Scribd
Upload
50 views37 pages

07 Random Processes

Uploaded by

jc_juank
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
50 views37 pages

07 Random Processes

Uploaded by

jc_juank
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/ 37
RANDOM PROCESSES DUCTION 1s for random message signals and noise encountered in communication systems are developed in jer. Random signals cannot be explicitly described prior to their occurrence, and noises cannot be by deterministic functions of time. However, when observed over a long period, a random signal may exhibit certain regularities that can be described in terms of probabilities and statistical uch a model, in the form of a probabilistic description of a collection of functions of times, is indom proce: TONS AND NOTATIONS OF RANDOM PROCESSES + a random experiment with outcomes 2 and a sample space S. If to every outcome €$ we -valued time funetion X(1,), we create a random (or stochastic) process. A random process refore a function of two parameters, the time t and the outcome 2. For a specific 2, say, A, WE time function X(t, ;) = xi(). This time function is called a sample function. The totality of all ions is called an ensemble. For a specific time , X(t,,2) =X; denotes a random variable. ty) and fixed A= 4), Xt A) = 3) is a number. indom process is sometimes defined as a family of random variables indexed by the parameter is called the index set | illustrates the concepts of the sample space of the random experiment, outcomes of the sociated sample functions, and random variables resulting from taking two measurements of ctions. wing we use the notation X(t) to represent X(t, 2). S OF RANDOM PROCESSES Expressions: random process X(t). For a particular time 1,, X(t) =X; is a random variable, and its tion Fy(3f:) is defined as Feast) = PR) <4} (71) real number. 165 166 RANDOM PROCESSES (CHAP. 7 Sale space age t ‘ ih 4 wo bean Outcome y “ Asai) hook Fig. 7-1 Random process And Fy(x1:t1) is called the first-order distribution of X(). The comresponding firstorder density function is obtained by OF (xis) oa (7.2) Similarly, given 1 and t,, X(t,)=X, and X()=Xo represent two random variables. Their joint distribution is called the second-order distribution and is given by Fy, mith, ) = PIX) 0 2 Six(-@) = Sex) 3 ° 1 xl. Sxx(@) do = Ryx(0) = EXO) F. Cross Spectral Densities: (7.36) (737) (7.38) (7.39) (7.40) op Then 1p on neopsin pt al t= ost sy And Sy() and Ryy(s) of band-limited white noise are shown in Fig. 7-4 Note that the term white or band-limited white refers to the spectral shape of the process X(t) only, and these terms do not imply that the distribution associated with (1) is gaussian. Sage Rear) me ® Te se Oe o 7 7 © o Fig. 7-4 Band-limited white noise D. Narrowband Random Process: Suppose that X(f) is a WSS process with zero mean and its power spectral density Syy(0) is nonzero only in some narrow frequency band of width 2W that is very small compared to a center frequency 0, as shown in Fig. 7-5. Then the process X(t) is called a narrowband random process. In many communication systems, a narrowband process (or noise) is produced when white noise (or broadband noise) is passed through a narrowband linear filter. When a sample function of the narrowband process is viewed on an oscilloscope, the observed waveform appears as a sinusoid of random amplitude and phase, For this reason, the narrowband noise X() is conveniently represented by the expression X(t) = VQeos [o,¢+ 0) (7.66) CHAP. 7) RANDOM PROCESSES 175 Syxhod =e, 0 % “ bet ki Ww Ww Fig. 7-5 Narrowband random process where V(0) and (i) are random processes that are called the envelope function and phase function, respectively. Equation (7.66) can be rewritten as X() = Vieos $(0) cos «t= Vidsin 6(Osin «gt or) = X(N cos @.t — X(Nsin wt where X.