Scribd
Upload
3K views189 pages

Digital Communications Notes

Uploaded by

messenger kkd
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
3K views189 pages

Digital Communications Notes

Uploaded by

messenger kkd
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/ 189
Onit-1 | A Purse DrarraL Mooutnt20N Communication Channel Demodulator Description of Each block ;— DO PR RRS TThjJonmattion Source : Tt is Classified into two Cobegonies based on thei Nature at ole nm finale Th onmaken Sounce % output Varies with time d Continuously ee Micxophone actuated by Speech 3) Discnete Tofoxmalton Source + output doesn't vary Xi output of teletype Machine eee a7 al Source Encoden $ Decoder 3- Symbols produced by the tajowmalion Souxes. axe given to the Source Encoder . These Symbols Connot be bransmitted dlinectly . “They axe Converted into digital form Cie bioamy Sequence of I's 8 O's) by the Lounce Encoden. Fach binany “Vand ‘o's Known as 0 bit - At the Heceiver decoden used to perfoxtn the. Hevexse operation +o thot | Aource Encoden . Tt Converts the binary output O the channel decoden into a ymbol Aequence . Charosl Encoder and Decode ?~"The Communicaken chennel adds nolse | and inten enence to the Aignal being transmitted . Hence. Emons ore intyoduced In the binary Bequence seceived ot the Heceiven nd “Thus , channel. Coding is done to owoid these typos of evers Channel Encode adds Some sedundant binany bits to the inpuh Sequeres and also these nedundant bits ane alioays added with Some propeslt detined logic. However these redundant bits ane Caysy No tnfommakin sother they ane used by the channel decodey to detect and Ayiyoct errs ony Modulalon ond Demedulaton 2 q Modulaling Aignal. is digital (Binary) then dlgitel modulation techniques axe used. The Conier Aigoal Used by digttod modulatos. Is alcays Continuous sinusoidal wove o| high _ (tequeney Signal. €x s- Amplitude shyt Keying (ASK), Phase shijt Keying (PSK) , dh{jenentiol Psk , Minium shit keying (MSk), ot the seceivey end, the digital moduloton Converts the binany Information sequence. {nto Aignol toavefoxms - The diagtal modulator cnops the bit'o! ba Wowefonmn Solt) and bit '!' toa tovefonm Silt). aa this monner each bit rer channel encoder is transmitted . sep Demodulator peyorm the neverse Operation 04 Modulatoy digital demodulotoy Convents the input modulated signal into the Aequence 4 binnwy bits Commanication Channel s— “The Connection between tHansmitter ond. Hecetven is established though Q Communication channel . They ane Alypes 1. Witeline Communteation Q. Wireless Communicolion Every Communication channel has ome inherent problems i) signal attenuation i) Amplitude and phase distoxtions li) Additve noise intenjenence iv) Multi - path distortions _Avetages of digital Communication systems — i] Digital Gommunication systems one simpler ond cheoper Compasted to Analog Communication systems because 9) the advancements in the Te technologies a Tn digitah Communication , the speech, Video, and ober data may be | | merged and tnansmitked over a Common Channel using Maltplxing § 2) vsing data encnyption only permitted steceivers roy be alowed to detect the transmitted data. This property is very Important in Military applications. 4) Since the transmission is digital and using Hegenenakive repeaters in the channel Hesulls mo accumulation Qf olse Hom nepeater to epeotey 5) Th digitol Communications » Channel Coding is used . ~Thengone. evrovs may be detected ond Connected tn the Heceiven 6) Tiarsmitted Signal is digital in nature , thenejore darge rnounF of Noise intenjenence is tolesiated . 