{|ak f(k)|}.

z/a z
Remark (i) Z{ak U(k)} = = if | z | > | a |
z
−1 z−a
a
z z
(ii) Z{ak} = since Z(1) = .
z−a z−1
2.24.3. Multiplication by kn

If Z{f(k)} = F(z), then Z{kn f(k)} = − z FG d IJ n

F(z)
H dz K
∞
Proof. Z{k f(k)} = ∑ k f (k) z
k=−∞
−k
INTEGRAL TRANSFORMS 177

∞ ∞
d −k d
=−z
k=−∞
∑ f (k) (− kz −k−1
) =−z ∑ f ( k)
dz
(z ) = − z
dz
F( z)
k=−∞

In general, Z{|kn f(k)|} = − z

FG d IJ n
F( z ) .
H dz K
Remark. Z{(k ak)} = − z
d
[Z(a k )] = − z
d z
=
az FG .
IJ
dz dz z − a H
( z − a) 2 K
z
Note. Since Z(1) = , we have the following results.
z−1
z z( z + 1)
(i) Z(k) = (ii) Z(k2) =
( z − 1)2 ( z − 1)3
z3 + 4 z 2 + z z4 + 11z3 + 11z2 + z
(iii) Z(k3) = 4 (iv) Z(k4) = .
( z − 1) ( z − 1)5

2.24.4. Shifting Property

If Z{f(k)} = F(z), then Z{f(k ± n)} = z±n F(z)

∞ ∞

Proof. Z{f(k ± n)} = ∑

k=−∞
f (k ± n) z − k = z ± n ∑ f (k ± n) z
k=−∞
− ( k ± n)

∞
= z±n ∑ f (r) z
k=−∞
−r
(r = k ± n)

= z±n F(z).
Corollary 1. For casual sequence,
Z{f(k – n)} = z–n F(z)
Also, Z{f(k + 1)} = z F(z) – z f(0)
and Z{f(k + 2)} = z2 F(z) – z2f(0) – z f(1)
RS
f (1)
= z 2 F( z) – f (0) –
and so on.
UV
T
z W
Corollary 2. Z–1[z–n F(z)] = f(k – n) = Z–1[F(z)]k → k – n.

2.24.5. Division by k

If Z{f(k)} = F(z), then Z

RS f(k) UV = −
TkW z z
z −1 F(z) dz

Proof.
R f (k) UV = ∑
ZS
∞
f ( k) − k
z
TkW k=−∞
k

= ∑
∞

k=−∞
f ( k)
FG 1 z IJ = − ∑
Hk K
−k
∞

k=−∞
f (k) z z
z − k − 1 dz

=− z z ∞

∑ f (k) z
k=−∞
−k−1
dz
178 A TEXTBOOK OF ENGINEERING MATHEMATICS

F I
=− z z
z –1
H
∞
GG ∑ f (k) z
k=−∞
−k
JJ dz = –
K z z
z −1 F( z) dz .

2.24.6. Initial Value Theorem

If Z{f(k)} = F(z), k ≥ 0, then f(0) = Lt F(z)

z→∞

∞
Proof. Z{f(k)} = ∑ f (k) z
k=0
−k

⇒ F(z) = f(0) + f(1) z–1 + f(2)z–2 + ......

Taking limit as z → ∞, we get
f(0) = Lt F( z) .
z→∞

2.24.7. Final Value Theorem

If Z{f(k)} = F(z) ; k ≥ 0, then Lt f(k) = Lt (z − 1) F(z)

k→∞ z→1
∞
Proof. Z{f(k + 1) – f(k)} = ∑
k=0
{ f (k + 1) − f (k)} z − k

n
z F(z) – f(0) – F(z) = Lt
n→∞
∑
k=0
{ f (k + 1) − f (k)} z − k

n
Lt ( z − 1)F( z) = f (0) + Lt . Lt
z→1 z→1 n→∞
∑ { f (k + 1) − f (k)} z
k=0
−k

n
= f (0) + Lt
n→∞
∑
k=0
z→1
Lt { f (k + 1) − f (k)} z − k

| Changing the order of limits

LM n O
= Lt f (0) + ∑ { f (k + 1) − f (k)}P
n→∞ MN k=0
PQ
= Lim f (n + 1) = Lim f (n) = Lim f ( k) .
n→∞ n→∞ k→∞

2.24.8. Differentiation Property

Let Z[{f(k)}] = F(z). An infinite series can be differentiated term by term within its region of
convergence. F(z) may be treated as a function of z–1.
∞ ∞
F(z) = ∑
k=0
f(k) z–k = ∑ f(k) (z–1)k
k=0
Differentiating on both sides w.r.t. z–1
∞
d
dz −1
F(z) =
k=0
∑
kf(k) (z–1)k–1 ...(1)

z–1
dF( z)
dz −1
= ∑
k=0
kf(k) z–k = Z{kf(k)}
INTEGRAL TRANSFORMS 179

dF( z)
∴ Z{kf(k)} = z–1 ...(2)
dz −1
Differentiating (1) w.r.t. z–1 again, we get
∞
d 2 F( z)
d( z −1 ) 2
= ∑
k=0
k(k – 1) f(k) (z–1)k – 2