(0) = V(a)cos (0) (in-phase component) (7.68a) X,() = V)sin $(¢) (quadrature component) (7.686) Vit) = yX2() + X20), (7.694) 0 = 1a EO (7.696) XO Equation (7.67) is called the quadrature representation of X(t). Given X(), its in-phase component X.() and quadrature component X,(t) can be obtained by using the arrangement shown in Fig. 7-6. ‘Low-pass , owe x xe xo 208 wt Low-pass x0 ‘iter H2sin et Fig. 7-6 Properties of X.(t) and X,(0: 1, X.() and X,(6) have identical power spectral densities related to that of X(7) by Sx(O- 0) +Sxlo+o,) lol SW $10) = 5x0) = 19 otherwise (7.70) 176 RANDOM PROCESSES [CHAP. 7 2. X.(r) and X,(f) have the same means and variances as X(t): dx, = Hs, = sx = 0 on) a, =o = 7.72) 3. X.(0) and X,(0) are uncorrelated with each other: EIX(OX,(0)] = 0 (7.73) If X(# is gaussian, then X,(#) and X,(0) are also gaussian. 5. 1FX()is gaussian, then for fixed but arbitrary t, V(t) is a random variable with Rayleigh distribution and (0 is a random variable uniformly distributed over [0, 2x}. (See Prob. 6.39.) Solved Problems RANDOM PROCESSES AND STATISTICS OF RANDOM PROCESSES 7.1. Consider a random process X(®) given by X() = Acos (ot + ®) (7.74) where A and o are constants and @ is a uniform random variable over [~, x]. Show that X(1) is WSS. From Eq, (6.105) (Prob. 6.23), we have 1 -ns0sn Jo =} 28 0 otherwise Thus, 1x00 = BIxen= | Avos (or-+ AYot®) a0 ar Aff coctarsaao=o 75) Reet, t +1) = ELX(OX(t+ 1) = FI cos tar Oeos xe + 9+ 0449 ac cot wt 7.76; 7 7.76) Since the mean of X() is a constant and the autocorrelation of X(t) is @ function of time difference only, we ‘conclude that X(1) is WSS. [Note that Ryy(2) is periodic with the period Tp = 2x/w. A WSS random process is called periodic if its autocorrelation is periodic. 4 eos ot +008 (a +20 + w)]d8 7.2. Consider a random process X() given by X(t) = Aeos (or + 6) 777) where @ and @ are constants and A is a random variable. Determine whether X() is WSS. Hap. 7) RANDOM PROCESSES im 73. 14, ix) = EIX()] = EtAcos (at +8) = cos (at + ETA] 78) Which indicates tht the mean of X(t) is not constant unless EA] =. Raxltt +) = BIXOX(+ 21 = BiAe0s (ot + eos folt +0) + ON Moos wt + cos (2ot +26 + we))EIA"] 7.79) ‘Thus, we see that the autocorrelation of X(1) is nota function of the time difference + only, and the process X() is not WSS, Consider a random process Y(t) defined by r= [xe id (7.80) where X(0) is given by X() = Acos ot 7st) where @ is constant and A = N[0; 67], (@) Determine the pdf of ¥() at t= ty. (b) Is ¥(t) WSS? @ Yay = f° Avon onds = 32% oan) ‘Then fom the esl of Pio 622, we see ta 4) gassan and variable with BY) = — EIA) =0 (7.83) ant vaaroo= (2220 ose Hence, by Eq. (6.91), the pdf of ¥(1,) is 1 Fro) = Tina, (7.85) (b) From Eqs. (7.83) and (7.84), the mean and variance of ¥(1) depend on time 1(j,), so Y(t) is not WSS. Consider a random process X(t) given by X() = Acos wt + Bsin cot (7.80) where « is constant and A and B are random variables. (@) Show that the condition EIA] = E(B] = 0 (787) is necessary for X(t) to be stationary. (Show that X() is WSS if and only if the random variables A