0% found this document useful (0 votes)
29 views30 pagesFourier Analysis of Signals and Systems
Uploaded by
sekharjyothika2001Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
0% found this document useful (0 votes)
29 views30 pagesFourier Analysis of Signals and Systems
Uploaded by
sekharjyothika2001Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
/ 30
GRAMI Gt, TH ae
PH 233 - ADVANCED ELECTRONICS
NN ON LCA
Unt i
deg D Fourier Analysis of Signals and Systems
oA” aon i , :
ue \ : \
es ee metrxte Fourier Series .
A’ pertiodeas Pamabtan) pL) Bansal eotpruncd wm the
form of Axigemoneduic Series as
SOD = Bart a count 4 a, cos aunts. + by Siounts
bz Sin 2ugt +-.- 3G
where ou = IF = on |
T '
F a the Arequency and @i and bi are the ~coclpicten’s-
The fousien series ead When the puncte for satipics
the following Ati tmditiors called Diichleli emdiliars.
DH w ewobll defined and singles valued exceept patsib
at @ finile number fF pant 4 Ft) has a earl
. average Value over the poviod T.
W ft muut poste only a fineke number of dustmotinushes
ap Ahe period JT.
W ft mut have, a finde yumobext of positive and
vuigakive roaxima & he period T.
© may be expressed: by the fousen seater
Sit) = 5 a + = Gn Comt + = bye somat] 0
“et k. ’ ne)
utes Gy andl by aie icoefpiciestl Balévalaaled -
Integrating OQ ever othe full Pee , h , ts
feo ae ame eo Bimilarly , mulliplymg both sider be fo rugt and
ae
a f" “a
ae an an ol sin must dt = ta a el a Gn. COTE Syma
2. é b 1 %, >
&@ { feb de ee ccede 9
Ne
Mattiplymg both sides ¢ equation @ ba coe eed 0d
and whgrating a fp & Se mut dt 0
aA ae a if 7 % af *
ue Abe Ge Mead | (te i
of ae hes ae. ik Qn const vio mast dt = ie [ain manreats Snr ora
2 os
(Gen nust+bn toning | ot
,™ ~S : 7 ~ Oy Mae, Tate
‘ _ =o.
cate 1
[a coruge de <0 4 fb eipnuat siomugt dt = is [em enemunt -
2, 5 are -% co @o-must Jar
Prd, ai
Th -{Ebn + rt
J Fe aa mupt de = red sok > eit
7h o% ae , sien
flew - “ ay) Gas) a - Tb, for man
sionus,t co muah aoe =:
a Tg Son iv rai, eb _ploowone Te
Tha Brinn) 1) ae Ferd fin mut dt ke E
= fF afte er+muat + cos ops {
oe ae ope
ae = by [sia @oampegt «= $i
4c
: non
Penk nee Siva, ee
y, 2
aw an, = ie fer) Ce mune
9 “ee. 7}
an.
i fer) 2 nate The « caine 3c ana
expiciiel os te ae
6
—
A a
when fd & ever ff sey ae - 2 frat
Bras 2 |
The sum ox product oF him OX mixe cues function
an even function
WM Fey S add when fetye - fey
$F fe 4 odd | pent <0
Es
Te sunt of tun on mone odd functons c am edd
duct of tun odd furcetens
functor and the
& an even funetim
ti) 4¢
cE4%) = Feb), oly even harm
3F Fte 4%)
Qte PAtsenk
+ SOD, only od. harmerta are puck
|
2. Exponeotial Form ct. founien Seaies
rm tb founder \serien ts,
The dnigonometsie
as) < tay + (Gn Gowen aby eon)
ble have,
ces runt.
sin nut
~ ee
ae (a) en om oy
ea re a
Take G = 2 (fbn)
and €
56-4
+ (@ ij bn)
Substituting the values oF a and by Sam @ ard ©
t
5 Go [4 [Reo co mun at ope [es 4 nat de
%e ‘
a i ca
& = = J Feb (ct nust =f sto nut) dt
Ve.
sinust
fen é alt
%.