4 Dig itol Communication is adaplve to othen advanced branches ot dato. Processing Auch as digitol Aignal Processing , Tage Processing data Compression ete. Disadvantages $ i) Due to Onalog t digital Conversion the data ate becomes high . Thenajone mone transmission Bandwidth is nequined ton digital Communication Fy} Digitol Communication needs Sypchongation in Case 4 rmodulalion Syechronous “Code Mo dulaton : (PCM) aaa fa [ ulse PCM is Known as a digitel pulse modulation technique . PeM is quite Complex Comparied to onalog pulse modulakion technique in the sense th E message Aignak is Subjected toa great number | operations ements of FEN Syste Tt consists of 3 main pants T) Thonsmitter gthes i @) Transmission ie th how pass Pai Sampler any ' agites @) Receiver _sigeol POM Tansmitten ;-— sil fe Dail eee “To Convent 0 Continuous time Aigral into digital binaxy ee “There 2 are 3 Steps ints) a hk tt ol 0 Eneading | Samplin Se Sampling Is the Process 4 Conventing Q Continuous time Aignal into discnete time Bignal. To Heconshuct the ontatnad Aignal. from Aampled Agnol is possible only when [F226] which Is Nyquist nate. | $32 Sampling qrewery = No-4f Samples see $m > maximum reqency present in the Aignal | Mathematical Eoalysis 4 RK message. oSignal_] =x) §- —z Mattpliey nse > SEW =" -AEHTy 0 Ts 25+ > - “mm APTS Tey Tinpube twain SU) = 5° § (k+nTs) a To Prepresent Aipal in fourier Series ~ Oy, YW= Oo + ZH wn@sningt +z 19 SiN Np E 1 Bor ~ LB qe innputse Hata ay = tfsudt 2 1 0 Ts a 1 Tg - Qn = “jg J Bee cosnut dt = 2 one tdk= 0 () sinnws e Ts]Q. J o &r, () = L 4 F& 2B cosnwst Pee ei [1+ acosest £2 Cosawst + J OS3Dst + a) Ts qt) = aU) Xx alh = ape + AXLE) Costogt + A2Ut) Cosaust + - - - aa) “Ts Taking. Fouttien Exansjoren on both sides X(W-Ws) +X (Wt Ws) W-KT — xt) Cos wWst = a G Ww) = oa [xeo) + X(w-Ws) +x Cored + x (w-AtWs) + XFL iS + Xlu-3usytxworsesyt - —- | ~ — i = = S XWends Ts ato ) ” | Few) = f Case it— From the Aquation | Qw) Ws= Wm = Wm Ws= QWm BH § = 8994 fry $s= fm a Catal damping Te Hecover onigtnal message Argnal,, ideol LPF is to be used Cose fi Gw) ) = ws ~ Wert “Us —LogtiOm— Wm OO WOM WW ing Wm < Ws—Wm Ws > QWm | Bhs > AX SMF m | [> Bia oo snp i { \ \| tHene We Con Hecover message Signal, Case iti = ——— - r) pos eae (0s-t0m lost wen Wg -Wm < Wm Ws afm > “here ane 3 ompling methods I Tdeal Aompling i Impulse at Lach Aomping instant a Notuel Zampliog 2 A pulse 2 shoxt width with aging amplhide 3. Flattop Sampling 2A pulse of shoxt width with Constant amplide Aenghtude Arvpltude, Hime | Tdeal Sempling Natural Aampling papi time Flat top Sampling Qecntgaion Is epresenting the Aasmpled Values q the amplitude _Quantyaton ee by a (lnite et oF Jevels , “the Signal is quant ged nto Gunlable {| discrete Jevels Known as quontiration Levels. Bach gy these. levels | Hepwesent 0 qed input amplitude. . Dusing quantdion the input ompltude is Hound at to the 1g KF { peanest quantized level | td. ce st psqe cs Total dynomic ange a signol Wrox —Amin | a s- Total na-OF hevels EE t Am - (Ai } jor Srusatel dpe = ACA) _ Bim / No-of disexiete Jevels on questizalion Jevels = L aut) | gelevel | Ag Ls) ' +> sampled Quanti er guantiged signal sigral Quonlyzakon Enron = Xq WT) — 200) © Chassienton, of .guantigntion process + classtrcation based on Atepsize | Quontiyation 5 | apeaeg SRC NET GL Uniform Quantpakion Non Untjonm Quarttization ees Midtnead type. Midnise type | midtread on midwise bype. Of the Ataincase wowefonm . a” ) i aA | } +2) } 1 1 2 [tie Hise the Atalncase oul ihe Unijonm quanti t- oa union quantger the Abepsiye Hemains ame. thyoughout the input nonge. The guontgen Characteristics Con be | Midtnead quoctigelign 3 the oxigin tes th the middle of the tread qolp Xqiots) NMR mea 1) Midwise uogtiyation. 