and B are uncorrelated with equal variance, that is, EIAB]=0 (7.88) 178 RANDOM PROCESSES (CHAP.7 (7.89) and (@) x(t) = EIX()) = ElAlcos wt + B[B)sin ot must be independent of ¢ for X(@ to be stationary. This is possible only if y(t) = 0, that is, FIA] = E18) =0 ()_IEX(0 is WSS, then from Eg, (A717) aol=s[x(E)]=R0= ek su xo-A ant x(Z)=2 Tha, BA) = B18] =o} = Using the preceding result, we obtain Raat) = EKOXC +O) = EA c0s oF + B sin oA cos colt +2) +B sin ot +2) = o%c0s ot + EIABIsin Qa + 0x) 790) which will be a function of « only if EAB] = 0. Conversely, if [4B] = 0 and {4} = E(B") = o?, then from the result of part (a) and Eq, (7.90), we have px = 0 Ryxltst +1) = cos ot = Ry) Hence, X(0 is WSS. 7.8. A random process X(t) is said to be covariance-stationary if the covariance of X(t) depends only on the time difference =r — th, that is, Cyultst +2) = Cxx(@) oly Let X(t) be given by X() = (A+ Deos t+ Bsint where A and B are independent random variables for which, EA] = E(B] =0 and E[A?] = E(B Show that X(t) is not WSS, but it is covariance-stationary. y(®) = ELX()| = EVA + 1) cos 1 + Bsin ) 1A + Leos t+ £1B|sin 1 which depends on t. Thus, X() cannot be WSS. Rex(tists) = EIXH)K(D] = BUUA + 1084, + Bsin 4 ILA + 1) cos tg + Bint) = ALA +1)? Joos 008 fy + EIB") sin sin tp + BUA + 1)B](608 4 sin f+ sin 1, €08 1) CHAP. 7) RANDOM PROCESSES 179 bfca 41] = 21a? + 28 + 1) = 1A?) +2814} +1 = 2 EA + 1)B] = E{AB| + EB) = ELAJEIB] + E18] = 0 EIB] = 1 Substituting these values into the expression of Ryy( sta), we obtain Rants) = 2008 1 608 + in sin = €08 (fy — fh) + £08 1 608 fy From Eg, (7.9), we have Caen) = Realty sto)— Hat ata) £608 (tp — fy) + 608 11008 fz ~ C08 11608 fy = c0s (2-H) ‘Thus, X(0) is covarance-stationary 7.6. Show that the process X(@) defined in Eq. (7.74) (Prob. 7.1) is ergodic in both the mean and the autocorrelation. From Eq. (7.20), we have 1p 0) = fim 7] aos (ott A hie AP cos (wea dt = (7.92) where Tp = 2n/o. rom 72, wee Ry) = GtOxtt +9) ea ou fin} ses orton ft 4) 004 1 Fl nn leos or reoseant +20 + ont 4 = Leos wt 7. 5 (7.93) ‘Thus, we have Hxlt) = EIX()] = (x0) = ¥ Rugs) = EKOX+ 2)] = WOE) = Rx) Hence, by definitions (7.24) and (7.25), we conclude that X() is ergodic in both the mean and the autocorrelation. CORRELATIONS AND POWER SPECTRA 7.7. Show that if X(Q is WSS, then FXG +) -x0F 21Rrx(0) - Rix] 794) where Ryy(2) iS the autocorrelation of X(0). Using the linearity of & (the expectation operator) and Eqs. (7.15) and (7.17), we have 180 RANDOM PROCESSES (CHAP. 7 Axe + 9 -xoF] = Ae+ 9 -2K0+ Ox +H] = Hre+o]-2exe+oxorrafeo] , (0) — 2x0) + Raa) = 2Rrx(0) ~ Rex(e)] 7.8. Let X() be a WSS random process. Verify Eqs. (7.26) and (7.27), that is, @ Ryx(-0) = Rex lt) ® WRrx()] < Ryx(O) (@)_ From Eg. (7.15) Ryx(2) = EIXOXE +2) Setting +4 =¢, we have Reg(s) — EIME — 9X) = XOX = Ryx(-1) ® Axo = x0+07]=0 or Heo =mxoxero+Ke+9]=0 or Afro] + 2EKKOXE+ 9) + APU +O] 0 or 2Ryy(O) * Ryle) =0 Hence, Ry > Rex(Ol 7.9, Show that the power spectrum of a (real) random process X(0) is real, and verify Eq, (7.39), that is, Sue) = Syx(0) From Eq. (7.36) and by expanding the exponential, we have Selo) = f. Roeder Ryx(oKe0s ot ~jsin wr)de (795) FF Retoeosends=[" Re (osin ode Since Roy(t) = Rry() [Eq. (7.20)], the imaginary term in Eq, (7.95) then vanishes and we obtain sa | Ryx(t)c08 ot dt (7.96) which indicates that Sy(w) is real ‘Since the cosine is an even function of its arguments, that is cos (~c7t) = cos ar, it follows