cee f fet) (Co neat +f Konugt) dé
Pian ans ae
| Gee Neti scr ne i
-%
there the, value of yen ontncgaloven 12 Viz eit festa
and aut ocluded, unde the summatin (Z) tigre
Aico Co may (bel trtlided under the 2 Sige by
using no *
a] Tree
4 fee es Cael
Aly: cleat thane) tha thd! piniedi, Reaktne #2
may be eaprened yathematically by an mafintle
cet of patel pate and negative Suequuncy
Uroponents
A poritaue Frequency may be associated will 0°
anki- clockwise relation ame a negative, Frequency
O° ting these values, we get
UJ de = E&
Ths 4 Parevals edentely ~
The quantily om Ris i the average value ot [fet] * ner
one peried a
Paxsevald theese stabs that the
tune ak fie riedal? quem lof Ieeecel each
& equal to: the Pua valle “Vet [Fes auellane
3. Fansevals idendily foy
oY Portion Serio ager
Ag finn Ot cepmcdertcd by pe Le ac
fouriar sates gy, the Fouletct coefpictoots 20 the functions that the
2 duce oti it é : a
eA (sig ratecte: ben sasbover af douche. Ercotially it oo toraenvatin lose er
a th ntoxe general, farm ik tak, thot
2A ZG $0 coonugt + Sd tml) ioe
( i [res] gate [ele (25 ee
t=
On ectegeatting a bky -% and 4) TAF as
% oA u pe?
+). (hee? avy Se f fee de ge a faa em nugt de) on\ stl
th, FS, Goat Wwe eth
ae h Powen Spectrum cf @ periodic Punctim!
+ Bal fey > nllaeeel Se ThE
2 fe) Sin nut The power, 4 @. periodic femnetias—ro the
signal Ft) in the me domain iu defined as, ®
Te
i Pers
3H
tence for a periodic funckin , the power » a
|
Kime wavelorm fet) and can be evaluated by |
adduag togelhis the powers centuinedd i cach
hatmenid The powed of Ike AI “harmonic component |
ar muy 4 GI ame that of -nt o (Cop |
fora single ral hatmo ni ,
{Gl + Ie_p}>
|
; j |
|
|
The vs values of Ake feet fern,
feo = 5G + 0, towb + a cor ub 4
by tnwl + be sin 2une 4
can be wyilttp as
Frwut
> Fourier Trans form
Te plot of “amnplitinds at debfeeeot trequency
components — fer @ periedic cuaue G kmewn ar
‘db cxele frequency, spectrernr’, beeawte the amplitude
values Hgmificaoce only at
lisence valour
of Ue, tec SF Ine vepetitien pened
T wercases, Ud
T tWlhes He repetiten puiod T+. | the
wave
fet) will bécave’” neo perio
meant the separatun
haimonie components call be
Theteforxe the ‘descacte spictaum cull
f a Ceptinewers speck’ telties Ah» ee Ihz
adyaceot pulses Ve ee Ee
pulse Ayato kedusees ts 4 ogc ne
peluees tue adjacent
Zeko, G, Utye0 =
been
9 wu Fee) 4 ipittually percoclée ne
wo expercectial Jor
2 inunt
fy + 2 Ge er)
one
, %% *conbinetcus nano)
“hihin evaluated Ihe quanstily Fo bracket’ <4
a functiw gF frequency ently and © demoted
a Fu)
D hineaniby
a a and b axe arbittary constant
so ‘
Faw) + [to eM gt] @
<0.
4h called (he “fourier ansform’ of fee)
uo, fet) « we gry) eet
Tin fe) 4 is Be a
° ae wig
pays tee f Fda O° due tT ay
“0 =
00 just
fb - f FGun ¢ dé |—®
He :
cx dnansforme
44 calleel the “imuedse fours
© ord ® emitilule a Famer Transherm pate:
for every Line demain suaveform “there U a terrespo-
ndring fouqueocy demain wave foxne and vice versa
fer. example a sructangulan pulse co the Lme
duroain —coimcideo ill vine. funckinm cm the
Arequin cy clervain » Muakily provides Ahat the
neverse also fue. a xeclengerlar pulte in
Vie frequency clomain ratthes cil a cing
funettns te Blend “damnelin 1 tlaive funcbesy
comapends te eagh cllter tn) Ihee
called taanshom pais -
that
manren Que
ash
tu
ih
Propenteco ot Fourien Truanitorn
4 gen RGU _
and fac => Faius) Khun
afi + bike) > 4 AGu + b BGue) >
w Sy
1
Pacot > 20 an
Xie Just
1 d
ee ee
an FEO = (" e(jus) aye
uhssx the dummy variable OG Aeplaced by ad,
ae laced bi ’
sreplaced by us.
ar foot) = f Fab) eI) ged
00
Finally ud & cuplacd by 't -lo obtain mark
Hetrgnisable tee foyrn , ;
ees
afew = fegy eae.
amicus) - PR;D) ae Jedaer eee
oF FGt).
