ELE $09 EINTRODUCTIOW To OCESSES EN STRVETOR ¢ PREF STEVEN KAY TEXT Book AYN TYTIVE PROBABILITY ANS RANDOM PROCESSES vsSinG MATLAe, S.KAY C REQUIRED) 2) SCHRYUM'S CUTLINE OW PROBABILITY RAArom VARIABLES, Anb RAnpeort PROCESSES, A WSe C RECO anenvers OTHER REFERENCES ; YD PROBABILITY AND RANDett PROCESSES, W, DAVE PeRT (OUT oF PRINTS, Me GRAW Hitt, 19°70 2p PROBABILITY RNP RAM OOM PROCESSES A, LEOW-GARYA 199% Bao See PACE F% of TEXT TOPICS |} CHAPTERS so-1a, CHAPTER 1% (Sone SECTIONS) | CHAPTERS pS- 1% SELECTED PoRTIINS OF CHAPTERS 2 AWP >} * ConTinvevs RAND tH VARIABLES Coavl) EXPECTED VALUES MVLT/ PLE CRV Ine > PIAENSjwAR CRY - LAW OF LARSE WURBERS CEWTRAL Linsr THEOREM ° BAIS RANDON PROLCESFES (RID ‘STATIONARY RP . LINEAR SYSTEMS ANP RE . GAYS LIAN RP - POISSON RP x REAL" woRip APPL) CATIWS % CompuTeC Sinuartows (Vir MATES © THROUGHOUT COURSE PREREQUISITES Vi) NTH 45) 0R EQUYALEWT COURSE JW BASIC PROBA ALITY (CHAPTERS 1-7 wh TEXT) ay CALCeLYS BP LW ERR SYSTEMS AND TRANSL ORAS Y Liver parvb MATRIX ALGEBRA ERA DING | "AWRY EXAM -~ 20 Fe HOURLY EXAM — 30 To Fival EXAM ~- 49 To WEERLY Hone WorK ASSIGNED. HOMEWORK DYE ove WEEK AFTER Assicwen! \N2 GeAre ps ErrHeK pass (V) oR Fan CV-). mest ATTEMPT ALL fRobLens - To pass. (roo many ¥-'s > cower ceave! Bont were = Awaryti/cAL ELERCSES + NATLAG EXPERIMENTS NOTES : COfyes of TRANSPARENCIES AT WWw. ELE, UREJCoeRSES READ GBEForE Crass / OFFICE HovRS * TUES THYRE, FRI 9yjAM _ RELLEY Annex A/22 LON TAST WFO 8 FOI B7p- SFO SV) Yop - 76 rx- baa (4X) KAYO ELF. vet, EDV lenale) Zn mopveriow (wp) Lommen PROBLEM? SICWRL DETECTION D RPRRR ay TOMAR 3) COMM UNICAT) AS Ww CHEM) CALS, BOLO Ges ETrS x SECTION NUMBER fy TEXT,TRANS jT PULSE ANWR LobK Far Eco FRin “TaRcer” TRRASYr Os THIEL, NOILE CopéLer AS a PANDO PROCESS) SUSPECT AN EH HAE - ZS THERE one 2 ate REEVE ASH UME WE SAMPLE KEP DO Xe, CHR Ww 2 How CAN WE MAKE A PEST fo JF XE OS Kw, CHUA yw oR KS We T wore @ SAM Foe ' SorvTow * Con PY TE SAN PLE _ MEAN 1 aw nN é KR Fe FoR Wojte oWLy ae ‘ oS FoR SICAL +M OSE wiy ° DEC)DE SIENMBL PRESENT IF Nn Svc > 0.26 = THRESHOLs gay fs nvZs qHis A Govop xerecroe % [ Tame eee bbs For Ayrwer |) CAN VIE A COMPUTER SIMULATION T° “ASSESS” PER FORRANCE py: i) C&NeERATE UsSiWl MATLAG W = /00 SAMPLES OF MO/ SE 2 Conpure sAnece nenw (2028 2) 2) REPEAT (1) BVT. Hen APD OF TO BAH sANALE 4 Compure Saris pea [>025 2) NIGHT WANT FO REPEAT £«¢PORMeENT SEVERAL TIAES - WAY -3 i 00 ar) 0 ao 0 gt Sample, é Sample, é (a) Noise only (b) Signal plus noise Figure 14.5: Received data samples Sigua is 9; = A = 0.5 and noise consists of TID standard Gaussian random vz& z & Sample mean on boise only i L el Z ? Nuraber of samples, NV” ” ‘Rumber of samples, N'? (a) Total view (b) Expanded view for 70 <.n value of sample mean versus the number data samples averaged. quests ove wd what wilr sra ple mea converge re convenses> % 2 HW LARSE..DES VN Havé re RE? vy How po Moyre STATISTICS AFFECT. RESvers ? D How ovoES AMPLITVDE OF . PULSE AFFECR ReEsv.TsS ? DP ALL THEE gvESTIWS CAN BC RNOWERED. BY STYPYIy és PRAnbdiMN PROCESS THEIRY [AND some DETECTION THEORY ¢ J (onCHAPTER 1/0 - CovwT RN eee RECALL THAT Ro RAWPOI7T VARY ABLE IS A Meepiné FRoit SAMPLE SPACE Te A NYATOER EXAMPLE : er a me (HA A Ivanne sSface la Le POSTIGLE ER PERIMENTAL Ovi bores) XfSed = © CPAs eS eb Pantort VRRIRGLE (eT pearey panooald (RV) NOTE = SLE of S..= 6 NABER OF ALVES OF X ® 6 THis 7S A PISeeeTeE RVD CovyThhE NYA BER OF VALVES, Cees HAVE X(SiPz CC oF 443 rae STint A pyscrsTe RY. For Tee PIE Exanere PI Xfi ec JFis FoR THIS EXAMPLE Coven AAve PU xb FEL ere ae, 43% 0 pes CPIxlsey > ej zr P Fr _ ft BENERGL DEweTe RV. BY CATAL LETTER X Awe Rive BY Lowen Case X, XPD = Ke 0 nae T3 Feveriew DEFiviTION Y= F(x) ), = (0,1) ¢ 10.1: Mapping of the outcome of a thrown dart to the reel line (example of inuous random variable). X JS PETANCE FRop Bute sEYE (cEVpERyY . How WAVY ovrlemes Ake Trine ? CAN WE tet PDK Sed = Ke} §~ pO FoR C= FB, ?Is ASS VUHE ALL cvrTrespes ARE EQVALLyY exery > Pleo sxe dl) = p [hex %y) = pltexe $44) Ere, ASIEN UWIF OR PLLA BILITIES OR Placxe4j= 4-4 ofa ezhes = LEverH oF Jw ERVAL Je oTH)S rencowane yp ovr copes Were PIS CRETE oR ae ® BG AO Maw Add EQVALLY L/MELY ° 036082 1 = (a) M-=10, Sr=01 (b) M= 20, Av = 0.05 Figure 10.2: Approximating the probability of an interval for a continuous random. variable by using a PMF. Pe lade PROBABILITY NAS FUN eTS Od Wore CPAFD yooh RECALL plasxeb)= TF pele) fe rasree a}Bur (ty = 1 be pyd as Dawe WE RECOVER ALL OV TLOMES NV Zhen be aexeens Zz (Xé, béva fog e* P20 9) =F ’ For OS K4/ ~ J tad > place cds " AS Bu wo. pro CALLED THE Phe BAbierry bewsiTy FUNCTI OW (PEF, be ot x TO Fidd PROBABILITS OF PARI LAnDive iM IATERV a (Bb) VST WT RATE PaoF oR 4 Plaexebl st J. Prim dx =i idee 4-4 OVR ORSEWAL ASSIENMEWT FoR EOVALLY LKERY ov reonest,am Wore + Px [eiy° ceweres Par va pelo DEW TES PRK THitk OF PF AS PRobARILITY Per UN{T LENE TR Stwle ASSUMIAS OT THe PREVJevL “DP/SCRETE ArrRox pATlow P xen ax 4X2 x. tAas) = Zo pal uid om = 2 palaesan (6+ ee Bt 4 Kee Ke Howe? fe: xe= no} Pan mace awe Xo = RAK = Px lates bow px (mod =P lwraeeXt xtra rr] Ax PPF FOR WAH pyle = eh ees ° © THER WEE CALLED A UNIFORM PDF. vse rand sw MATLAR TO SEN ERATE AN OUTCOTEAnoTHeR E%Ag PLE dl CAVERAL WE fer PReBABITIES By Plasxe bj = ST privex a 0 07 08 t a Figure 10.4: Nonuniform PDF. 1 Pe Rese CAW Aza % pxfesz=jzrt FoR Are > prix) =e Se patapan = fae Fal eet sy WHAT ABovT DISC TIN CITY AT Lo, pxlo-D = © | plod = pxfots=A ?I THIS PA PRoaLed ° WHAT 4f pl-esx se) as erot Z. PL(t) Ax = Se peta dy. = € fk (oy > oO RECARPLESS OF YALVE of pro). FeR CONTINVOVS RAV D6 VARIABLE Pl star, wrrvallJy+>o Aws PL X =x) = 0) (wipTy oF wreevAL=e). Peovtce, PLease b)= placx 24) = Ppl atextb) cPlacx2e]j LAN PORTANT fF S 4+ DD UNIFORN pulp bra 24K a ° 0 THERE a “ ado de sel Baap toe & 1 Maa} ” taper dont a Xv fab) othedenf oped moby yf PISS be buTe 2 A fe © CRED STAVOALD WeRmMAL PDF JM sIRTLAB rankn 6 ENRATES STANDARD VWORHWAL RV. NOTATION © X ww lu Or Mle tp = STANDARD WoRHAL PDFo HOW De WE Koro Se Px (upd =) PROOF : J. es eRe dx = x a SL. eka fog FF dy t RI by SI ¥ nan LAnALegovs re 2 & Zn Ci lary) one at aay " PROVE THIS IF UNFAAIL TAR LET M=N FS FeR EXAMPLE) TRANS FeRa TO fos AR crxoRDswarEr x= Ces 7 =r gone Va AT Wlep | [en ree) yf reg t Sof) os wire ING TF Cor ® = | r corte tr sire] = p> P Wey DILe 7> 2) (why wer r~-1?) Fp LAPLAQ AN - Pre ia = =e 0 £000 Pelt) = rG= SEE hook FoR FAM PLES Exaspce ) Five PL IK >t) pon tee PUixpar) = pl xrrexerys) = PIX N+Pelxes) way? ~ 2 Ply ry way? = mo eH EK =a / KC dx oo Se Fe oy isD “ft CAUCHY . io. reo 4 XL ow px OY * Tit gry * : wore: gs PAO-E) = pale) “ (EVEN) E ; : WHAT 0 THER. : PDFs ARE Figure 10.11; Cauchy PDP. Even? &) GAMA CAN MODEL WIDE RANGE 6F PDF sr FoR NoWWEEAT/YE Ree ~ _ wn) -a pel¥p~ Fo ee I ye rey > we Are «de (Te Wreérate To 1) os rian fo 47 e tae NOTAT] oof : PC2) CALLED Carts FUVerisa xw Pls adpx(z) rr er asa a=? Figure 10.12: Pxamples of Gamma PDF. PROPRTJES oF My. pp Pla) > @Phzy ay PI) = (170 22 PRI = SEE PRB. fo jb FROM PROPERTY J Peoop preys [et ede £er = win 9 ht > udu - bur r(k= A jaR tM ee : . ys ef ete Og [~ eR Kde nos Vs _ = Vea = VT Sway % Secs AL Cases oF Fle V) y TP DD 1 - pales = Det yore rip