2 ∞
d F( z)
z–2
d( z −1 ) 2
= ∑
k=0
k(k – 1) f(k) z–k = Z{k(k – 1) f(k)}

d 2 F( z)
∴ Z[k(k – 1) f(k)] = z–2 ...(3)
d( z −1 ) 2

2.24.9. Convolution Theorem

The convolution of two sequences {f(n)} and {g(n)} is defined as
∞
w(n) =
k=−∞
∑ f(k) g(n – k) = f ∗ g

If it is one sided (right) sequence, take f(k) = 0, g(k) = 0 for k < 0, then
∞
w(n) = ∑ f (k ) g(n – k) = f * g.
k=0
Statement. If w(n) is the convolution of two sequences f(n) and g(n), then

Z{w(n)} = W(z) = Z{f(n)} . Z{g(n)} = F(z) G(z)

LM ∞ OP LM ∞ ∞ OP
MN ∑ f ( k) g ( n − k)
PQ ∑ MN ∑ f ( k) g ( n − k)
Proof. W(z) = Z{w(n)} = Z = –n
k=0 n=0 k=0
PQ z
∞ L ∞ O
= ∑ f(k) M ∑ g (n − k) z P
−n

k=0
MN n=0
PQ
(by changing the order of summation)
∞ L ∞ OP
= ∑ f(k) M ∑ g ( p) z
− ( p+ k)

k=0
MN p=0
PQ , (putting n – k = p)

L ∞ OL ∞ O
= M ∑ f (k) z P M ∑ g ( p) z P = F(z) G(z).
−k −p
MN k=0
PQ MN p=0
PQ
Note. This result will be true only for the values of z inside the region of convergence.

2.24.10. Another form of Convolution Theorem

If Z{f(t)} = F(z), Z{g(t)} = G(z), then the convolution product is
∞
w(t) = ∑
k=0
f(kT) g(nT – kT) = f * g

and Z{w(t)} = W(z) = Z{f(t)} Z{g(t)} = F(z) G(z).

180 A TEXTBOOK OF ENGINEERING MATHEMATICS

Proof. (Here we are dealing with one sided Z-transform only)

∞ ∞
F(z) = ∑
m=0
f(mT) z–m ; G(z) = ∑ g(nT) z–n
n=0

R| ∞ U| R| g(nT) z
∞ U|
F(z) G(z) = S ∑ f (mT) z
|T
−m
V| S| ∑ V|−n

m=0 WT n=0 W
∞ ∞ F
∞ n I
= ∑ ∑ f(mT) g(nT) z–m–n = ∑ GG ∑ f ( pT) g{(n − p)T}JJ z –n
n=0 m=0 H
n=0 p=0 K
R| U
S| ∑ f ( pT) g{(n − p)T}|V| z
∞
=Z –n = Z{f ∗ g}.
T p=0 W
2.24.11. Time Reversal Property
If F(z) is the Z-transform of f(k), then Z{f (– k)} = F(z–1)

2.24.12. Correlation Property

If Z{f1(k)} = F1(z) and Z{f2(k)} = F2(z), then Rf
1 f2
(z) = F1(z). F2(z–1)

and cross-correlation sequence rf1 f2 (l) = Z–1[ Rf1 f2 (z)].

2.25 SOME IMPORTANT Z-TRANSFORM RESULTS

S. No. {f(k)}, k ≥ 0 F(z) = Z {f(k)}

1. δ(k) 1
z
2. U(k) or 1
z−1
z
3. k ,|z|>1
( z − 1)2

FG − z d IJ n
4. kn f(k)
H dz K F(z)

kn
FG − z d IJ n
z
5.
H dz K z−1
,|z|>1

z
6. ak or ak U(k) ;|z|>|a|
z−a

1 FG z IJ , | z | > 1
7.
k
log
H z − 1K
8.
1
z log G
F z IJ
k+1 H z − 1K
INTEGRAL TRANSFORMS 181

1
9. e1/z
k!