# kD > FGus)
treo F(t) > 21 Fe-us)
WD Scaling
a ky S FGue
hin fat) = at F (ky)
Be riccibeliey »
Peon
Sh aco . ther Ure Fouiex uansform at Feat) x,
Pap 2 f fan aM ae
: foaggee :
Put. xeak ; da = adt
FE (fead] =F [feo].
Sean meen
”
- F ,
pe in pues” dx
saps Tey)
gaia *,
; by
Sr ao. then Ffegey] = FOapy
Corwbining Uae fur pesealts., coe gee
ca eee
Lange the durratim of dhe Aime funektin, smaller
a kt © bandwidllc of tts “spectoutm by hz same
sealing factor. Covetely , sroallee he ducati of
the kine funetioy , Lengex a the bandusdlle of
ds Ygusy
2
JO bade = YGus), = Xtusy, HGua
F Eee] = ¥Guey”
oe - fue OW ae
Le
iG | eo RE ar) at
ao
‘e e —juck
a zcof f hove 2] dc.
: » 9a
ie “ic ei
» Sicquenceps ceovolaskon +
Put ae t-t
then, b= 45t and dt-da
yoo = ff “acof f i “nte) ee
© dc [nc ~
a ao &
Go) =
Hence “the Fousien Exansformn cf [hz cenveluttin of tuo
signals G the Lime demain & equal fe lz
mull plicatun ct this individual fousucx tsansforms|
to frequency demain. gt rneans that 4 we convelute|
ue signals in kine demain, equivalent to
saulliplysog thes specte & facquency deroein.
X Gus - ce
OF
and. gcth > h(ju)
then, fet) get) > 3 jue” GGfas) =
fp > PGjus) :
Paceuk ¢ ——_ ~
The invelue Gransipglo cil hecene jee] Se
5 [FE] ta) frt* cae
But, 3 iz
Fu)” jee) al PCa) Gu far fs) dia
aa
ist lmaett
pact at a 1|28 >
ue = tu
dup = dx.
e a a co) ey *
ae) Can) feo | ois. 2°
a
a = 2, j (a+
=) Frau f aco 6 "2a
ae
ee ay
1: i
= (ar) Jr oan fogadag
4 am
» 2 Pet 1 gee)
1 iid 1
Feo. ge = Uw) Gicqeo)
tence the roultiplicatiin “d+ Ewe signals in Ihe
time domain 4 equal fo the Cmvoluttts of
dividuala fourier transforms to Ihe Arequen
; a)
demain. This preperty) & called ‘noultiplicnten theaueet
Shibteng
ome -shibting delay): -
dF ft) & EGus
| P
[theo Fb Fun ee
Crook
Tee fouster txamsfoxmn ot fet-b) Go givtn by
2 .
F (sa) - f fore oh
2 5m
Pur ne x
Ba Rot Eateaeand™ nea
eceasey) f a pone) a
oo jun —juse. 5
eee eons dx -@
<0 e406
Just
= Fiju).é
“juste
Fer-t) ey Fijw) €
udheo Ake function fet) G delayed byl te’ the
fouquiocy spectrum ante multiplied by
erivete see thee tr mo change in Ihe amplitude
spectrum and “each frequency emopenent tr shifted
we phase bY an ameunt -arh! |
SF ft <> jus) “theo
fee et LY PG uetfate)
Cxoufs \
By debiniten, ——
fs 2 fage it 4
ftp el a fee ee Pal
f a
«ftp, pieeren?
é leh a
=F (jus jute) > Pruguenay diffetenc ation
SS
i) Fjur) tuo
\(JO feo d Fgw)
aut
Sotageation
Time _solaguation
fer) a Fijus) thio
@ dee MF Fro) se)
= Fu th
id)
mao
e f Fijus) des!