2 a} the onan Jies in the middle q the jor Aq ints) we We we we = = 435 Be [fp Hints) wit ew we wis ANE (Guontgation = 4 2 ip XCOTs) > _Quosliakon Evo" “(E) = Quankger olp — Quantizer lip 0 Xy(0Ts) — XWTs) dom Exomple —y) x(n) =A then %q (nT) = 4/9 2 Aa, 2 - e* | Zim Sake Strolasily it XT) = -A ten Aq (0) = xf | €=-4 -(-a)- 4 Fiom the above Axomple ibis obsexved that the maximum quontyation Eyvoy is A a ‘oding 3— Encoding s Encoding is the puiocess that oteas the Aampling end quantyation Ereoding process Converts the quontiged Aarmples into Gdewoxds. Ino bioawy Codewond each Symbol hove Sithey 0'0' oy ‘1! SH & quested Somples 3 Bains {0 a ol \ na Ts 21 ats UT + Novo bils 40 Hepresent “LE en0-o4 guontizotion ovels t's from the above Axomple de log.4 =2 Aten Encoding VOllO110100 (Binary Aequence ) | Rilse Code Modulation :— (PCM) PeM is Known oF & digitah pulse modulosion frechnique. | TE Gonsists oj 3 main pants | I) Thansenittey &) “Tronsmission path e 3) Receiver om Low pass Binany filter with Eneoden cut-of} fry, Gigit3e") $e= $m Continuous time mMessone signal Pawallel Sexiod Convertor $57 5m => Signol HUD is ginsk passed though the Low pass gilten with cutqy frequency fm (HZ)- This Low pass Jitter blocks all the qroquencd Components which axe above fm Hp “This means that now the Aignal Ct) fs bondlimited to fm Ha. => _ Sompler then Samples this ignok at the Hake of fs. Sompling deg, fs is Selected Suyfficrenty above Nyquist nate to ovold aliasing iey fs z asm . The output q the Zampley is denoted by 0Ts) . “This Aignal XUOTs) is discrete in Hme ond Continuous '9 amplitude => aA g- bevel guon tigen Companes input 7Lt0Ts) with its gired quant zalton Aevels . th assigns any one a the quantization ewe to xcnTs) which Hesults in minimum dishaton on yey. “This Ewor is dalled quantizalton evvoy \ ) => Now the quontized Aigral Jevel %g (Ts) is gen fo binary Encoder, “This Eneoden Converts input Aignal to 9 digits binary word . Thus %q (Ts) is Converted to v binary bits. This Ancoden also Known as a dg ges >v binosey bits ane passed though parallel fo ental Convexker go thet we can tansmit data Aenially PEM “Transmission Path ¢— The poth between the Pom tiansmittey- and the PeM Heceiver ovenwhich PCM Aignah travel is called as “Pom tnansmission path". Tn the thonsmission path we use Hegenewabive Hepeaterts to Hestone the Onigina Aignal {om distorted Agnol “These. Hegenevokve Hepeaters ane placed at proper distances in between the channel so that the ¢4ject Of Meise on the Aignal ges neduced pistonted Reqenewative ~ = =~. 3|Regenenative pcm Hepeaten ofre. nepeatey | 0 [1 LO clean eee Clean PeM PM Block diagram Regenerative Repeatey , pian Distorted, Per! Amplitude basin Equi i oti ]e mm F Timing cimeuit => ‘the Amplitude apie Ahapes the distorted PCM wave so as to Compensote the ejects of amplitude ond phase distortions . => The Himing cincuit produces a pentodie pulse train which Is denived vor the output 4 omplttude Aaguiliger « “This pulse tyain is then Applied to the decision making device. => The decision maxing device uses this Pulse brain fox Aampling the Aquliged Rem pulses. The Aampling is Caynied out o the Instants where the Aiqnol fo noise xoHo is maximum DSThe decision making device. takes the decision using thyesheld Level . ot Aampling instants i Aignal Value is flower than threshold Level then it gives output ‘0! . if Signal value is greater than thsieshold fevel then output is, '1" --—2- Threshold. Value Ohta ae PoM Recewvex 2-— Clean en Pass bi, Repenenative — s Hepente => “The Negenenotive Hepenter ok the tant Of Hecete Is fo teshape the pulse and fo nemove noise. “his Aignol ts then Converted fo Paxatlel digit words fon Zach Sample => The digital words one passed thwough digital fo oneleg Goverkey (hich qives the Aampled Voltage Values ond these Values one passed thou Sample. and hold Cieuit which gives output %q Cb) D> Alb is allowed fe pass thwough a Low pass