You are on page 1of 16
fioung cays fo saB01s anf aul CE = N YM Leled LIG 804 I “mojaq umoys Ayfsoyn eateries. ‘esoduexy my ! ‘= 70 Aq U2aid st poyeadas st aw ym pareroosse gouanbos uauodxo ou} SaUuH Jo Joquinu yp st Yoryas “(-1yxq) 107984 eadox juauodya ou “Ol “soryranng Jo suoNas 40 sjos Jo Jaquinu , _-7 sey oBeys yor “6 (prs) (rT ORT V0 z aks Aq uaaid st pur J, Xoput e8ejs oy} Jo UOKOUTY B oe sjuoUodxo a) JOjoey a[PPIay oy, *g ‘N Bo] N’ Aq want si suonsppe xajduioo jeioy, “L “N Bo] Z/N Aq waatd st uoneordnjnus xe[duioo joy “9 “UO OS Z = / 98BIS PUODOS JOJ pue | =/ dBe}s ySI Joy “97 XopuT adeIs ayy syuasaidas 7 oxoym ‘soydures , _ ,Z Aq payesedas ore Apjioyng yove Joy syndyno pue snduy *¢ “Z/N S} 238}S YoRe Ut SarysaHNG Jo “ON “F “Nn Bo] = 7 st ydess Moy [eUsIs ox ur sode}s Jo Joquinu JeIO} OU “E Jest9Adrtq ysnomp peysnys s} 2ouenbes yndur oy “7 T ‘yoSoqul ue St 7 ‘o19y ‘7 = N So[dutes qndut jo Joquinu ay, +} unpprto3[e Lay LIM Z-XIPes ur paaroaut sdaas Aouaroyya jeuoneindwog ves “f= ont? = {Md “Gy a a= SM MOM, So, of” ering. ~>- — €28e5 7 aseg Lasers “8 = N ‘soulonbes papraoid om wor “uopnjog mynsodpo Leb IG Suasn (YX aunuiasog {1 ‘bb ' Z‘1} = [ule soy womsB 91 17 8's ajdurex sy a "| -/7 Aq Woars st poyeodas st 7 (PIM poyeroosse gouanibas jwouodxe oy} soumH Jo Joquinu ayy st Yoram “(pyq) 0}%y yeodox quauodx> UL ‘Or “satygianing Jo suonoes 10 sjas Jo oquinu , _,z sey e8es yorg “6 (I= ;-7 "ETT O=7 * were =4 Aq UoAts si pue j, Xaput a8ejs oy Jo UoKUNy e axe syuoUodxe MM JoVR a[PpiMy aL *g. “N “B0] WV St suonrppe xa[duro9 Jo Joquinu [eo “L “N Bo] Z/N’ St uoNeoTdnynur xe]dutoo Jo Joquinu [Ro], “9 “xeput BES ay) St /-dxoym ‘sojdures 1=7@ 4q paresedas ore sindjno pue sindur ‘Ayjsoyng yoro 0g *s ‘UN St ots yore ut satysoyng Jo quinn, “p ‘N Bo] = 7 st ydes3 Moy jeusts up UI sodejs Jo Joquinu je10) oY, “¢ “Zopso Jempeu sod se st couonbas yndur UL ‘Z ‘eBoy ue st 7 ‘azoyas “77 = jy ‘sojdures yndut yo Joquinu ay, *] + Uoy1ad 0} aay om YOM Lad AIC wo ynoyom diseq oy []Jo} sdays May SUIMOTIOY OUI, + MNHOSTY Lit ATC Z-x1pey 103 sdayg 6 ‘WHiitioSs¥ ido: i # + Thetotal numberof complex mutip ions req + Also, the otal number of complex additions forcomputi og samo $3: Pata sequence is given as X(K) = (0. 