Fourier series:
To represent any periodic signal x(t), Fourier developed an expression called Fourier
series. This is in terms of an infinite sum of sines and cosines or exponentials, Fourier series uses
orthoganality condition
Jean Baptiste Joseph Fourier,a French mathematician and a physicist; was born in
Auxerre, France. He initialized Fonrier series, Fourier transforms and their applications to
problems of heat transfer and vibrations, The Fourier series, Fourier transforms and Fourier’s
Law are named in his honour,
Fourier Series Representation of Continuous Time Periodic Signals
A signal is said to be petiodie if it satisfies the condition x (t) =x (t+ T) or x (n) = (1 +N),
Where T = fundamental time period.
oo= fundamental frequene}
There are two basie periodic signals:
-oscx0t( sinusoidal) &
Zjoor{complex exponential)
These two signals are periodic with period
2ax/on
A set of harmonically related complex exponentials can be represented as { Pi(t)}
u(t) = {eo} = {MF wherek =041,£2..n .....(1)
All these
als ate periodie with period T According to orthogonal signal space approximation of a function x (t) with n, mutually
orthogonal functions is given by
Where «= Fourier coefficient = coefficient of approximation.
This signal x(t) is also periodic with period T.
Equation 2 represents Fourier series representation of periodic signal x(1)
The term
Ois constant.
The term #1 having fimdamental frequency « , is called as 1* harmonies.
The term £2 having fimdamental frequency 29, is called as 2" harmonies, and so on,
The term én having fundamental frequency noo, is called as n" harmonies.
Deriving Fourier Coefficient
We know that
Multiply e"Y on both sides. Then
a(t)eIrwot — Ss azedhot e—Inwot
Consider integral on both sides, T T ©
[ aera = s aged ote ineol dt
0 0
k=—00
T x
-[ s azeik—m)wot de
0 K==00
T i) T
[ a(te™tdt = S> a, [ ellh-n\wot dt... (2)
0 0
wn
by Euler's formula,
T T T
[ le meotdt, = [ cos(k — n)uydt «if sin(k — n)wot dt
0 0 0
T
[[orman {ten
0 0 kn
Hence in equation 2, the integral is zero for all values of k except at k= n, Put k=n in
equation 2
=/ a(t)e I dt = anT
0
1 —jnwot
=> = INwor dt
oT [ “
Replace n by k 1 ft y
sa=7f/ e the0t dt
0
oo.
2.2(t) = 35 ayett-mt
k=—00
1? i
wherea, = = [ e Tkeot dt
Th
Properties of Fourier series:
Linearity Property
if a(t) ees IT, gg yg) Persie cent
then linearity property states that
aa(t) + by(t) PE alte oF kb hn
Time Shifting Property
ita(t) Jourier series coefficient Ion
then time shifting property states that
a(t — ty) Comers ethene, nt f Frequency Shifting Property
5)
to to
x(t) + sin nyt dt
? nayi dt
2 peer
2 dn =e x(t) + cos nwt dt
T Sy
b,,
2 pein |
zh a(t) -sin nut dt Series (E1
Expone Four
‘eiment } (n = 0, +1, +2...)
len. This is a complete set soit is
Consider a set of complex exponential finctions {
which is orthogonal over the interval (f,fo°7). Where
possible to represent any function f(t) as shown below
f(t) = Fo + Fret! + Feu +... + F,eimmot +.
Fe! + F pe Pty, + Fe imaty,
oSQ)= s Fmt (t t
T/2 1/2
Fla] = ATSa()
FT of Impulse Function:
FT w(t)] = [f° 5(e** at]
FT of Unit Step Function:
U(ws) = 760) + 1/50
FT of Exponentials:
FT
e u(t) —> 1/(a + jw)
etu(t) 7 1/(a + ju) FT of Signum Function :
ET
2
sgn(t) <—>
® jo
Conditions for Existence of Fourier Transform:
Any function f(t) can be represented by using Fourier transform only when the function
satisfies Dirichlet’s conditions. ie
‘+The function fit) has finite number of maxima and minima.
