0% found this document useful (0 votes)
132 views41 pages

Itc Notes PDF

Uploaded by

hello
Copyright
© © 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)
132 views41 pages

Itc Notes PDF

Uploaded by

hello
Copyright
© © 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
You are on page 1/ 41
- a mp re means take an ed ar eeitiie ssi. for ee a te PO ‘sovece coding Hheosern, if thexe’s 0 much of suspwise is avaulalle ,then use less noof bite fox that sovsce ~Gesely occuxing symbole GSEInSzaISOORIES. Geqelariyy o Ulen sovsce 1S evorttiog oymbolaguswecessinte SSuspmse ox oncestdinty is sel e ated +0 +he proba: bility of ocipzence - % a Probability of ocevsence of Sovonce Symbol 8 = {bo 61). SKA” Soe CS=S,. = Py Keo. =, 'k4 > — | Pel and Py =o eK S=S~ how mod Bag iee ig pesent gts a ee measuse = ™ Code wosde ase prodoced by encoder {| ao Ea eo” decade uniquely ~ effivency of encodes ee tfhaenc ep is morse 2. *# TL >H(S) ‘Sousce estodiog by mmintaining efficiency one © les END Rechix Se fot, nd use Prob of a coder ey pyre occurence Code codes be os ° ° . 4 os ‘ 10 o; wi ons co lo on 53 S12 2 thy oly $3 Sa Sa Sp Nok, Prefik x pet. o* ilo gue code Not S Bpittal = og state Hoffman codieg a gepostant HS) ) * Kato es PRN '= deviation K=0 a x 0.2 ola codeword 1 oa a ee) i" 10 Keo ‘ eects = 1:36 000 “ Proioos mia ie bast bs 3.62 0010 00th Ay | | iqinsadtaninge we need 40 stose patvate probs vias read © Soage all the. info 89) Lampel~ Ziv coding ge a0 sioples +: (01 0olol o11too — fain Sieaain. «pass +his data- @ we create a code back & aseotre 0;1 axe olseady code voosds: ©}, O11 10 O10) ol Q Present Seqvenes Gtosed 0,100) We toxe them in sumerical positions > £* 4 Nomesical positions 1 2 3 4 s orl an? 1 ( cocuot_ott ol eee ~ Sequences 2 ia = 1 @otB 1000 Ilo, = ooo coll 0! Cio Noy Binary encoded blocks: be done, s cooly consce coding PO “a co aes Je Jy x GR yx 7 = XEN Cnloise) p(tj) T=. 34 3) P(A) = PUx=j) fee x J Gump, - . ..; OD i 0) eum .. ae ge, xf 95D aso rebar) j = fod 24, at iad PCYfa) = PCY= Hk) “Gnansition PrebAbitites : . a} PU Y= [X= 35): POX) 0 25 P(Sw/aj)- POS) i ce P(Yeleo) PLY)... PC Ska/acg) Peis) PUY)... - PCa /aly v2 Peles) PUY/ag)

[a] woe dot diectly) bd Dot fivst obsesve 4 29d hen find bas much uncedtainty is these & calculate - Channel Inpet X =% HOA] Y= 4) ys ‘ud a jee 1 [os ” ak sia |e 705/94 stgas)| Pd ko L j=e ¥ 9 gels of nyu) = PCM) log ( 1 welt = FB PCI Gay) Kel! nay = 3 PC%, Ye) tog ( k=0 J=¢ — tjoint) PCa, HG) -HCx[g) —S aocestainty about the channel inpot resolved by Obsesving channel cutpat DUK 4) = HOO ~ ory) + 2tys%) = Hey) HUY la) © channel copays Cc: plogaP + C-Pdfog,,(-p) my PCY /j) C=-HtP) T-1 K+ Dip 2 = PC), Yc) "99 Lo PCH), Yad= PlYk/aj) POX) J-! Ply) = Z_ PL Yw/xj) PC) é jz0 which has Skill