Hecanshuchon jZltey to get the oppropriote Oniginal Message Aignal xct) + Ts 215 31s MTs BT 6S FR BT mmission Bandwidth of Pom 3 +0) bik: “Thonsmission tate (0) Signalling nate (%)= Noa] bits = fo-of bits x Sample Sample Sec = vfs Bandwidth = 4 x signalling nate Je a ple v xYos + $s ot nyguist nate fs = asm Bandwidth = 30 ASm = 0fm Bondividth of PEM = Vm Where VD + 0-9} bits pen Somple. Sin Message Aignol (requeney fs: Sampling qrequency Signal to Quantization noise Hatio fox PeM system 2— = Noxmalized Signal (SQNR),,, = Nowmalged! sigral power Nowmalized noise powex Qustizalion noise | ever Power =— IAA AAR AAA Mosimum quantyakon roy in POM = +3 Xo Untjoxmly distibuted sondom Vawlable, then Probab ity density (urcton is 5x 1 a ees P| D Othenoise oO b Let us assume Quankzation CxxoH Is also Unionmnly dishibuted then &y using the above foxrouta, U ; Fel) | $f. = #-[4) ases$ oO othenoise = —=9 in pe eID In Mm ” p> 1 - = a . Othenwise. ¥* Mean Aquane Value measures the average Atrenath on Power g a signok . * Meson Aguane Value of Hondom Vawable 'y' js Bapressed ae fs ELM)= S22 fowdx 00 using this oneal 7 we Con Calculake Noise Powes, ao e(e] = [ee© de q -~ 4b ao 1 37 4l2. fe pee - 4 68] 42 Ab 2+ [af a [= -[s]] Ce 3x8A 1a Quantization Moise Powen = Signal powen (Pp): — Gase ts a) toput Aignal ts Ainusoidok Am f 2 Noxmolyed Arqnal power (P= Vans R =Amfe =~ Where R= |2- Vyms = Am va “ Nonmmalijed Aignod powex (p> { on Ainusoldal = (Amis )* Am R “a => (SON Roem = Nowmalized signal power (P) Now malized Yuantryation noise. power (= 2) a Q / pr Ta An. 8m c Qo we Know that Akepsije A = where Li no-of gpankgakon devels vi No- oj bits (SQNR)pcy = Rin 3 cM, 2 [22 7 Be pox =) a Ape 12 = 3g 2 a ad) (SONR)pey = FS ¥ 8 (SONRDpem = 15% gr Gren | Sn) (SQNR) gg = tolog (sed y _ fol hs + ioleg 2” 7 o 10 = h46+ acdlog a (S4nQ =(46 +69) de Coseits— Jon non-sinusoidal Aignal (P< |) (SQNBpem = Nonmalized qpankyation noise power 35 ) p=} a (SONRDpem a a CONVpem = Ne), 12 (S8™%ren Am \2 ( 2) 3 2° (SGN gare” APs Assume Am = | SNR) pom < 3x7 (SQNR) yg = tolog, (axa) Ss lolog 3 + I6log 2 SONR < (+ +6¥) dB Problems on PoM 7] Fl PCM system howing 16- Level quantigen then find the no 9 bits, tuansmisston Atgral sake , bandeoidth aq the Pom , Argnal to quankyattor Nolse Halo where messoge Signal me) = sin yoort mgt Ag Given L=16 Mét)= sin4oort Wm =afrfm = Yoogt fm = 0043 * No.9 bits = y= fog = log = y bits ¥ Signalling Hote Y= UK = 4(afm) = [600 bits/sec [Bi = Re = x00 = go0 bits/sec SONR =(|.46 + 68)de = (-46+6x4) de = a5.t6de 3) Caleulate Aignal fo quantyation noise valeo , Lransmission Hate z N0- oF qpantaatton Jewels, Borduidth jo" a-bit pom ond its Sampling time is ams Given Y= Sbits , Te=1ms fo \ > — = Iki Ts im > 'K3 ¥ Signal tb quanhyation Noise Hato < (4.7 +68)de % Tansmission yoke (¥) = fs = SXIKH = &K bits|seo. » @Bandiotdth Joey Egy = gxsk = UK bits/sec * No-of quantization Levels q = 3% . 9% - 966 Levels Advantages of Pert:- De provides high noise Iromuntty 3) Due to digital natuve a the Algnal, we can place Aepeakers behween the txansmlttey and the Heceivex . Thus , Hepeakers Hepenevole the Hecelved PcM sigral 3) we can Alone the pom Aignal due to its dtartol nature. 4) We Con use Vaxlous Coding techniques Ao that only the destted person Can decode the Hecetved Signal Drawbacks o Pom: I} the: Eneading » decoding and quantiging Cineattyy dth as compared to the othe Agptems of PcM is complex a} Pom Hequines hasige bandwi parding 3 _ Commpanding iS Q non-Unijorm gueatigation Tt is implemented 4o Improve the Signal to quantzaton noise Hallo gor weak Argnals SQNR = Sagpol powee LP Bho Quankyaton noise power “Th Case oj unix quantiyotion the Atepsize is fixed: the quantization Noise, power Hemains Constont . HoweveH Aigoal power