10. δ(k – n) z–n

11. f(0) Lt F(z)

z→∞

12. Lt f(k) Lt (z – 1) F(z)

k→∞ z→1

∞
13. rf1 f2 (l) =
k=−∞
∑ f1(k) f2(k – l) R f1 f2 (z) = F1(z) F2(z–1)

14. h(k) = f(k) ∗ g(k) F(z) . G(z)

15. f(– k) F(z–1)

EXAMPLES

Example 1. Determine the Z-transform of

f(k) = δ(k + 1) + 3 δ(k) + 6δ(k – 3) – δ(k – 4).
Sol. By linearity property, we have
F(z) = Z{f(k)} = Z{δ(k + 1)} + 3 Z{δ(k)} + 6 Z{δ(k – 3)} – Z{δ(k – 4)}
= z + 3 + 6z–3 – z–4.
Example 2. Find the Z-transform of

(i) {ak+3} (ii) {k2} (iii)

RS (k + 1) (k + 2) UV
T 2 W
(iv) {abk} ; a ≠ 0, b ≠ 0 (v) {k (k – 1)}.
∞
Sol. (i) Z{f(k)} =
k=−∞
∑a k+3
z − k = a3 Z(ak) …(1)

z
We know that, Z(1) =
z−1
z/a z
∴ Z(ak) = = (By change of scale property)
( z/a) − 1 z − a
za3
∴ From (1), Z{f(k)} = .
z−a

(ii) FG
Z(k2) = − z
d IJ 2
Z(1) = − zFG d IJ LMFG − z d IJ FG z IJ OP
H dz K H dz K NH dz K H z − 1K Q
=− z
d LM
z
=
z( z + 1) .OP
N
dz ( z − 1) 2
( z − 1) 3 Q
(iii) Z
RS (k + 1)(k + 2) UV = 1 Z(k 2
+ 3k + 2)
T 2 W 2
182 A TEXTBOOK OF ENGINEERING MATHEMATICS

1 1 z( z + 1)
+3 −z
d
Z(1) +
LM
2z RS UV OP
[Z(k2) + 3 Z(k) + Z(2)] =
=
2 2 ( z − 1) 3
dz N
z−1 T W Q
=
LM
1 z( z + 1)
− 3z
d z
+
2z
=
FG
1 z( z + 1) IJ
+
3z
+
2z OP LM OP
N
2 ( z − 1) 3
dz z − 1 z−1 H
2 ( z − 1) 3
K( z – 1) 2
z−1
.
Q N Q
(iv) Z(abk) = a Z(bk) = a
FG z IJ
H z − bK if | z | > | b |.

z ( z + 1) z 2z
(v) Z{k(k – 1)} = Z(k2 – k) = Z(k2) – Z(k) = 3
− 2
= .
( z − 1) ( z − 1) ( z − 1) 3
Example 3. Find the Z-transform of f(k) = u(– k).
z
Sol. We know that Z{u(k)} =
z−1
By time reversal property, we have
z −1 1
Z{u(– k)} = −1 = ; | z | < 1.
z −1 1− z
Example 4. Find the Z-transform of
(i) {k + nCn} or {k + nCk} ; k ≥ 0 (ii) {k + nCn ak} ; k ≥ 0.
Sol. (i) Z{k+nCn} = Z{k+nCk}
∞
= ∑
k=0
k+nC
k z–k = 1 + n+1C1 z–1 + n+2C
2 z–2 + ......

(n + 2)(n + 1) –2
= 1 + (n + 1) z–1 + z + ......
2!
(− n − 1)(− n − 2)
= 1 + (– n – 1)(– z–1) + (– z–1)2 + ......
2!
F z IJ
=G
n+1

= (1 – z–1)–n–1 = (1 – z–1)–(n+1)
H z − 1K .

F z / a IJ
=G
n+ 1

(ii) Z{k+nC n ak }
H z / a − 1K | Change of scale property using (i)

F z IJ .
=G
n +1

H z − aK
Example 5. Find the Z-transform of
(i) sin αk, k ≥ 0 (U.P.T.U. 2009, 2015) (ii) sin (3k + 5), k ≥ 0.
∞ ∞
1
Sol. (i) Z(sin αk) = ∑
k=0
sin αk z–k =
2i ∑
k=0
(eiαk – e–iαk) z–k

1 LM ∞ ∞ OP
MN ∑ (e ∑ (e
iα
= z –1 ) k − − iα
z –1 ) k
2i
k=0 k=0
PQ
1
= [(1 – eiα z–1)– 1 – (1 – e–iα z–1)–1]
2i
INTEGRAL TRANSFORMS 183

=
1 LM 1 −
1 OP = 1 LM z −
z OP
2i N1 − e z iα −1
1− e − iα
z −1
Q 2i N z − e iα
z−e − iα
Q
=
1 LM z(e iα
− e − iα ) OP = z sin α
.
2i N z − z(e
2 iα
+e − iα
) + 1Q z 2
− 2 z cos α + 1
∞ ∞
1
(ii) Z{sin (3k + 5)} = ∑ sin (3k + 5) z–k =
2i
∑
k=0
{ei(3k + 5) – e–i(3k + 5)} z–k
k=0

1 LM ∞ ∞ OP
=
2i MN ∑
k=0
e5i (e 3i z –1 ) k − ∑
k=0
e −5i (e −3i z –1 ) k
PQ
1 5i
= [e (1 – e3i z–1)–1 – e–5i (1 – e–3i z–1)–1]
2i
e 5iF 1 I F
e −5i 1 I
= GH
2i 1 − e z JK
3 i −1
–
2i
GH 1 − e −3 i z −1
JK
e F z I
5i
e F z −5i
I=1 LM e 5i
z( z − e −3i ) − e −5i z( z − e3i ) OP
= G J
2i H z − e K
–
2i GH z − e 3i −3i JK 2i N 3i
(z − e ) (z − e −3i
) Q
1 L (e − e ) z − z(e − e
5i −5i 2 2i − 2i
)O 2
= M 1 − z (e + e ) + z
2i N 3i −3i 2 PQ = zz 2
sin 5 − z sin 2
− 2 z cos 3 + 1
; | z | > 1.