+ There must be finite number of discontinuities in the signal f(t),in the given interval of
time.
‘+ Itmust be absolutely integrable in the given interval of time ie.
SL [sOlat < 00
Properties of Fourier Transform:
‘Here are the properties of Fourier Transform:
Linearity Property:
a(t) o> XW
& y(t) <3 ¥(w)
‘Then linearity property states that
aa(t) + by(t) #7 aX(w) + bY(w) ime Shifting Proper
ira EE, x)
Then Time shifting property states that
a(t — ty) o> eH" X(w)
‘Excquency Shifting Property:
If a(t) <> X(w)
Then frequency shifting property states that
+ ay
het a(t) <> X(w — wy)
Time Reversal Propert
Pr
Tf x(t) —> X(w)
‘Then Time reversal property states that
x(-t) <3. x(-w)
ET
If x(t) —> X(w)
‘Then Time sealing property states that
x(at) Tar xe
Differentiation and Integration Properti FT
If x(t) <> X(w)
Then Differentiation property states that
0). Fes, X(w)
at
dzx(t) FT |.
we + (Jw)”. X(w)
and integration property states that
Ja(t)at “2 x(u)
ff. fo(o at = rx)
‘Multiplication and Convolution Properties:
FT
If x(t) <> X(w)
FT
& y(t) —> Y(w)
‘Then multiplication property states that
FT
a(t). y(t) —+ X(w) * ¥(w)
and convolution property states that
a(t) * y(t) “> 2 X(w). ¥(w) Statement of Sampling Theorem:
A band limited signal can be reconstructed exactly if it is sampled at a rate atleast twice
the maximum frequency component in it
‘The following figure shows a signal g(t) that is bandlimited,
Figurel: Spectrum of band limited signal g(t)
‘The maximum frequency component of g(t) is fin, To recover the signal g(t) exactly from its
samples it has to be sampled at a rate f3 > 2fm,
fi is called “Nyquist rate”
‘The minimum required sampling rate f&
Proof:
Let g(t) be a bandlimited signal whose bandwidth is fy, (C= 2)
@
Figure 2: (a) Original signal g(t) (b) Spectrum G(o)
dr(t) is the sampling signal with f, = 1/T > 2fm. a Bw)
OO ee
a ya
@ oy
Fiewe 3; (®) Sampling signal 6r() (b) Spectrum dr(.)
Let gu(t) be the sampled signal. Its Fourier Transform Gs(oo) is given by
Flgs(t)) = Figlt)or(6))
$00
= [ae ys «umn
\ a
= [ou x soma]
Gsle) = YD Glw)+5(e—new)
G.(w) = F[g(t) + 29(t) cos(wot) + 2g(t) cos(Qwot) + ---]
Ge) = & YS Gene)
a0
Figure 4: (a) sampled signal g.(#) (b) Spectrum G(w)
If wy = 2vm, ive. T = 1/2fm- Therefore, G,(w) is given by $00
t YS Gw-nem)
‘To recover the original signal G(w):
1. Filter with a Gate function, Hy,,,(w) of width 2umn.
2. Scale it by T.
Gv) = TG,(~) Hr... (
Figure 5: Recovery of signal by filtering with a filter of width 2u;q
Aliasing:
Aliasing is a phenomenon where the high frequency components of the sampled signal
interfere with each other because of inadequate sampling ©, < Op
Figure 6: Aliasing due to inadequate sampling
Aliasing leads to distortion in recovered signal. This is the reason why sampling
frequency should be atleast twice the banchwidth of the signal, Qversampling:
In practice signal are oversampled, where fs is signi_cantly higher than Nyquist rate to
avoid aliasing
Figure 7: Oversampled signal-avoids aliasing
Problems
1, Find the Fourier transform of the rectangular pulse signal x(9 defined by
lil<@
x()= wo-{4 lion
a)
Sok: By definition of Fourier transform
X(w) =f" pave de= f° ede 1 sinwa sinwa
= a (e™ —e°*) = 2: =2a
jo . wa
Hence we obtain
sinwa . sinwa
pAlt) 2a
. oa
The following figure shows the Fourier transform of the given signal x(t)
Xo)
2a
Figure: Fourier transform of the given signal
2. Find the Fourier transform of the following signal x(t)
x(t)
Sol: Signal x(t) can be rewritten as
xyeema
ean a>0
em 10
aad <0
Then X(w)= f° erteme des ferme dt
- ln
O ta- ju] af fat jaye
= fi errr eer 2a
ate
Hence, we get
ene
a+?