chaonel_is cox eect, aia ee Voit is ip estos out of (00 oe bits teansmitted mose iofosmation) bot Aine Sovste* (So we do We do channel, coding 4o viedvce BER hannel coding - Gocseases nesigtance oF a digital corpmonication systero to’ channel Noise le do some mapping of incoming data Sequence Toto channel toput “sequence Hex) Peers | ; oh keane ed amy) Diecedle 2 " x cod din fess C= Dewy) ari > a ae cee Poms TPeead= PER S= Ya) ® rr Q PCy /aq) =1-P ey ; r P Ciera) = 2 oa PUY /x.)=P et Disesete chandel ise] channel fers [tebe sovsce| | encoder) * channel Ee } ins wees Te Discuete |_| Encoder Iniaveleam) |detectos/ | [eect }1en) to aone| | Ldeodex pkey poe +e Cocle B io advance ae om & j he este transmitting 1 we teansiit positive polee Signal Y= X+N 6 Xnoise Signal 14 } coengy of tbat, bit) ! y <8 —_—_7 ve An ———_s Sn channel coding scheme Bsolo opesations -2 i Addition Subwaction oto =0 0-0=0 ae ots! Block codes CLinear) >. convolutional codes Joke K POF 0: messome— Mo ,Pry ipfesmation bits and Gereates 9 of bits (Genesally N2K) “chanel > nr mn iS encoder 2 oP idan, mesuosgnyy, we ane, adding to mq signs Mt Kk poof bits i each 1H k bits > 2° possible messages oI, : > ms oo mse ke 2, elo, dake ort q to _ Repoaee code’ Book code wosds "|! code 2oed =— = diff code @osds) Cojo word bod %, M1 Go-1 Sequences 65a 0 dita stream toanslated 10% aK of thero axe actoal We do all this +tbnovg? a Meq, sequence = Mo, Pasi bits b= bob (ode wosd C= C1 grrevaea From req bits: ana" genescled — (ish os. bi is th check [re mm, Mj I |Fo Pon Ro Py Fao Pray ¢ [be b& +... UR EDK IX DK, c -[bim) See. erm ek KXN-K 2? vectors ones ae bo -K-1 ++ Co-2,Co-4 ; - o-1 sequences . ane gpresatos mabe, IXK OD Vectoy KO-K 2100 vechyy 1xn now vectow ee Po.0-Ke=1 Plow Pat Pak Pow] p ee CkxO-K) DEKK e- wen -CZDP ea KXN-K KX K ieee a A Winean Block code has @ tnatsix gives Find C K=3, O24 2 ble médges 23 -g possible ay 9?» 23 combinations of code wowds M -(0 00) m, 2001) OK=F-3 <4 s =(219) 9 bern. m - (ol 1) m, = (100) mg = (101) m = Co) m= tt) C= mq Co= 8 =a 1x3 KXD 3nd modvb 2 ae eae addition Efe Gioac ii ee aa a qteys Crates) 9 GeMmq Cc, =™M24 K : 1st er Leb sent, = ™M% E EGr 1 Ete = MG Ct code word We ave to xemove nedundancy > Given a code book, bow +0 extract K Po.o¢ bits 42-22 Sore tock Words algo. user C= [mb] develop bole OF maby: 5 using this we con genesate all 2 code woxds i menting [To-n! Teel a = Tyg PT PT a Ha" = pty eT fo] DKK E O-KexK, pe itgor 3 qwitithe - Vek Kxn-K a8 fees per eked ox.) Posh SPAR FE One tan er ct gy is gx. Detenenine all code Words beginning caith 1 —i —_ =: oa enone oe? Soq a m=[m, m, 4 sas To] slits Stab a 6 =[3,P] “ aie G Elion 00 mJ appended with [bo by byl c-(m ™, o1 © 0 {ot roo 'rltlo ange a] eon) |, 25 eeftrscoel / ~slasting uct eq