Is not Constant Tt is proportional to the Aquarie 4 Aignol amplitude CSONRX PL vile) Signal fo quankyaton noise fou the weak Algnals is Vesy poor. “This | wolll -2pect the quality 0) Aiqna. Hence. the temedy is to use Companding Comparing is a tenm denived rem two words 1.2, Compression ond Expansion ety Grpanding = Compnessing + Expanding It is diygeatt to implement the non—Uunyjowm quantization because iF is not Known fo advance about the changes in the Agra. evel Th ‘Vontsum | Py ps Pema | ole Componding model, The weak Algnols axe ampliied and Atvong Argnals ane attenuated befone applying them to Unjjoxm quantgen . This process is Called Compression and the device which performs this operation 'S CompnessoH. At the Hecelver Sxactly opposite Is Jollaxced which is Called Expansion. The device. which performs Expansion is xpanden . he Compression of Aignal ot transmitkey and Axpansion at the Heceivett Combined together Called ag Companding, cages 4 ovrp =| Gain Jon Weak 5 Goi fo Sheng Input sige s Ip is high high | Compressor Input Sipander, Input Gm) pressor chanactenistics Ex porden Chanoctenis bios Companden chanactenistics outpot valle Compressor Janeot ronnstentsttes TThene axe t00 methods 4 Companding t) dow Companding 3] A hau Gropotcing H- haw Grpanding +— Th 4y—Low Compandiog 1 the Compressox Characteristic fs Continuous . Tt is Opproximately Aineax om Smolen Values input ond Aogasithmic (o high Input Aenels - The Jas Compressor Chanackenistc i's mathematically Expuessed as 20) = samen a) Amox < 14) bo Oy) where O< 1%) |fneodex |-_-s, input DPeM Ts) + Signal [Prediction filter %q.00Ts) rom above diagram a Predichon evvor ecnTs) = xUNTs) — * 01s) “the predicted Value is produced by using Predichon (ten “the quantizex output Aignol eq{Ts) and previous predichon is added and given to th k Input qj predicHon iter Chis is nepnesented ab AglnTs) .this makes the prediction more ond moe clese 1 the ackiol Aampled Argnol we Can obsenve thor the quantiged error Signal 2g (NTs) 1S Very small and Con be encoded by using small Po-a bits . Thus no-44 bits pox sample ante seduced in DPCM. Quantigokon Ayn its) = 2g6nls) ~ e OTs) CTs) = 2XTs) — Qewts) 2g (07) = L(oTs) + Cy (Td = RWTs)'— @ MTs) + eg (NTs) 2g (Ts) = XT) + 4 OTs) tence the quanted vexsion 4 Aigoal qn) is the Sum 4 original. Somple and guantization evvoy and it is independent of predicon ilten chanactenistes Receivey oy DPCM system ¢— The decodes first Heconsbudls the quantized evvoy Aignal (7m iMeoming binaxy signal. The predicton filter output and quantyed exwor Signals one Summed up to que the guontied Vexsion a] the onal Aga “tus the Arpal ot the Hecaiven diyers rem chia Aignal by Quontyotion Cvroy gints) , which is. introduced Permanently In the Heconstnucted Aignal.. DPem Advantages q_pem ; | Difgerence between xontsy and Riots) is being encoded ond, transmitted by DPCM technique. a small dijerence Vollage. ts to be Quan bred and encoded . 2. This will nequine fess no-04 quantigation fevels and hence fess No-o} bits to present them 3. Thus Aignaling Hate ond Bandwidth Q) DPem System will be Sass than POM Delta Mallon +~ We have Seen in PEM that tHansmits oll the bis which one used to Code a sample - (840 = vfm). Hence Aigoaling Habe and tuansmisston Channel Bandwidth axe Quite Jorge in PCM . To overcome this prtablem delta — modulosion ts used evror : ‘15) Bina: exnts),{ one bit | bM Bicden m4 quantiger output Rn Ts) ee 4 Ulr-0s Sampled Input ere) q t ‘ 1 ' i i ( ucnTs) ( (ot eee eee | Accumulotox Delta —modulotox transmits only one bit per Sample. Hone the present Aample Value is Compared with the previous Aample value and ges the Hesulk 4 uhethey the amplitude ig tncneased on deceased and this Hesult is transmitted. Here Atepsize is (red