» Teme reversal
4t RD & Gus) thio
fc- <= FCfus)
Pix _ cerjugation
4 Fee)
$e)
= FOur) Iho
}e PC)
> Dually
pt Fed) e& giju) Eheo
Git) ean Fc Jur)
Pauevali thyeoreen
[O dpemiae ae fh Tequesta
f
2
Thu relate ts calteel Paxsevals theonem oY
Rayleigh energy thepxenr.
usok: ‘
E of [feo|? dt oui fen fty at ef
= ff Fen (& fou Ms)
2 ee
-£f fore Fon) caer
to de
ok Ef aa
2 f cae eqs
SJ Fue Fiquy
Ea
cal [Fejus>|* dees
2 i ©
eee J [feel” de > ell [Fejeot? dus .
~~
7. Fouxiex Traniform of seme imposetant function
© Gate _femidbinn ’
Rentider a tage’ Function (rectangular pulse) a
shaun tm agua -
See east atin = (tore je (elt y
=f) od (=
Ht has Ake analytic expacrion
Fy = i ae
©» olkeuerse
Tre fourcer transform of -feey wv given by
Fan = Fe sey = f fery okt ge
40
*
ill welt de
he
rh
mee
ee
= 1 x -a sin flv)
ee £
’ . & Sino %)
= 2% wee)
oT
+7 singuom)
©%)
T . sine (ey)
Fu) =
Fourtien transform of gale duoctin,
Ua) = T. sincQwomy
thou lke roplitade spectiun
T [snc wn)
cpeebuun & ,
LE (juss |
gives by ae.
sin < @U%) <0
The peak value cf amplitude a one
SD eet pH =Oul 23,
= EE EM, 1, BM,
Aa cet alc by F radeans
polarity Beem
ur
‘Amplchite spect
Cio adele
Svequeney) =
&
Recangulas Pulse
Genider @ sectanguilax pelie shown do figure. The
analyre expression
few a
» octet
» clhensise
The fouxier txansform
a E
FGu - F (Feo) - ff fen eb de
P
aA Covey
= 7. 8% sen)
Hy
R lee
Fou) + > Sine GUT)
eurier tanto cf ~ rectangular pulse.
panes
ou)
-2
Ferd
of fet becemer .
aa
Just
= Afb eeatie de:
°
fot
= icin
Sie
ot
. pene
a EN
i
on ie
Jas
ae
jae
ac
we
et jum
Th
7 sin (ws7, )
. Sine (UT)
. — Aj Sh@syy 7D
:
Th jute 5
et” gan)
Ga: |
heer,
The aroplitude spectuum & exactly the same ay
Mat of galt’ functtw. But there Yan addileona)
phase shifE fackn glum which changes Ihe
phase speclatuume « Gy weg the time shelt Ween
& [ite =a, = eejuny) PPE tee
Gus) |= T Frc Gur) sIhe abe Yesult can readily
be cbtained «
Fu)
ae
20
4a :
@ Exponential Pulie
Amiider an exponential pulse shown fo fogee - : s ¢ ‘
The founien trarsfoxi -of expecotial pulse cu,
2 ajust
fom) < f RY eae
“0
te as
” fe f ete ae
<2 es
2 Ca 4jur
. f e dt ;
es ariut
Anoplilnde cpeetnaim
= @rjus)
oma
arr)
ae envider ad Ulex pulse Shown to 5:
ajar ; parades pul “Shou 7 Figc
Se Aton)
G@auncaju
a-jup
Oe ust
fone (2108 (ota) is
fate
Fagus)
e
FiC-w0) © Bins)
eq? ot Line FR & given by ,
aus h
le Pt) OE ERT
iw) [ee eee ae
dc) - “tad (uy)
|Fouer | =
fet) = oe +h)
Fourier transform ot exponmtal pulie cy
pe 2 AC PEHD fr heteo
FGw) « oe
ep ct Line FP oo given by,
s ie :
The armptitide and ‘phaue spectrum of Iho ; Jip 28a
exponential pulse ane, > ae F Qu) =
Thus Ane analytical expression Soy Ihe driang ie Cae Sls
2A Tite!