BS MMe lalhen we Settle added, 60 instead of infosmation, some esror is et © os vice versa Sea ~ deo) - (cig) ca re, fieotitd epaporanip:|! o i] d(G,G)*3- ssegponds to single code eum “rr SE vag in a code Gowd. i ght of G = WCC Hemming ig w(l2) =3 7 = Mipioorn dictorce QBs (be, ormlest barony beticeen any “too code Words: pice, e= ello) « | alas e 210000 esme bas occosed i0 vector , pose these ase 2 oP eRSO® patterns 0 2 e apie Sheacch Sees hs =eHT - 7 S2=@aH" = CetC)H may ave Same Sung. (can be) ona i Sa= Si cHT 2S GD =.1- of code ond dCi, 2) iS least Hamming weight of essvs blo ¢i,1=t- d(C < dCGj,70 Bf 2 civcles axe none inkexsecting BCerepyH zee Henan coill be identified by decodes O% Sj: 6,3) ae. check” equations for | n-k=3 k23 Oxf oe slim. m, m+ bp b ba] mM ™ Ma Oe or ° e Pp. 0 in © 0 1 [mo m, me) 3 Doky co. D aes: | Fok wet 7 ee To 6 ; fo) Genesate Hand hen G matrices ity check peers vsing pastity cK mathnk ef. ee! 2 ‘Gue implement wit Polyromial sepsesentaton OF cyclic codes Haroming, codes Bose chowdeny (Bch) codes: Golay codes Bie can wx Moltiple exons (G10 9 & decoding can be implernented with shift weqistess ~ easily): ar Memery not needed-(h No deed of ok vp tables) - L aphiicinans psi SBR 0. Ay ‘ Bonst exss - essor happetd in a continous bits, Vi bay k boy qowentwe aca aies) C= Co, G ay cyclic shift wight Cor Co q 2 cydic shifts +o sept “D2 | [> Cyt Co ers | nt eyelic shifts +o wight } de G Cp Cy tse one left cyglle shift Eaves Kroc cyclic shifts to sight is same as that of giviog th-K) cyclic shifts % left tle assodate code wosd with polynomial of (Gagpee <0 cle) ox €(p) \ SECO Cre TC. Code pelyromiat COM) = LotR +x. ten x” Che) = CoP? lng cup + Co and they belong 40 dinay field with GF(4) mx We = KEK oa +b” m2 2 ie S 2 ajb,x ‘G0 i-0 YO= dot aut a. ZC = bot bix + yes) 200) = {ert zla) = (a9 Hoe) Hath ™t -_ + Cantbn)~” + nn red x(P)q(P)'= has degree 16 ak ABromiat fxr}: 28 These ‘can be, fosmed fxorra giver GUD Cos + G0P)_ _ ce) PP 41 Coo (P41) + ,(P) = (P41) CCP) OP) = Pert x ~ +)+%t : fee) + 4 y ()> 14% ° Ero E 1 = (14e)0+%) ro cae atx 40H te tae atz)= x34eH oven GAL 2) DIZ) = a4Ex4]) aya yreyxti( rt] a xryx i se la Reet! os TR pasity ck pom = bt bat bxty 440" Dba) Xa cole Polyrorniat due qnee > D-ktK-1= 9 x (x) = ACx) ow) + 0X) modulo-2 avithmal ‘ wr mC) _atxy4 bx) ae qcay add Cx) “mox) = ca) Se Gly Linear leek code Show that these axe 2 gene Polypornials re My wet Te oe WHOLE AE (oe p-K= 4-4 lity checks pits added- Lea a 3 parity vote 9 (ade V+ a pae3 PD Yr 12> sft La v net arg Be ; - ence) xh) 241 (ater eS x+ ee aye a = x? me) {ease Te ea 2 a2(t4 423 © 24a 403 are Sat ies i) Find memainder Go as , 20) _ ox said a@ ws emda Creme) ELE 4 Pay ry oy 4 at4aF — qpoteiot ya oe oo, es poicamist xP net ae t - a4 034 2c? 2 