Example 6. Find the Z-transform of

(i) ck cosh (α k), k ≥ 0 (ii) ck cos (α k), k ≥ 0
FG kπ + αIJ , k ≥ 0.
(iii) cosh
H2 K
∞
Fe αk
+ e − αk Iz 1 LM ∞ ∞ OP
Sol. (i) Z{cosh (αk)} = ∑ GH
k=0
2 JK –k =
2 MN ∑ (e
k=0
α
z −1 ) k + ∑ (e
k=0
−α
z −1 ) k
PQ
=
1
(1 − e α z −1 ) −1 + (1 − e − α z −1 ) −1 =
1 z
+
z LM OP
2 2 z−e α
z − e− α N Q
z
. 2=
LM
2 z − (eα + e − α )
= 2
z ( z − cosh α) OP
α
2 z − z(e + e ) + 1 N
−α
z − 2 z cosh α + 1 Q
By change of scale property,
z z FG IJ
Z{ck cosh (αk)} =
c c H
− cosh α
K =
z ( z − c cosh α)
.
FG z IJ − 2 FG z IJ cosh α + 1
2
z − 2cz cosh α + c 2
2

H cK H cK
∞
F e + e I z = 1 LM (e
i αk − iαk ∞ ∞ OP
2 M∑ ∑ (e
Z{cos (αk)} = ∑ G
H 2 JK
iα
(ii) –k z −1 ) k + − iα
z −1 ) k
k=0 N k=0 k=0
PQ
=
1
(1 − e iα z −1 ) −1 + (1 − e − iα z −1 ) −1 =
1 z
+
z LM OP
2 2 z−e iα
z − e − iα N Q
184 A TEXTBOOK OF ENGINEERING MATHEMATICS

zF 2 z − 2 cos α z ( z − cos α) I
= GH2
2 z − 2 z cos α + 1
= 2
z − 2 z cos α + 1
JK
By change of scale property,
FG z IJ FG z − cos αIJ
k
Z{c cos (αk)} =
H cK H c K = z (z − c cos α)
FG z IJ − 2 FG z IJ cos α + 1 z − 2cz cos α + c
2 2 2

H cK H cK
R F I U ∞
F I R| FGH ∞ kπ IJ
K
FG kπ + αIJ U
H 2 K |V z–k
∑ e
1 +α −
kπ
T H 2 JK VW = ∑ cosh GH 2 + αJK z = 2 S|T
kπ
(iii) Z Scosh G
2
+α –k +e
k=0 k=0 |W
1 L O
∞ ∞
= M
2 MN
e ∑ (e
α π/2
z ) + e ∑ (e
–1 k
z ) P
−α

PQ
−π /2 –1 k

0 0
1
= [eα (1 – eπ/2 z–1)–1 + e–α (1 – e–π/2 z–1)–1]
2

LM F
1 α z F z I OP = z LM e (z − e ) + e (z − e ) OP
I α −π/2 −α π/2

MN GH JK
GH z − e JK PQ 2 N (z − e ) (z − e ) Q
= e . + e–α
2 z−e π/2 −π /2 π/2 −π /2

L F π IO
z M z cosh α − cosh G − αJ P
=
L
z 2 z cosh α − cosh (α − π/2)
M O
P =
N H 2 KQ .
2 N z − (2 cosh π/2) z + 1 Q
2
z − 2 z cosh π/2 + 1 2

Example 7. Find the Z-transform of Scos G

R F kπ + αIJ UV ; k ≥ 0.
T H 8 KW
R F kπ + αIJ UV = ∑ cos FG kπ + αIJ z = ∑ FG cos kπ cos α − sin kπ sin αIJ z
Sol. Z Scos G
∞ ∞

T H 8 KW H8 K
k=0 H 8
–k
8k=0 K
–k

F kπ IJ – sin α Z FG sin kπ IJ
= cos α Z G cos
H 8K H 8K
R| U F I
z ( z − cos π / 8) | G z sin π / 8 JJ
= cos α . S V – sin α . G
|T z − 2z cos 8 + 1|W
π 2
GH z − 2 z cos 8 + 1JK
π 2

| From Ex. 5 and 6

FG z
π 2 π IJ
z 2 cos α − z cos
π FG IJ
=
H8
− z cos
cos α − z sin sin α
8
=
K 8
−α
.
H K
π π
z 2 − 2 z cos + 1 z 2 − 2 z cos + 1
8 8
z(z − cos aT)
Example 8. If F(z) = 2 , find f(0).
z − 2z cos aT + 1
Sol. From initial value theorem, f(0) = Lt F(z)
z→∞
INTEGRAL TRANSFORMS 185

z( z − cos aT)
∴ f(0) = Lt 2 = 1 by L’ Hospital rule.
z→∞ z − 2 z cos aT + 1
z
Example 9. If F(z) = , find Lt f(t).
z − e−T t→∞