1, 2, 3, 3, 2, 1, 0}. Determine x) using DIT FFT algorithm, olution. From definition, Given that N=8 Hence, = cos n/4 ~ j sin n/4 = 0,707 -/ 0.707 = cos w/2~j sin n/2 2) ~{28)3 W} =e 43) =cos 3/4 —j sin 3n/4 =~ 0.707 -/ 0.707 Now, from the signal flow graph, we have X()=12 x3 X(0)==5.328 -j2414 X@)=0 X@)=0172-j 041s X(5) =-0.172 +j0414 X6)=3) neh yeh 2 ps 1/240 ee 336 EE @ Example 5.4: Provided x[n] = {1, 2, 4,5, 5.4, Solution. It can be seen that N= 8, then 2, I), Determine X(k) using DIT FFT algorithm. We =-0.107-j0.707. iy X(O)=24 Z Xi) = 4232-414 x@=0 Hence, WA =0.107-j0.707, We = Stage 1 Stage 2 Stage3 6 2 X(@)= 0.232 + j 0.586 0.707 Fig. 5.7. Example 5.4. Therefore, AW = (24,4.232-/ 1.414, 0, 0.232 + j0.586, 0, 0.232 0.586, 0,-8.242+/3414)- @ Exampiess: Given, x{n] = 2" and N = 8. Determine X(k) using DIT FFT algorithm: Solution. Given, x{n]=2" and N=8 ; then, *@)=1, x()=2, 3@)=4, a@)=8, x(4)= 16, x(5)=32, x(6)= = picaicies x)= 64, x(7)=128 Hence, cmemeessermnrangy ———— an waa 387 we 1 ' sob Mi =0.707-70.707, 92 = - 707 70.707. Stage 1 Staged X(4)=-85 =30 aa WR, XG) =-78.63 =}46.05 X)=8 NY X(6)=—51 -j 102 Me Weed “1-20 F120 wo 707 XO) = 48.63 + 16605 0.707 Fig. 5.8. Example 5.5. So, A(R) = {255, 48.63 +j 166.05,—51+j 102, - 78.63 + 46.05, = 85, + 78.63 ~j 46.05, -51 —j 102, 48.63 -j 166.05}. © Example 5.6: Nis provided x[n] = (2, 3, 1, 2}. Determine X(k) using DIT FFT algorithm. Solution. Given, an] = {2,31,2}, Hence, N=4. X(0)=8 xay=tsi X@)=-2 X@)=14) a Weed Fig. 5.9. Example 5.6. ivol+ sags-= (X0%M rio! zero-= OX sirol+ zr 0-= (OX EM wip [= see's" (x 0=@) xo o=(x O=()xo'M oz = (0)x eases of s 5 Ga 128%g é °8 = N ‘oouonbes d “aon mlysi408} pepraoid ay wory “woHmleS 1? LAH AIG Bussn (y)y aunusarog {12's pp '¢ 2} — jae oT + gig ardor B s0uanbes F Mewodys ow} sown Jo r>qunu ayy 4) ae a St poyeadas st'7 yim paywig0sse TMSTYM “(hY) soyo"5 ywodax quauodxo oy “OL *solgsoyng jo su 396 ——_—$—_—__________————= [JBiGiTATSIGiAR FRocesaniay From definition, Hence, we find, We =1, .707-j 0.707, We We =-0.707 —j 0.707. Hence, X(K) is obtained as X{&) = (20, ~5.828-j2.414,0,-0.172-j0.414, 0,-0.172-+j 0.414, 0, 5.828 + 2.414). ww Examples.9: Determine the DFT’s of the sequence x(n] -08 na/2, where N'= 4, using DIF, FFT algorithm, Solution. Given, ata] = cos Hence, x{n] = {1,0,-1,0}, N= aks Bean From definition, Wi Ai) Wh =1, Wh ser? 0 X00)=0 x@=1 x()=0 Tx@=0 xq)=2 x 2 Wye i xG=0 0 Wisi tT Xe)=2 Fig. 5.17, Example 5.9. From Fig. 5.17, X(&) is caculated as {0,2, 0,2}. @ Example 5.10 Determine the DFT of the given sequence x[n] = {1, 2, -I, 2, 4, mathematical stage wise explanation. . -1, 2} using DIF FFT algorithm with Solution. Till now we have performed DIF FFT without expanding every stage. If N = 8, then Ws = 1, Wy =0.707-j0.707, We Output of stage-1———_0)= (0) +x(4)= 14 4=5 Al)= x1) +x(5)=2+2=4 A2)= x2) + (6)=- 1-1 §3)= #3) +x(7)=24+2=4 Wg =~ 0.707 -j0.707. yy = % FO~ LO) — 20) 8 may 25 45)= ()—x(5)1 7 =) me =0 49~ WQ)=16)] 92 = 141) 2 < 4D='G)-21 W = 2-2) Wp «0, Similarly, output of 2 stage is calculated as, 800) = [A0) +f2)]=5-2=3 8(1)= (Al) +£3))=4 44-8 8(2)= [A0)-f2)] We =542=7 A . 