4 LA ree a am O42, a e4eth a pene oe acad> X32. oe $04) bee)= 1 pri tail : gc, £9 0 se ” Se th (MOM DIOR ip ip, P > 0-15, m8 cen) Lacan ie g| eee yaa 43 |" 1 Vocoma « 6 Finding Genesatos potyoorniale 16 © difficolt. Pe inact i GHT=0 m eae OT Kxp, gore =0 modulo (27415 OM tet qc? bx) axe factoss of 27H 0 0 IO a i A-kx0 4 o-K KT re 1 mP=b 2 mp-b ere en renee mele bE (Pelt till Peqinaat Pee) ax 0 eh aa em, om “ex Ay = cat) CaP DCP HAH). : para Be sma be x+x7 Po tHe xara Wxte (ay ra a o PAX = ° © ho) Por check mataix st qcm bex) Mod (x41 =0 gods eat Srey ail RS aor aay is bee) bead > each SH takes” block by In fname Sroalles frames St has gi a finite Sbift wep stens cabal Je) Sequentiatty, + I” state AA a 017 Tevet ooo ; !oo Wwe FaKe linear nD Gf) Before Socerning bit ° 1 ° ca | ' for ° 1 ° 1 Convolution encoden Kel Leo OTe) affes Cossent State a} oo o1 Oo! to 10 ae (bool ois dropped et bi oo let fel oo oa to 10 O71 and the convolution encoder can be mlepnesented by a state diagsam aoonert gtote OF shift deqistes Be does 2 SRS HOt dene, 5 leDend 4op of Anellis. line mepresent ofp of encode Pill entise possible cases ©O11 1 Gol OMe ofp combo W101011, 5 op bits es decode this olp Cowsent outgeing Ei state nS OF 0° SA °° to a mt ay d, Dox other 3 cases also (ot, 10,11) 1 Si 30 a I>2 ol\o ay) So lated to discond ups State diagnam ea n- yedtor gn 5c c wosd dete: Zi weedattah vo noe me AG xo o10ll? n=00,000° qa 1) Ag6uere Har dict On line Jowarde left side cove olpy Right oie Harorning distance comet into’ pictose 19 vitertoi diagno? “ (sil) starring ut ot Be 2 ies eS eof pe a calwlote: sor of ham, de differ ther we oo as Ngee thi Re aa least harnming dictance path which bas CE path) gf bab” same, both one ees BR _—eeeeeee cosuival ot d ¢ iene 4ing " m) (4,4), mnmnor U4 ¢ onal code A 0 : a0oxe channe yed= x44 ith OPP yee) =mtk) # bE) + re) tle need toasern BEN nga esign a filles x) — ie £ hie polve is ‘fox aoe eer which bit ntt) = yet) * He) q carey 44d) HEED enenally 4 rect seoveloren cant be Qererated bez. it bas — shanp nisiog and fallin, q si system / Received Fitter g edges» < ee xc is ——— ole dy al pent tes bt ; AX in hat, +f Signal 40 6 0 4 pal * anal Maxire be sqrt fo Ae pw ¢ ul Si — { | sselation s ke vl ied comp we take expectation (avg vale of Sigoal p0 zapectation is linear Pe E(B) ~ eff ma-7) HC) 87 fr 5 2 e tool” f ef | Pte mpndarl § we of amplitude power of data symbole q PR), bit) 3 fixed a 12 efow}o : mean Rando? waniable 5 pier bird fr qF Ff bo bl) drat \ can be as auto covselation fee ae Ss. £ efor 0lT-¥), 7=t tT, ott! nl + 7-7) E ce Pb 4it ae efncr- 706 og aime seepsesenteg beg oe are any Ne BUTT) bObCT) ard 3 a ty? ye No bir) fi pen s(rtar at 2 nn Exists onl: when e-% ~ Ne fherd boar = 2 Le 2 Sed coin S08 «fA