Sol. From final value theorem, we have

Lt f(t) = Lt (z – 1)F(z) = Lt (z – 1) z
= 0.
t→∞ z→1 z→1 z − e −T

Example 10. Find Z–1

RS 1 UV given Z RS z UV = (– 1) . –1 k

T z + 1W T z + 1W
Sol. Z S
R 1 UV = Z RSz . z UV
–1 –1 −1

T z + 1W T z + 1W
=Z S
R z UV = R|SFG − 1IJ U|V k

|TH K |W
–1 = {(– 1) }, k = 1, 2, 3, ...... k–1
T z + 1W k→ k − 1 k→ k− 1

Example 11. Find Z G

F 3 IJ .
H 3z − 1K
–1

R| 1 U| LM R UOP
Sol. Z S
R–1
3 U
V =Z –1 = Z Mz S
| z |VP –1 −1
T 3z − 1W S| z − V| 1 M | z − |P
1

T 3W NM T 3 WQP
=Z |
R z U| = FG 1IJ or F 1I u(k – 1). k− 1 k− 1
–1
S| z − 1 V| H 3K GH 3 JK
T 3W k→ k − 1
Example 12. Using differentiation property, find the Z-transform of
(i) k ak u(k) (ii) k(k – 1) ak u (k).
d FG z IJ
Sol. (i) Z{kak u(k)} = z–1
dz −1 H z − aK
d a az −1
= z–1 −1 (1 – az–1)–1 = z–1 . =
dz (1 − az ) −1 2
(1 − az −1 ) 2
d2 2 a 2 z −2
(ii) Z {k(k – 1) ak u(k)} = z–2 (1 – az–1)–1 = .
d( z −1 ) 2 (1 − az −1 ) 3
z −1 2 z −2
Note. If a = 1, then Z{k u(k)} = −1 2 and Z{k(k – 1) u(k)} = .
(1 − z ) (1 − z −1)3
Example 13. Find the Z-transform of f ∗ g where
(i) f(n) = u(n), g(n) = 2n u(n)
(ii) f(n) = 3n u(n), g(n) = 4n u(n) using convolution theorem.
∞
z
Sol. (i) F(z) = Z{u(n)} = ∑
n=0
1 . z–n =
z−1
if | z | > 1
186 A TEXTBOOK OF ENGINEERING MATHEMATICS

∞
z
G(z) = Z{2n u(n)} = ∑
n=0
2n z–n =
z−2
if | z | > | 2 |

By convolution theorem
Z{f ∗ g} = Z{w(n)} = W(z) = F(z). G(z)
z z z2
= . = if | z | > | 2 |.
z−1 z−2 ( z − 1)( z − 2)
z
(ii) F(z) = Z{3nu(n)} = if | z | > | 3 |
z−3
z
G(z) = Z{4nu(n)} = if | z | > | 4 |
z−4
By convolution theorem
Z{f ∗ g} = Z{w(n)} = W(z) = F(z) . G(z)
z z z2
= . = if | z | > | 4 |.
z−3 z−4 ( z − 3)( z − 4)
Example 14. Compute the convolution f(k) of the two sequences :

f1(k) = {4, – 2, 1} and f2(k) =

RS1, 0≤ k≤ 5 UV
.
T0, otherwise W
Sol. F1(z) = Z{f1(k)} = 4 – 2z–1 + z–2
F2(z) = Z{f2(k)} = 1 + z–1 + z–2 + z–3 + z–4 + z–5
∴ F(z) = F1(z) . F2(z) = 4 + 2z–1 + 3z–2 + 3z–3 + 3z–4 + 3z–5 – z–6 + z–7
Taking inverse Z-transform, we get
f(k) = {4, 2, 3, 3, 3, 3, − 1, 1} .
A
Example 15. Determine the cross-correlation sequence rf1 f 2 (l) of the sequences :
f1(k) = {1, 2, 3, 4 } and f2(k) = {4, 3, 2, 1}.
Sol. Cross-correlation sequence can be obtained using the correlation property of
Z-transform.
F1(z) = 1 + 2z–1 + 3z–2 + 4z–3
F2(z) = 4 + 3z–1 + 2z–2 + z–3
⇒ F2(z–1) = 4 + 3z + 2z2 + z3
R f1 f2 (z) = F1(z) . F2(z–1) = (1 + 2z–1 + 3z–2 + 4z–3) (4 + 3z + 2z2 + z3)
= (z3 + 4z2 + 10z + 20 + 25z–1 + 24z–2 + 16z–3)
∴ rf1 f2 (l) = Z–1{ R f f (z)} = {1, 4, 10, 20, 25, 24, 16} .
1 2
A
2
2z + 5z + 14
Example 16. If F(z) = , evaluate f(2) and f(3).
(z − 1) 4