Ceonstont) . The dlijexence, between the input Aignal UTs) ond Stain case. approximated Aigo ts Gonjtned fo two Jevels y.e:, +A ond —A .Now present Aomple is Gompared with previous Sample . } the dijfexence {8 positive then Opprokimated Aignal increases by one Atep . the dlp jerence is Negative then opproximated Signok dee eases by one Atep Mathematical Expressions :— “he evvoy between the Aonpledl Value (07) and Last approximated Sample iS given as ems) = x01) ~ Row) where els) Is @nxon oF present Aample A a Hols) = U(W-)Ts) depending on the ign error e0NTs), the ign 0 Alepsize A ig decided . bint +4 Lens) > ReR) “A -2XTs) & Reve (ony bois) = Sqo (eam) a bOTs) = +A. then binay ‘I ig transmitted bUTs) = —~ then binary 'O' ts tnansmitted Receiver Part -— The accumulate x generates the Alnitdase approximated Signal output ond ig deloyed by me Japp Peniod Ts. Tt is then added tp the inpat signal 3p inpub ig binang ‘" then It adds +4 step to the previous output Cuohich is delayed) |] trput is binary 'o' then one Step ‘A' is Subtnacted qrom the delay ed Signal Also the Jow pass qilter has. the Cut prequency Aguol to highest qrepency aulb). This LPF Amoothens the Stalncase Algnol to Heconstnuck oweginal message Aigoal Ut). -a-- ron demodulated Accunulato x Advantages of Delta modulation 3— fon one ample thenejone dth ts quite |. Delta modulation tHansmits only one bit wate and ttansmission channel Bandwi fy Gamponed fp PCM. the Aignalin Small fo delta modulatto The tuonsmitter and Heceiven implementodion is Very Much dimple for dette modulaiion . “There is no onalog to digital Goventter Heguired tn delta. modulation . _Prauobocks of _Delto. Modulation ¢ ~ Delta modulation has & major dnawbacks J Slope overload dlistoxtion Q Gronilon (ow tle noise 3 Slopes overload dislotion, —This distortion arises becouse 4 Jovge dynamic worge. of the input Agrol. Fim the fiure fhe sabe Of ise of input signal Ub) 1s bigh thot the Atuncase Arggo. Connot opproximate, iE. Hence, theve is a Jove exnon bebween the Staincase approrimated Aigral and the original input Aignal. XL). This enon (on) noise Is Known as Slope overload distortion 3) Granular on Tle, noise, :— Gxonulax on Idle noise oceans when the Atepsize Is too Jonge Compared 49 Amol Vaniosiors In the Input Aignah (as shawn in fiquie ). This means very Small VoxtaHons in the ‘input Aigoal ;the Ataincase Aigoal ts choniged by Longe amount (A) becouse 4 Jorge Atepsize (igne Shows Inpur Aignol is almost gist, the Ataincase Signal tut’ Keep on oscillating by 24 oxound the Aignal.. The Srror bekween the input ond Opproximated Algnal is Called os granular Aoise y gatorton Stamcase Aves approximated signal Grooulat noise. FL F Clenoisey Condition fo the_slope overload distoution het us Consider XU) = Vin stnairfmt a moximum Alope 9) AtaiHicase Wouefonm = pSrepeijemee A 4 Sampling pentod Ts Slope - overload distortion occurs tf ros 4. 20| ees A Ts mox 4 Gin sinarfnE) oA Ts Vin fm Cos tangnt) | > = s max Vm fm > 4 Ts ae Slope-ovetload distoxton occurs ad” Sie | To ovoid Slope overload distoxtion Vm < aE Signal_to guontigation foise xaHo in Delta modulation ;— (S8NR)oyq = Notmalized signal powes Nokmaliged quantization Noise powest Quantization noise powex in DM; Moximum quantization enor in DMis E= tA TE x sO unitonenly distributed random variable on (a,b) then “then the probabil density function is 1 hO= fb to oo %=6 0 othenwise Stell the quantyakion ennor Is disbibuted on (- A, A) then the, density urction is fels) eo © = ee Afedde A ff .si4 4 = Qa SF). a Quantrakon Noise power = d* " poset = a2 “The delta modulated Argnal is passed through © Heconsbuction Aow poss qllten» “he bandwidth of this Joo pass file is fe (utp poqpney) +Anea under poten Spectre, density gives paver Psd ra Bhs fe