ene- B(feenbmst)> U,. frm i eign be) SS fibesPar Jou which , valve of pes SNR | 10x -. open We eqpation Se ae wo inequality tbeosenn ocho i jacd V@dL?dt ef \udo|Pat ti (Meat ones % “Lees vl? dr [ hese ay Ne [ierstan fa 2 é fy flees) dy Ne/> a 2 “en < Py f [Pa-r)I"dr eh) uct) LV) =5 ult) = KVL) tonstant P(t) & h(t) (7-7) =Kb) Gf pulse « bE) max pom pi = © ace) Design ttt) bor) =P CT-7) er) patched ae , itt ee coe of “be patebed 28 a freer at pear aa oS eae = oa lon Ooh No/y No/y a “ f eoendr ie, dg oe. —— f 2 z ever C0 mmAKe’ SOMME EHEDKS bey some AWGN is present Ore metric fo measuse pesfosmance no Ain: wb moiniroige probability of eros x) = aPC) “2 ade b= POT-B) jo i ben = PCT) i 3 ee a@® =f xc T)bCT) dr + { ocenberidr as i signal component ~ 6 To maxinnige SNR PCT) = b(T) ha a n(#)= Saf Pendt +B se hl ' a» oo 30 tems Energy L AF Signal pore OF ay - pelse flee saurpling aie 7 ao ay > teansmitted bit Isymbol etaeo elt = Ne (Order = No z SF GB oNoice offer sampling 2h hos % property » additive wide cea Noise gah mean 0. NP Ep 2 jm (or Seer) voniance no, ) Standand eo av ale) = a0 &p +0 get & CAAY ex yotrto Jos weal san gmissioP lo 4A Pdf o9-A of acv)= AEprh fos a= A fees t= -A Spt fos Oo= -h 2 > of neN (he ) i Bate) pie HK se propentional fo sormat ner) 7A Spt 0 i with ome mean & Vantiance ? gictsibbuto by NCHe?) 4 “ col. mean is changed SMEG decide a A tO — Pr nobability of es Considen, transmission of Qo=-A 5 let ASp 2A Enos occvsS When %CT)>0 wt) Hind iotesm fon eyo Nea) eel) 4 ae Tse | Ww ist poser] given " Receiver = | iHes at tne *°T bie) | 4) > puge signal jantance gpm reco Nels a. ———=——————— Receiver: detect pulse - optimem 7eccivert the effects, oF noise minimize signal + noise poo io oP Sign od - jpstantaneous aici pom poe l gecoP efrayy > wang pie) ~ filles seopers® che qe +ult) xt), a yoo= x) * bw) Y= x(PHO) = {gen +N ENE) 2 GEA) HED) FNAHD cee Siparmrrrcot cee RO Retr [SacHe Pe tL <@ saropling v se when filtes olp sampled at *=T rise is a mandem signal, we ; cant compote a epeteuny aheectly i ht Su) - —— x tolds + value | 29 oo \ (REE) op GANT) af |? Piven ae ae eqn Ne f HEH dF ana » jor a ae valve of polse of pontticulay how to design Ht. oI that # fap eB) ouch thar # makes a tax vale of fficiencs i 3 2D PDK OMe > VSi0% schwag; inequality tbeosem capt me petals ‘ Holde eqpalliy only hen one ig factos/ Scalar mottiple of cthen. 