Sol. F(z) =
1
.
LM 2 + 5z + 14 z OP
−1 −2

z 2
N (1 − z ) Q
−1 4

By initial value theorem, f(0) = Lt . F(z) = 0

z→∞
INTEGRAL TRANSFORMS 187

Similarly, f(1) = Lt z{F(z) – f(0)} = 0

z→∞

f(2) = Lt z2{F(z) – f(0) – z–1f(1)} = 2 – 0 – 0 = 2

z→∞

f(3) = Lt z3{F(z) – f(0) – z–1f(1) – z–2 f(2)} = Lt z3{F(z) – 2z–2}

z→∞ z→∞

LM 2z + 5z + 14 − 2 OP
2
= Lt z3 .
z→∞ N (z − 1) z Q
4 2

R 13z + 2z + 8 z − 2 UV
3 2
= Lt z .S
3
z→∞ T z (z − 1) W = 13.
2 4

ASSIGNMENT

1. Find the Z-transform (one sided) of the following sequences {f(k)} where f(k) is

(i) FG 1IJ k
u(k) (ii) (cos θ + i sin θ)k (iii) (– 1)k u(k)
H 4K
kπ kπ FG 1IJ k
(iv) 3k sin
2
(v) 2k cos
2
(vi) 4k +
H 2K + u(k – 3).

2. Find the Z-transform of {f(k)} where f(k) = k 2k

3. Show that
z( z − cosh θ) z sinh θ
(i) Z(cosh kθ) = (ii) Z(sinh k θ) = .
z2 − 2 z cosh θ + 1 z2 − 2 z cosh θ + 1
4. Show that
ze a ( ze a − cos θ) ze a sin θ
(i) Z(e–ak cos kθ) = (ii) Z(e –ak sin kθ) = .
z 2 e 2 a − 2 ze a cos θ + 1 z 2 e 2 a − 2 ze a cos θ + 1
z( z + 1) z2 ( z + 1)
5. Using Z(n2) = 3 , show that Z(n + 1)2 = .
( z − 1) ( z − 1)3
2 z2 + 3 z + 4
6. Given Z(un) = , | z | > 3, show that u1 = 2, u2 = 21, u3 = 139.
( z − 3)3
z ( z3 + z) cos θ − 2 z2
7. Using Z(k) = , show that Z(k cos kθ) = .
( z − 1)2 ( z2 − 2 z cos θ + 1)2
8. Find the convolutions of
kπ kπ
(i) k(k – 1) ∗ 3k (ii) 3k ∗ cos kθ (iii) cos ∗ sin .
2 2
d
9. Prove that Z{kn} = – z [Z{kn–1}].
dz
∞ ∞
10. Evaluate the Z-transform of the sequence {f(k)} = ∑
k=0
2k ∑
k=0
3k.

Answers
4z z z 3z
1. (i) (ii) (iii) (iv)
4z − 1 z−e iθ z+1 2
z +9
188 A TEXTBOOK OF ENGINEERING MATHEMATICS

z2 z 2z 1
(v) 2 (vi) + + 2
z +4 z − 4 2 z − 1 z ( z − 1)
2z 2 z2 z2 ( z − cos θ)
2. 8. (i) (ii)
( z − 2)2 ( z − 1)3 ( z − 3) ( z − 3)( z2 − 2 z cos θ + 1)
z3 z2
(iii) 10. .
( z + 1)2
2
( z − 2)( z − 3)

2.26 INVERSE Z-TRANSFORM

Inverse Z-transform is a process for determining the sequence which generates given
Z-transform. If F(z) is the Z-transform of the sequence {f(k)}, then {f(k)} is called the inverse
Z-transform of F(z). The operator for inverse Z-transform is Z–1.
If Z{f(k)} = F(z), then Z–1 [F(z)] = {f(k)}.

2.27 METHODS OF FINDING INVERSE Z-TRANSFORMS

We have the following methods of finding inverse Z-transforms:

(1) Convolution method (2) Long division method
(3) Partial fractional method (4) Residue method (or Inverse Integral method)
(5) Power series method.
2.27.1. Convolution Method
We know that Z{f ∗ g} = F(z) G(z)
k
∴ Z–1 {F(z) G(z)} = f ∗ g = ∑ f (m) g(k − m) .
m=0

EXAMPLES

Example 1. Find Z–1

RS z UV using convolution theorem.
2

T (z − a)(z − b) W
–1
Sol. We know that Z {F(z) G(z)} = f ∗ g

Let F(z) =
z R z UV = a
∴ f(k) = Z {F(z)} = Z S –1 –1 k
(z − a) Tz − a W
G(z) =
z
∴ g(k) = Z {G(z)} = Z S
R z UV = b
–1 –1 k
z−b Tz − bW
Z–1{F(z) G(z)} = f ∗ g = ak ∗ bk

∑ FGH ab IJK
k k m
= ∑
m=0
ambk–m = bk
m=0
(a G.P.)