= 2h (owe =sin0)) = 9a (ee siolotH
(ee
pule ur,
Were -74 BA tiowm ~2AQ sin (0H) —
transdurm of daiangulat pule G . nen
: end i
The fourtey
“ 44
F (ft) - Fw) - [ foe dt
f + AAL 2d (OA)
ae -
. i (i+ 2Eyede +f oa
-% me
ea
; Te
aa fe nile ae eae ; ;
a) 2
Fousion. ETRE hang pile ey,
AT ene" Cap)
ort
gee
Vo they eo tm
T% shut : : te rr ihe
: 2 ee et pear
» ages eM ar -
“a P(e ee ae
T
art
> aA ey =
jar * ur %
(a
je Gwr
cpu )%
> 24 2 a
[= eee) es
tect
4|
we
BAe
T i
The pulae Aefined by
Gaunian pulse a
The fouier transform of
E (6 aed? - (kine
%
Dea yale
i ng sat dt
“0
2 ew)
Fuu- J é dt
Let att 4 jue Vet + lwo y
*
Fu) ir exp (@ bp jun
ere
dt
AF Ceres
“2 a
(Tt + Why
af é eter ge
<2 s
em 2
Way £4 IW, Ge)
; Faw = e e ee
“0
es te sm F + jus
. air
and de = fr. dt
©
a Fw) « ca edu
r
-%
called)
Gained pulie a
Wik 00
© Cte) i ene oes (tier
cc a
Sages
va iH
gts
eye ee
tne. Be. v
a, Ike fourley transform of oa gauuvian pulse
& a gaunlao pulse Uself
26.07. 2029
qusnday
ees,
FroBLeMs
1. check Uhethey @ Fourier sexies
Giloumag . distete © tine ovignall -
a) ©) = 2 sin VB an
6b xen! ele
2 xo = 2 wb G1n
Lat (na) 2 2. BUGIN) = 2
exists | Sov
$5 (Gm 4 Gant)
For pertodietlg V3 nA) should ‘be ao integral
roulkiple of | 291
ket VaaN 2 Mxet ,where M and A are begers
eS ae ei
? Nv ie
Here N cannot be \an. intages for , so Powe
seen) U papperiodce signal” and! fo) Foca
seria doern’t ear
Jeno
ee j o20+N) ame) jean
kt x(man) = € * = é
Here also for periodicity SUN should be
iotegcal rpultiple of a
ket sai = 20N usher M and N axe integess
oe oN 4M
a
Here Noa inter. fy Mz 510.10).
lib M =5
nee ‘ :
Here 0 us pesidic wilh Sandacnertal no - Stee”
poets :
N-4 and funde tal ch
apes AUD GS, ig 28 “2
sats eli cy wl.
fat eo
% Find Ike fourier dvansforre af xen) cherie By the ae oF FE
ne =f) ozne 6 gud tpi PORE x aa ne
i volucaite F(ju) + xX - Sn = Oye
a X St prt: ata Speedie
X 2
’ Skid aie ease flee bee)
By Ihe definthio ot» feuviey Transform, :
— Jntcarnasordh BE, plaids
FEE RED) = Se. tye ‘ rover, a APSE EI ES Ze. i
a3
eee ee eee ae
ole
fini, —
ae
é
ss
peace’
nad! . BP
Be 6
a3
= 0,933
eqgc = Se >aju)
2 5s ep b.oe7 é
golot] 202
e
Oe , pucukk time fourier Trans fomnn
The pevtodic function any dan be
expressed in a
Fouricn senior
yepreentation, Consider a deicaete
signal x(n)" that periodic wile Period N
debined by Hin) = x (D+KN) for any integer
value of k-
The pevtedic function 20m) can be synthentsed
as Ane sum oF tine! Yand covine sequences
0% equivalently a Linear ctombnatinn of cmple
expontals where frequencies are roullipladl
epultiples cf Ihe fundamental frequency , 20
a
Tid 4 dene by comltucting @ periodic
Sequince for ci each period 4 identital
toch finde Lingth sequences
ponential fern! of DET
The expineotial form of founien series for a
pericdie desnrt Hime Siqnal ir given by
gkuon
a
xm 2-1 5 x(k) 2 ani
) pone aha sz.
where the coefficient x(k) aed the surdameota|
digital frequenay wy, are expressed as,
Jun
A-T
XW = Fam wk
n=0 ae
erect) Us, em
hs za
© and @ are called descecte Foucex dransfomn (DET)
synthets and “analyst pate
Where Xk) ahd xm) are periodic sequencer.