1. neers flares ee One) eats: Dac, No f pene SS [e@lae assuming * HG) & GCF) on related depend on bit), depends, only o signal enengy and noise power spectsal density oe consider ge © % ee se. ot aa aia FA (LAM Sy translate’ © ryooo ot lolll wt, Toc) of commun enor Ry aftey is teanemnitted Data »a What ig DR when ore bit sequentially? aa bit > Ts & 166 peed to + the data ratte Obl sec. ao ; ‘ Sei atomee 00 30 tirning pesiod ap vies 2 ttyl = Pr nal 200 apace concept at) ale 2h Xb) <2), TCE signal HWaa@ll= \% Pawoah cst cola col = > docoming message signal mitt OM these sigrals ane peste onto Se modvlated — Sigoal PNENPETIobee eBtovination BF HAsts fSottion + ostbonowmal ve ul: ut she et ig vowrmalized by ie > We gpnesate this 9) transmission through extbowomnal - Clineax comto of basis fonction) wv oi These ane orthonosmat Real valued signals Si), zc k, emit) each of 7 dysatiod S10 = Boy 7 Fi Ne. ae ane ages © Siw 0 ae = called Ras pron al BERS EOD i z — exthoooxmal basie fonction dimension fonctions, yaived OS'S = ofthonosmnal Real v be), Oo). ya) ghee fooctions one nosmaliged voit erergqy aie Meceiace y> element of N dirnensionat vectos Si Si- [8S Sin] Sin ihe 4.0) 2S) Se SS) 6,0 oe t butt) = {ae gcodt ats al— Six Sia. it aoe ies L_@—fpaj> Sin Gy? : 6, poxmalized signal if P Iprer Yuide product es {wre 6 il 5,051 : ge, Gu 18 COeHbCen gpa be e Be + 2 3 Nay sij= [aepgcos st Su = (aes e®yy = Er ). St) at = i Gi 2 = [ott Se 6S NG Secord ete =6)) Sore fes w od, ij ea? a cae en 6 gobtoachog aul qh) progeetions Trig & powmaligi?? g2(t)- $21 di E-Sai S foesqy of Si ah jdt =! i $i) E5580 eat 4,(0= ae gro So 2 - Fswjarde Kapow. ni $04) = Gi) To rae. a ayy ieee & iodependeot- 3,0) Seat) are Poeasely Nam (+), Sqq@h) ane EOF Voeanly independent nem 3) sal) S30) ne ea pe may Of hut Puech 4 tut dont Know we frat “a if Se 2uldt = 273 273, Ey=T ong |. octe 3 v% & otheatoise e - otheswise Son =f Sal Hat HE = i ro ) Pele icy Find 45 tt) & posal ase ° ae pe a € {! aget Jo. baseband Re ise eratet Nee fee. y > Co: GBBso hey we ate, modolating es freq mange, the bandwidth 1s EBs antenna propoyaticn Amp, Ineq and Phace ane 4oi0F ti hed v ( gk actosding + signal aceowdiog to req gral L conde SP dshBS dows and a com modulation, *8 30 4bi freq, axe embedded otto cases ated L 2 bik, ak O time 0 ‘ Now tnansmiesion +o detect theee sigaales \ Solies we ved a8 optimal “filles Same We. ose How matched costain i Ssves, that can, be Concentrate _ off |. Max dol wate 2: Mio probability of ossor. CRecelven shold give + Pe) Scone 3. Lees power transmission: Ye Minicom baed width. (channel) Max 5 pictmore ABlotance -BIMEE Fesence 6. Minirodm — Cisentt complexity ig COsties™ Wave. 1S 152 ager 22 M22? 28 00, osi05" ae . ' me ve eee 16 done PRR ET tte e sign’ desig? oe looking, ok P S,(b) v bit” Bact)= Sievilan Wy Ee + (& sintnc) ~ “unfe average prc of esx the enerqy o: diay fo majntion same BER of esay >F 1' comes on-off level encoder dye [Fcooan dit gers eed divectly ! > 9E'S comes mit) is 0 60 prodod © 0 at inverts he sigoad te UP moHiplication becomes negative talk aloout Pane oxmerseseu tion ware fowrns abt ag bet *t mipol enenge, enesq yg, Father

You might also like