R| FG a IJ − 1U|
k+1

= bk
|S H b K |V = a k+ 1
− b k+ 1
.
|| b − 1 ||
a a−b
T W
INTEGRAL TRANSFORMS 189

R| z2 U|
Example 2. Using Convolution theorem, evaluate Z–1 (z − 1)(z − 3) . S| V|
T W
Sol. We know that
Z–1 {F(z) . G(z)} = f ∗ g
z
Let F(z) = ∴ f(k) = (1)k
z−1
z
G(z) = ∴ g(k) = (3)k
z−3
Now, Z–1 {F(z) . G(z)} = (1)k ∗ (3)k

∑ FGH 31IJK
k k m
= ∑
m=0
1m 3k–m = 3k
m=0
(a G.P.)

R| FG 1IJ − 1U|
k+ 1

= 3k
|S H 3 K |V = (1) k+ 1
− (3) k+ 1
=
RS
1 k+ 1
(3 − 1) .
UV
|| 3 − 1 ||
1 1− 3 2 T W
T W
2.27.2. Long Division Method
∞
Since Z-transform is defined by the series F(z) = ∑
k =0
f(k) z–k (one sided), to find the inverse

Z-transform i.e., Z–1 [F(z)], expand F(z) in the proper power series and collect the coefficient of
z–k to get f(k).

EXAMPLES

Example 1. Find inverse Z-transform of

10z 2(z 3 − z)
(i) 2 (ii) .
z − 3z + 2 (z 2 + 1) 2
10 z 10 z −1
Sol. (i) F(z) = 2 =
z − 3z + 2 1 − 3 z −1 + 2 z −2
By actual division,
10z–1 + 30z–2 + 70z–3 + ......
1 – 3z–1 + 2z–2 ) 10z–1
10z–1 – 30z–2 + 20z–3
30z–2 – 20z–3
30z–2 – 90z–3 + 60z–4
70z–3 – 60z–4
70z–3 – 210z–4 + 140z–5
+ 150z–4 – 140z–5
∴ F(z) = 10z–1 + 30z–2 + 70z–3 + ......
190 A TEXTBOOK OF ENGINEERING MATHEMATICS

Now comparing the quotient with

∞

∑
k=0
f(k) z–k = f(0) + f(1)z–1 + f(2) z–2 + ......

We get the sequence f(k) as f(0) = 0, f(1) = 10, f(2) = 30, f(3) = 70, ......
i.e., we can get f(k) = 10(2k – 1), k = 0, 1, 2, ......

2 z( z 2 − 1) 2 z −1 − 2 z − 3
(ii) F(z) = =
( z 2 + 1) 2 1 + 2 z − 2 + z −4
By actual division, we get F(z) = 2z–1 – 6z–3 + 10z–5 – 14z–7 + ......
∞
Comparing the quotient with ∑
k=0
f(k) z–k = f(0) + f(1)z–1 + f(2)z–2 + ......

We get f(0) = 0, f(1) = 2, f(2) = 0, f(3) = – 6, f(4) = 0, f(5) = 10, f(6) = 0, ......
kπ
In general f(k) = 2k sin
, k = 0, 1, 2, ......
2
Example 2. Find the inverse Z-transform of
4z
F(z) =
z−a
for (i) | z | > | a | (ii) | z | < | a |.
Sol. (i) For | z | > | a |, we have

4z 4z FG 1 − a IJ −1

∑ FGH az IJK
∞ k
a
z−a
=
z H zK =4
k=0
if
z
<1

∞
= ∑
k=0
4ak z–k, where | z | > | a |

Z–1
FG 4z IJ = {4a }. k
H z − aK
(ii) For | z | < | a |, we have
4z 4z FG 1 − z IJ −1

z−a
=–
a H aK
4z F z z 2
z3 I
=–
a H G 1+ +
a a 2
+
a 3 JK
+ ...... , if |z|<|a|

4z 4 z2 4 z3 4 z4
=– – 2 − 3 − 4 – .....
a a a a

Z–1
FG 4z IJ = {f(k)}
H z − aK
where
RS
{f(k)} = ......−
4
,
−4 −4 −4
, , .
UV
T a 4
a3 a2 a W
INTEGRAL TRANSFORMS 191