The equivalent forro of @ fy x(k) &,
a Na} ke
BiG enim
pet {xv} - x (N-k)
YD Conjugation
Ber fam} = x(k)
per Sixt} = X(N)
Gawler Srequency shite
Der {zon} x(k
Der {xa eae Gee)
vw)
vi Multiplicctin
per {zap = x (k) io, ae
pet {xm} -+ Ixloo mt w= am
vid Civeular Convelekion a0) 5 eee
HUST. Seftan> Xi, Cue
DET m= XE), 4, 9
DFT %0) = X[k)
DFT fxm® xem} = Xilkd Xalk)
1 Greular correlation
Det fxom} = x(k)
Det tym} = Yk)
Na rai
DET frhey Om } : Beal xm Yeo-my}
= xi) YW
Parse vals relation
DET fom} = XiCko
DAT {amy = 2th)
Nel * re
BE ee eins eb Se xo xk F jxal
2 NT Ro ise ee
ae
Prosi ems - A
4 ;
4. Reropule — civectlax emvolution oF follawing Seq!
wang DET. ;
ao = fool} and ©) = fi, 2,2}
hing sien 23 : idl sie
2 asa
DET fantyt ngeiCO, = a CO ce — Ko.
4 —jemok/,
DPT {mcm} = %W) = Exum e ee
fi BU
x.) oe
AW, ik ark
-1+2e 42 +26
Keo > Xi), = ae 1asalt I) va 6
kero mune 1 ym! +2 =0
ke2 2 Xx@> 1a +3)] 42 --2
keg > %GD= 1 +af-! =25 389
fe aC HNN NG (Oo 22 Ol
yee
pet {xi © %O} - ah) = (KX)
te, MH Xj(09 = Xi) ¥x0) = 2x6 =1/2
x09 = XD X20 = One
5
x@)
X, @ = X1@) © -2x-2 = 4
Xg(D = XQ X,G) = 0x OF <0
2 Xs =| favo, a, of
ciyculax Convolution af x(n) and %aCn) @
Xa(k) ={12,0/4,0}
Find x (k)
vb
for
2m)
{v23.4]
kle haue
XO =41 23,44
New
5 0k
deer
44 oj
an) e@
oa
3-0
x(k) =
2s -2-2j}
defination (Fart 2)
A roathematical tool te
sein and
be both
we Use
DFT
ae
the foustier
A signal — ean
For _ peniocti
and for nan periadiz signal
vepresent” signal
fourier brant form -
periodic and nmpertocts-
fourier sero
fourier 4 ransdorm.
signal
we Wwe be made easier When Note ; a
- ¢
ws oie
ae ee arena og 0S i
mt signal 0 XC X09 5
ee Shiny fourier drars foam , é Ae
* a DFT U Whelely sed in GPS uit. rap ae
a or PROBLEMS .
¢. Perform sane creuler cravelutin of tha fol
| | sequsnce duing DIT - IDET meted
t al el £
Pa yee
aur = fevaniy ord m0 fi.2,3,47
Aare ae oe
iY TEU AAS) aod fits 4} 9 oe
M5 fuss ap ;
Lt iran tie
The divence fouter dnansform | the discrete st
ex gmt te fuer artansfoam apetation.
Ricacte signals may be Like,
+4
pep ithade \ !
Divichit conditions ;
Te conditims FY susknane of fourier trans
2 FCO) mut be abiolubly. intigrable
b Fey must hay
oO Se \
PES seis = 80D TAD
ar fo ve ciff to, 242), -2/-2-25)
a foo, 2-2) 1-2) 242i}
Now, we have !DFT 2apk
mm 1 é€
SS oT
a a i> | xX.)
1 1 1 oy
& 1g ac aa
pte Sy) |i
\
SEMINAR — cc. oF Baap
‘ope ole
Fast Founien Dh eee
Trans form
atic FET U @ method (algorithm)
a DET utth Acduced no, of
Groputatonal
a divide add
fox computing
calculations . The
achieved by adopting
Conquer apptoach. This
effi clemey |
appreach
By Peed on Ars deersporilem oF ay Np
DET into succesively smaller DE Ts. This
wie Ahead,
Can tovernt the ducukk dain 40
(mtinucus data Aype available at
equencter
various
The mopudatems can be of fe types,
(a) Decimation tp time FFT (bit Fer)
(b) Deeimatioa to Trequency FeT (Dir EFT)
The Der of gq sequence 20m of length.