Js ty DM signal is passed through hewpass iter with fo = AF Nonmaliged noise power of fitter output én wa a y fe = &x 3 a Ny the Sinusoidal Algnal Nonmalized Signal powen 2 — det us Consider mUt) = Vm sin 9 fm E wm VN Noumalized fetge p Signal power = Vims “WOE R whene R= 1 Sane 2 = [ay = Nm bey = TO avoid lope overload distostion Vm < A Ql fmTs at minimum Voltage Vn = 4 QmfmTs Noxmmaliged Signal powet = A> 809 TS “Signal #0 quanfgakon noise statio = Nermalyed siynal power Noise power _ BI fit Te /@ x fe fs) “= x 3fs 3 Wirt Fe” SMe EE | (S6NR) 5.4 = where fet \ Si fnfe TS ° 4 fe =f then (QNR)oy = a (sy Eaoblems ms |. Colalate the Hequined Voltape to avoid Alope overload distoxtion oO tessoge Alanal cath qrepeney IK ond stepsize IV ond also Colculoke Algnol to guantyation Noise Hato Given dato, Siepaye Aa ly $5 =2fm = aki $m = tt to avoid Blope overload distortion Vin < ahh ‘es 5 ir mer Signal to quantization noise Hotio (SNR) py = (2) = 3 aks $7 (KK) =e A a (SOND pu, = 01303 | “To overcome the guontyation Ennons due to Slopeovertoad and gronulas noise, the Atepsize is. made adaptive fo vaxiations in the Input Argnol Att) - Fox dypasmic Hange of Aignal 21) , the Stepsige is incneased and stepsije is Heduced jon Alouly varying inputs This method is Known as Adaptive delta modulation (ADM) hogic Jor Stepsize Control “Honsmitter pant : ‘input suits) one -biE quontige outpal E A (Vanioble Sige) +t 2—+— Summer “The Jogie {o" Atepsige. Conbrol is added tn the diagram The Aepsige Incneases (ch) decreases accosding bo output 4 One —bit quantizer > AE onebit quantizer output is high the stepsize may be cloubled! Jo Next Aasple > Tf one-bik quantizer output is dow then Alepsize is neduced by one step | The Heceiven Of ROM Contains boo portions. The qinst portion produces the Btepsye- jyiomn ach Incoming bit The Previous input: ond present impat decides the stopsige. This then applied to Bcenrulator wphich builds up, shinanse woavepnt The LPF thon Armoothens’ out the Staincose wanefowrn to seconsbuct the Cniginol Agral Advontng jes | ADM x The Signal fo noise nolo becomes betteH than Ondinary DM because. of Heduction in Abpe over distortion and granulan noise # becawe of variable Slepsige ; the Ayrosmic mange of ADM is widen than simple OM * Bandwidth vblligatton is better thon DM G poxisen Among PeM, DPcM »DM 8 ADM: Parameter | Companison Pent prem DM ADM | No 4 bits TE Con use 4,3 | bits canbe Uses only one | Only one bit is On 16 bits per /™eMe thon one. bit Jor one | used to encode but are Less L Somple one Sample P than pom —_ | 22mple r &. hevels and step Novey Level | simila to Per only too Levels Recording to Sige depends on No Oi bits (1-2 fixed 0-4] | stepsye is qed Aignol Vasiostions devels axe and Cannot be Atepsize. Varies Stepsie is Kept used Vanted G-ey adapted) fixed ° 3: Quantization Ever | Quantization quantaation Slope overload Quantization 7 and distortion /evor Is | noise is distortion and | noise ts on no- bits | present Panulax noise | present but Piiniihl one present | othe enrons without Compandiy very high A One seal distontion 7 44. Transmission | tighest bard- eee] eee ee aot wath fuoidth is noquined is |-width is Bond width is Gandwid : required os | fowex than Hequixed Sequited Tro of bits arc hat Pom | Poxameten POM DM Rot by Compaxison ee » 5. Feedback | There is no |reedback | Feedback Feedback geedback tn exists exists Exists transmitter On Keceiven 6 Complexity System oy simple aa o — mMple Q Tmplementa, Compl P Simple - ton i UNIT- 2. 0 _UNTT~ 4 DIGITAL Mopuratron TECHNIQUES oc ee — Types 9 Digitol modulation “Techniques ?