Example 3. Find the inverse Z-transform of

1
F(z) = for
(z − 3)(z − 2)
(i) | z | < 2 (ii) 2 < | z | < 3 (iii) | z | > 3.
1 1 1
Sol. F(z) = = −
( z − 3)( z − 2) z−3 z−2
(i) For | z | < 2.
1 z FG 1 IJ
z
−1
FG IJ −1
F(z) = –
3
1−
3 H +
2
1−
2K H K
1 1
= – (1 + 3–1z + 3–2z2 + 3–3z3 + ...) + (1 + 2–1z + 2–2z2 + 2–3z3 + ...)
3 2
= – (3–1 + 3–2z + 3–3z2 + ...) + (2–1 + 2–2z + 2–3z2 + ...)
Here coeff. of z–k (if k > 0) = 0
1 1
Coeff. of z–k (if k < 0) = – − k + 1 + − k+ 1 = – 3k–1 + 2k–1
3 2
Now, Z–1{F(z)} = {f(k)} = {– 3k–1} + {2k–1}.
(ii) For 2 < | z | < 3.
FG IJ – 1 FG 1 − 2 IJ
1 z
−1 −1
F(z) = –
H K z H zK
3
1−
3
1 F z z z I 1F 2 22 3 2
23 I
=– G
3 H
1+ +
3 3
+
3
+ ......J – G 1 + +
K zH z z2 3 2
+
z3
JK
+ ......

= – (3–1 + 3–2z + 3–3z2 + 3–4z3 + ......) – (z–1 + 2z–2 + 22z–3 + ......)

Here, coeff. of z–k (if k > 0) = – 2k–1
Coeff. of z–k (if k ≤ 0) = – 3k–1
R|− 2 k−1
, k>0 U|
Now, Z–1{F(z)} = {f(k)} = S|− 3 k ≤ 0 |W
V
T k−1
,
(iii) For | z | > 3.
1 FG 3 1IJ −1
2 FG IJ −1
F(z) =
z
1−
H z
–
z K1−
z H K
1 1
= (1 + 3z–1 + 32z–2 + ......) – (1 + 2z–1 + 22z–2 + ......)
z z

∴ Z–1{F(z)} = {f(k)} =
3 k− 1 RS
−2 k− 1
, k≥1
.
0,T k≤0

2.27.3. Partial Fractional Method

Here we split the given F(z) into partial fractions whose inverse transforms can be
written directly.
Example. Find the inverse Z-transform of
z z 3 − 20z 8z 2
(i) 2 (ii) (iii) .
z + 7z + 10 (z − 2) 3 (z − 4) (2z − 1)(4z − 1)
192 A TEXTBOOK OF ENGINEERING MATHEMATICS

F( z) 1 1 A B
Sol. (i) = 2 = = +
z z + 7 z + 10 ( z + 2)( z + 5) z+2 z+5
1 1
− .
1 1 1 1 FG IJ
=
3 z+2 3 z+5
∵ A = ,B= −
3 3 H K
1 z 1 z
∴ F(z) = −
3 z+2 3 z+5

∴ f(k) = Z–1{F(z)} =
1 –1
Z
RS z UV − 1 Z RS z UV −1
3 T z + 2W 3 T z + 5W
=
RS 1 (– 2) k –
1
(– 5)k
UV LM∵ Z(a k ) =
z OP
T3 3 W N z−a Q
z 3 − 20 z F( z) z 2 − 20
(ii) F(z) = or =
( z − 2) 3 ( z − 4) z ( z − 2) 3 ( z − 4)
F( z) z 2 − 20 A + Bz + Cz 2 D
Now = 3 = 3
+
z ( z − 2) ( z − 4) ( z − 2) z−4
1 1
⇒ D=– , A = 6, B = 0, C =
2 2
1 2 1 FG IJ
∴
F( z)
=
6+
2
z
+
−
2 H K
3
z ( z − 2) z − 4

or F(z) =
RS 12z + z − z UV = 1 RS z(z − 2) + 4 z + 8z − z UV
1 3 2 2

T (z − 2) z − 4 W 2 T (z − 2)
2 3
z − 4W 3

1 R z 2z + 4 z z U
2
= S
2 Tz − 2
+ 2.
( z − 2)
− V
z − 4W 3

1 LM∵ Z R az + a z U = k a OP 2 2
Now –1
f(k) = Z {F(z)} = {2 + 2k 2 – 4 },
2
k 2 k
MN ST ( z − a) VW
k
PQ
−1
3
2 k

= {2k–1 + 2k . k2 – 22k–1}
2
8 z2 z
(iii) F(z) = =
(2 z − 1)(4 z − 1) 1 FG
1 IJ FG IJ
z−
2
z−
4H KH K
F( z) z A B 2 1
= = + = −
z FG z − 1IJ FG z − 1IJ z−
1
z−
1
z−
1
z−
1
H 2K H 4K 2 4 2 4
2z z
∴ F(z) = −
z − (1 / 2) z − (1 / 4 )
R| FG 1IJ − FG 1IJ U| , k = 0, 1, 2, ......
k k
f(k) = Z–1{F(z)} = 2 S| H 2 K H 4 K V|
T W