8G) = [A1)—f3)] Wz = (4-4) G)=0 2 = geen r. 84) = f4)+£0)=—3 8) =f5)+£7)=0 8()= (RA) -£6)] We =—3 8(7)= [A3)—A)] WP =0. 7 Hence, output X(8) is X()= 8@)+ 265) XQ)= g@)+8G)=740=7 XQ)= [a(6) + 8] WE =-3 24) = [(0) —g(1)] Wp =3-8=—5 (5) = [s(4)~ g65)] We = 3 X(6)= [g(2)-g3)] Hf =7 X()= [g(6)- aD] We =—3. 1,-3, 7, -3,-5,-3, 7,3} {l A a @ Example 5.11: Compute the FFT for the sequence x{n} = {1, 0, 0, 0, 0, 0, 0, 0}: Solution. From definition, Hence, Given, Hence, X(k) is obtained as, we (NTU Apr. 0,3) =-0.707-J0.707. 0,707 -j 0.707, 3[n] = {1,0,0,0,0,0,0,0}. XM = (1,111,111, 1}. Fig. 5.20. Example 5.12. T Example 5.13 : tf 50 7 Compite'the 8-point DFT of the sequence x{n] = | Sa using DIT and DIF algorithm. O ; otherwise Soition. DIT Algorithm We =1, We =0.707-70.707, Wy =-j, We =—0.707=f0.707. i 9 a Fig. 5.21. Example 5.13. Einaialn.n.0.0.0), 400 DIF Algorithm 2 7 8 xO=1 \ ie = w,? 2 0 x@=1 t Wy, 7 0 X@=1 Wei 0 x@=t a w;) 0 x@=1 ie a a x@=1 of 0 x@=1 S we e 0 x@=1 = Fig. 5.22. Example 5.13 Ch) = {8, 0,0, 0,0, 0,0, 0; @ Example 5.14: Compute 8-point DFT for a continuous time odd signal x(t) = sin (2nft) where f = 50 Hz Solution. Given, x(0) = sin 2nft) and f=S0 Hz 1 T= — = =0.02sec. then om see. ‘Now, we have to find x(¢) at different values of t. Hence at t=0, x()=0 1= 0.0025 sec, x(i)=0.707 1= 0.005 sec, x(#)=1 1= 0.0075 sec, _x(f)=0.707 1=0.01 sec, x(=0 1= 0.0125 sec, x()= 1=0.015 see, xi 1= 0.0175 sec, x(t) = Hence, we have found, x{n] = {0,0.707, 1,0.707, Also, twiddle factors W,' are calculated as 0.707 0.707. = 0.707, ~1,~0.707}. 1.707 -j 0.707, W2 = We =—0.707-j 0.707. Fig. 5.23. Example 5.14. I 5.5 DETERMINING AN INVERSE DFT BY DIRECT DFT ‘Mathematically, IDFT is determined by ot ainy= 2 xGHW yh, 2= 9412 W 1) 627) 7 N fo Taking complex conjugate of eq.(5.27), we have wet vel 5 ow wA5.28) % is Kats _ Comparing with the expression of DFT, we have 1 29) ' a] = (4) DET [X*(4)] we “ meee the complex conjugate of other sides, we have N=1 y a= 4) Dxt@w™ oe N\ geo (5.30) xin)=(4) ert poy’ 401 X (0) =0 7X (4) =0 X (2)=0 WX @=0 X M)a4 We X@)=0 x(@)=-1 X (3)=0 We 4 x@}=-0007 er 3X M=i6 Fig. 5.23. Example 5.14. I 55 DETERMINING AN INVERSE DFT BY DIRECT DFT Mathematically, IDFT is determined by wel aio] = DY XM y™ 2 = 0, 1,2, WD (5.27) N k=0 Taking complex conjugate of eq.