N’ & given by a complex valued sequence
xO | as, ;
ae - Jeo fy
Xo ite are eckeNn-l
hits tattodutce. ai ned vtetm llehy , the coroplen
Valued Sphawe a taetyn ilennsst valent etacs factor’
ov, Multiplication factor’ pick u an ,
ji cect of
wrth ~ieiy
Ea diy? stadia
AE an
Be + Za ae Bly OekeN-
30 PET we exploit
the followimg proper bcs af
tuiddle | fretor it
D periodicity a &
lly 2 ely
t) Syranedruy ie
eae
Radix “y Fem (Vector vadix PET)
do. an N-point | sequiente
2 IFS can be expressed
a RE racy w
ushere
= Varo a 27, then
N! = \7eeeeeoeeetiee Sequence can be ace ia a
YD eo ee
Y- point! DFT can be \ assumed. there the sje of
le OFT. %, U
Caer Aa Ne adie’ of Ihe PET
algortthm -
| ®
1
~ “XW. = wen Why + Foner by
ce rea
fv vadin-2 FFT algorithms . the value of N
sheutd be such that N= 2, Ahat the
N-point Sequence U decimated tno 2-point
gequenues and contequently 2 - point DFTs .
Ung this 2 pint DFT we can calculate Negt
Point DET
Decimation 9o Time Fart Foucier Transhorns iter
Te simppicok and pethapsadhe -best 1 kanlgy
methed i Peru tadix -2 devinaty,
to Kine algarithen
ah N-point
computing
Huot by “decoroputing
time
demain stnal BRN Hime
demas signals each compond ta Single Powe:
th this cae. Lets auume that xem)
seprorents a N Values “here
NG a0 integen power
3. decimated into tio
consisting of the
sequence af
OF 2 .The given sequenen
Né point Sequences
cven- numbed Valier of zm)
and edd numbered valucr of xm)
The point DET sh Stgumee xy uy
ua is given by,
XK) = 2 x0) bly + 0LkeN-
520
pumbered
Break cm) into even and- odd va: Sivoo
nl nk aa
SUE Ta Gat, lvich ts axon By
PO, Benn =O, nedd og
Substituke,
ae
Peartl fey n odd,
an ae) Ne! @rpk
>
ae 2 yrk ae
Xi Co ae + kly Z
= xan (hy) MN
Me
XQE SS
3
Wn =
Butter Th Diagxam .
ae
x Gr (WI)
Basic bultenbly Hn geaph fey ke computation of
DIT FFA / algorithm er
7 =I
eepant DIT FT
dt represents a
ee
x2)
Bea-b hy
Floe gxaph
Xt)
x
xQ)
xX), xo
x00
Xap
=a) al
z@
xo
x0)
xa
xe)
es
x0)
xo)
2@
za) ae
Repu ses
wd - fe IH © cas mt, — f Hom
- Ka -i Xa
2 lane He poor dasen
nk = € é co Mj vom,
3 sara =
yaad = ‘ema - Gs amy 2j sin an,
PRoBLeEMs
“Find” DIT FFT of The ollocumg .
x0) > ces rm. bh .
WD 20)= fo. j-2,3} Neg
“Ao fon2, 3) 495,67}, Whee
DW £0) = 2! ee
HOD = si225 42), ae
~ (0) = Exo 2) :
KU, = CoS 2 Nye
x= (Th 6
AB) = Ct 3% - Yo
nad=0 = xa
Dede xo
xa) o > xem
yay =
cr EXC {re 1-8j 1, 14 vajf
te -
@ xo = forest BN =e cee ies
x@ <0
xv =!
a@y=2 a +e
x (3) 3
10)-0
HQ
XODde)
AOns
wo xt) fe, PO+af p22, -2- a} w
=e 9. eof) A 44) pea! oe], +, - aR
SAAS 49 bei}
258 = xo
XO) “1255, At.64 4166.06) , 814 102)
Bl aca ise
uy ,
$6644 166.06) 00
Pit ies) = Key
2164+ A606 x0)
= x@
6.64 446.067, 6%
Giciente nays coe
y £4) {i 2/34/4981}
xX (kd -{20. 5. 823-2. 414] , 0, -0-1716-0 -414f_ 0.
=O-1716 4 o-414§ 10, -5:82 42-4145 F
b Decimattan, © Frequincy fast Foutier Transform Der|
The deimation im Lime FFT alaosilimn
decermmposes Ahe DET by Sequen fialle, splitting
am) 0 Ike time demain chto— sets cfs mailey
amd sroalter Subequences. DIE FET decmmapaes
The DET by recursively ‘spliftng Ihe sequence
elements XO ty (ke Srequca