— Cohenent Non - Cohexent [- Comnien at the Receiver is I. No phase Bypehtoniged Canniey phase locked with Conniey is needed of Recelver. at tansmittey q. Less Complex 8 Mone Gmplex : ance is Injexior fo 3. Bolten performance a 4 Cohenent Mote : Th digital modulating fechnigue the Input is olays a digital ignah Qnty senig cess, ht eg Tee, ib Bask | DEPSK iv gpsk | PSK PM BFSk | M-any ASk DPSK | M-any PSK M -axy PSK => Th BASK | BFsk, BPSK tansmits only one bit at a time hence there ane only Q Vasiotons erthey 'o' on) ' => Where os in M-014 , there axe ™n0.q Vawiahons ane there. a Gneidex 4-any Fsk then Q bits ane passed at % time Where thee axe 4 Vvaniotons © scanned with OKEN Scanner Cimplitude, shift Keying) oe i) Bina 85H [ov=oFF Fey. + “The ammpttude qj the Camen is vanging [eying | switching accede, 4o the input digital Signal . ¥ Thene arte only @ Vaxiakons Compltudes > only he 08 | | Woweoxm ASI wm MAM _tnaa! oem DT the Heceiven ecaives Commion in Ti, duration then it | i genenates ‘1! i Consion is absent in Th it generates ‘Oo! That is why BASK is Called as ON-OFF Sigratting Generation 0) ASK Sigral Digits signal Product |, Ask Signal Unipolar Modulatoy slut) NRZ J Coxtiey PeCosamfict © scanned with OKEN Scanner when ‘input bit is't' s(t)= Ix AcwsRMfet = Acasa © Wohen bit is'o! Slt) = OXAcosSaMfct = 0 Hence ASK signal Con be Hepnesented Ob Accosatifet do biti! Slt) = Ey! 5 qo bit 'o Gihevent detechon (0x) demodulation 0) ASK ? phose o the Cannien OF the Hecetven fs Same as phase of the Camtiet at ttansmitten Ask signal Tabeguato Thues hold Digital fide device “Agra 3 Actos aiife t “The input ASK Signal is applied to muttipliey . The multiple Oulput-is Recs alifet X Ae Cos ATfet q = Ac Cos” aiife t u AE (It Cosutfety See a = othen input to the matkpliey is Q Sinusoidal Caynien which ts generated with the help of Jocol oscillates > The output 4 muttpliey [product modlulaton is given to Inkegnato > The Integrate operates on rnulkpley outpuk fon Buccessive. bite _ {nenvals (Tp) and pexprns dow pos filtering ackon © scanned with OKEN Scanner © Th Tb Titepaio outpt = & ose dt + [BE suet dt oO ° => The output 9 inenralor gees fo thueshold device. Th Gmpaves the intennaton output with predained thneshold value J Inkegualox o]p > “yeshold value i generates Ambo ‘1? qy Inkegraton olp < thieshold value it generates Agbol 0 Non - CohexenE detection 1 BASK i OCCA OCS COON > Here Connien is not used ? 4 the Heceived igool is having frepencies othenthan fe then iF is ejected with noise > T Aiminate those noise Components We use Bond poss qilter(8PE) Lolth Conken frequency “f.' wahich is shown below ' y Ask ' vol Spal] BPP F—1> Rechper r LPR [1s threshold Se ( ' t T ' eyice, ( 1 © scanned with OKEN Scanner => The BPF attenuates othen (ue Components other than f2 and its output Is. qn to 0. neckyion © > Reckyien Govents PC to pulsoling D.c and LPF Smoothens Dc “The Combinalion 9} Heckyer and LPF is Called os ° Envelope. Deleclox ' “Thnoshold device : Tt the dC Value > thweshold value ‘I is produced qt the de Value < threshold Value “0” is produced Signal Space diagram a ASK: 6 using this ; probability gf Ero is detenmnined y} probity 4 Bron is mone. ie,‘ Pe! then Agsten gjgicinay is less ond vieeversa we have SLD = Rec@samfcr gon 1? (on) Va? Gosarfet [Oproer,p = che)” R Ie R= ax fe Cos atfet iy R=1 ee ee P= Pela = VE fe cosamtfet ti vm] b pe sw =f& am (CO %W= Je cosa | Sth =0 for! bit ‘0’ © scanned with OKEN Scanner ae TY the distance bekowen & Aymbole is rmoxe then probebitihy O] ension fs Jess then Ayptem Ajpictencyy 18 Move ) > he Argel Apace dhagearn owe Ask Is B,C) ° 5 u distonce = [&, => Hene distonce fe Less So Pe is moe $0 Aystem Agiciency 'S ess => Th Ask , the injormabion is Contained tn amplitude it ts costly Adected with noise Bondwrdth q Ask ; A ° ‘Tb fox one ectongular pulse the (reey domain epresertlion t's bos bf) o © scanned with OKEN Scanner

You might also like