(5.27), we have Nel Next~j= Dx" (5.28) imo Comparing with the expression of DFT, we have 1 (5.29), 2 (2) DFT [x0] Taking the complex conjugate of other sides, we have (5.30), uy nk n= 3] Ewe | : (530) So aE i) (FFT EX"D" 402 x@=0 xq) x x@ x@ x@) x0) x@ Fig. 5.24. Signal flow graph determining an inverse DFT by direct FFT. w@ Example 5.15: Determine the 4-point IDFT of the sequence X(k) = {6, ~ 2+ 2j, ~ 2, ~ 2 — j2}. Solution. From definition Wize Hence, We xO=6 4 < o 48 ()=0 x (I) =-25 a = = * Wa"? x(j=2 - aE a 1 yet Fig. 5.25. Example 5.15. Hence, a{n)= {0,2, 1,3}. gene nenronn ALGORITHM a ee 403 s Example 5.16: Compute IDFT x(n] of given sequence XW) = 136, 4 solution, From definition, $1965, 4 4-4 15 16, = 4,47, W=4=$1.656,=4 4H -4-}9.656) Given, Therefore, wo tee %' =0.707+j0.707, From the process explained in IDFT, we have fol a alge a #5 0.1074 0.70 8 Signal flow gra 0-707 x(-3 ES ‘graph x O=1 (F996 5x (8) =5 Qa IS x (2)=3 x(3)= 41.656 7 ax )=7 xg (1-2 Waxn6 x0)=-45 1656 xQ)=4 x0=-5 10S pee ee a Wat eee ls Fig. 5.26. Example 5.16. 4 Ther 7 se afr] = (1,2,3,4,5,6.7.8) Is. DECOMPOSITION FOR N A COMPOSITE NUMBER eae sections, we considered two FFT algorithms, decimation-in-time and decimation-in- and are oa, 1¢ case where NV is a power of 2, that is, ‘N= 2M". These algorithms are highly efficient lage represented by butterflies that correspond essentially to 2-Port DI < £565. Ho) whenever possible to deal with sequences of length 2” and this can be See Wwever, there are cases where this is not possible and it is necessary to develop algorith Where Vis a composite number and can be written as product of factor suc 88 N=Py P21 Pm FT's. Therefore, itis done in many ums for onf5.32) ay esos cr Here, V=Ry oy ; N-1 KZ)= V2 DY gin yk-nP2 ee where, gl] = xf] (7, egy» « pore (2, toe ME )= V2 YD) etn) Kn) = pin oe HA) = hag} Hx) Here, the DFT points fall on the circular ¢ ; sense ofthe filter is given by ; = Contour and is a linear filtering approach, The unit samp le Afr] = [Ry oop? = Ree ier 2 Ithas magnitude of Rj” and the phase angle Jv. = ona n= 7. gp Example 5.23 : Prove that Wy = Wry 2n ? Solution. We know that, Wy=e 7 In this equation, replace N by N/2 a a Wyn = € 88 = @ Proved. B Example 5.24 : Ifx[n] is periodic with a period of N and 2N such that : x(n] <> XK) and x[n] <> XK). Find the relation between Xj (K) and XK). (UPTU 2003, 04, 05, 06) mie aA Solution, DFS x(k) = Hv uae ot Now, x)= D Mel wk aD Z wel wet iy ok yak? Nel weak 92S xin a EN and ayy = Dal wk = eet iw a mo S. at a XO" Tet en

You might also like