The Fourier transform Xv) of x/i) is shown in the following figures
“0 x)
°
(a) (b)
Fig: (a) Signal x(t) (b) Fourier transform X(w) of x/i)
4. Find the Fourier transform of the periodic impulse train
xo
Tr 0 T M% ‘
Fig: Train of impulses
Sol: Given signal can be written as
8,(t)= YL a(t - kT) the complex exponential Fourier series of 5;,(¢) is given by
1 Qn
-=_— ik yt -_—
m=z 2 7,
de =
F [559] 7 LD 6(w-kwy)
0 kane
=a Le i(w—kwy) = 05, (a)
kane
LE G(t- kT) oy L 6(w key)
bate k
‘Thus, the Fourier transform of a unit impulse train is also a similar impulse train. The following,
figure shows the Fourier transform of a unit impulse train
X(w)
0 wy 2 °
Figure: Fourier transform of the given signal
5.Find the Fourier transform of the signum function
Sol: Signum function is defined as
1 >0
san(t)={_ 120
The signum function, sgn(t), can be expressed as
Sgn()= 2u(t)-1 We know that
a =
ay 88m) = 28(0)
sentt)
Fig: Signum function
Let
san(1) —+X(w)
‘Then applying the differentiation property , we have
joX(w) = F[28(1)] =
2
sen(r) +> —
Note that sgn(t) is an odd function, and therefore its Fourier transform is a pure imaginary
funetion of w Properties of Fou Transform:
Aperiodic signal
Fourier transform
Linearity
‘Time Shifting
Frequency Shifting
Conjugation
‘Time Reversal
‘Time and Frequency
Scaling
Convolution
Multiplication
Integration
Differentiation in
Frequency
Conjugate Symmetry
for Real Signals
Symmetry for Real and
Even Signals
Symmetry for Real and
Odd Signals
Even-Odd Decompo-
sition for Real Sig-
nals
x0
wo
ax(t) + byl)
x(t ~ fo)
ei x(t)
ro
xO)
wat)
x" yO)
xy)
fo
[ x(t
oo)
(0) real
-x(0) real and even
x(t) real and odd
elt) = Er{x(} [x(0) real]
xo(t) = Od{x(0)} [x(0) real]
X(jo)
Yo)
aX( ja) + bY(jw)
eX)
XG(w ~ w»))
X"(- jo)
XC jw)
1 y(2
laa
XGjo)¥jo)_
{ “Xu@H Mo O40
as
jwX( jo)
Xin aXO3(w)
vd
FXG)
X(ja) = X*(-jo)
MX jay} = RelXC— jo)}
Snl{XCja)} = —SrolX(- jo)
|X(Ja)| = [X(— ja)}
EXUjw) = ~XX(~ jo)
X(jo) real and even
X(jo) purely imaginary and odd
RelX(jo)}
iSAXCjeo)} Fourier Transform of Basic Functions:
Fourier series coefficients
Signal Fourier transform it priate)
Sane an S ado - han)
~ 2x5 ~ en)
coswat IB lo ~ 09) + 800 + 0)
sina Z[6(w—oy)— dw too)
F1B(e — on) Btu tool
wah m=O keO
a =1 2n8tw) iss the Fouir series representation for
any choice of T > 0
Periodic square wave
Lost
xo- [a nels
and
2M tan
VT: gine (HOOT) — sin ke Ty
sine 2
L Wi
80)
xU+T) = x0)
Sou-en
xo
sinW
aa)
3u-0)
ul, Rela} > 0
te Wn), Grea} > 0