that Sexl-@) = Sxl) Which indicates that the power spectrum of X() is an even function of frequen. CHAP. 7) RANDOM PROCESSES 181 7.10, TAL. ‘A class of modulated random signal ¥(¢) is defined by ¥(0) = AX(Nc0s (w.t+ ©) (797) where X(j) is the random message signal and Acos (w fy. When fy~f, > Ty then f; and ¢> must fallin different pulse intervals [Fig. 7-10 (q)] and the random variables X(t,) and X(p) are therefore independent, We thus have Relist) = EIXCG XC = ELX(C EI) When ff, < Ty, then depending on the value of Ty, t, and fg may or may not be in the same pulse interval o (7.106) [Fig, 7-10 () and (0) Ife let B denote the random event "t, and are in adjacent pulse intervals,” then we have Rela) = ELK IX()IBIPB) + EXC )XE BIB) Now FUX(H )X(C)B) = EKG NEUE] EAX()X()BL =? Since PCB) will be the same when 1; and fy fll in any time range of length 7 it sulices to consider the case 0<1-< Th, as shown in Fig. 7-10 (0). From Fig. 7-10 (0), PB) = Ply < Ta < tp) =f fnttoats From Eq, (6.4), we have PB) =1~PB) Thus, Realty) =A(1 7.107) where t= ft Since Ryy(-t) = Riy(®), We conclude that . 184 RANDOM PROCESSES (CHAP. 7 , Jaw “| [— Ol ala th & ue ofa |e te 7 wit wop tt na Ie * Fig. 7-10 At it I=7, Ral = 7) RST 7.108) oO I> 7, which is ploted in Fig. 7-11(@. rom Eqs. (7.105) and (7.108), we see that X() is WSS. Thus, from Eq (7.36), the power spectrum of X() is Which is plotted in Fig. 7-110). Ratt) az See) = enf® coreap (7.109) ot /2 Seal) Fig. 7-11 CHAP. 7) RANDOM PROCESSES 185 7113. 714. Let X(2) and Y(#) be WSS random processes. Verify Eqs. (7.29) and (7.30), that is, @ Rey(-0) = Rint) O) Irv] < VR ORO (a) By Bq. 718) axOre—9l Seting 1 = #, we obtain Rey(-0) = BIKE + DYE)) = EOE +91 = Ral) (b) From the Cauchy-Schwarz inequality Eq, (6.148) (Prob. 6.52) it follows that {ELXOOVE + DIP = ELON + 91 or [Rar(oP = RexORy(O Thus, Wr $ VR ORO Let X(#) and Y(t) be both zero-mean and WSS random processes. Consider the random process Z(?) defined by 20) =XO+VO (7.110) (@ Determine the autocorrelation and the power spectrum of Z(t) if X¢) and Y(¢) are jointly Wss. (b) Repeat part (a) if Xi) and ¥() are orthogonal, (©) Show that if X(@) and ¥() are orthogonal, then the mean square of Z(0) is equal to the sum of the mean squares of X() and YC). (@) The autoconclation of 2(9 in given by Roath) = EZ) = BUX) + YOR) + VEIT = EKG )MG)) + FIXED + ELM) + EOI) = Rexistn) + Ravltyt) + Rallis) + Rnttit) ny If X() and ¥() are jointly WSS. then we have Real) = Rul) + Revo) + Rix) + Rr) guy where t= fi ‘Taking the Fourier transform of both sides of Eq. (7.112), we obtain Sx) = Sexo) + Sey) + S60) + Sx) 7.113) () IEX(H and NH are orthogonal (Bq. (7.34), Re) Rex) = 0 . 186 7.16. RANDOM PROCESSES (CHAP. 7 ‘Then Eqs. (7.112) and (2.113) become Rezl2) = Rex) + Riv) id) and Sz2(0) = Syxl0) + S(O) sy (©) From Bgs. (7.114) and (7.17), Rez{0) = Real) + Ry) or AZO) = XO] + EO) 7.116) Which indicates that the mean square of Z(t) is equal to the sum of the mean squares of X() and Y¢0) Two random processes X() and Y(0) are given by X() = Acos (ot + ©) (7.117) ¥(@) = Asin (wt +) 7.1176) where A and @ are constants and © is a uniform random variable over [0,2n]. Find the cross- correlation of X(2) and ¥(), and verify Eq, (7.29). From Bq, (7.18), the eross-correlation of (0) and ¥() is Rett +9) = EXON 9) = BIAc0s (ot + O)sin [att +2) +O] S blsin Qox tax +20)—sin (-00)] sin ot = Ren) 7.1180) Similarly, Roltse+ 9) = BWR + 9) = Bla?sin (at + @)cos [ole +) + O}) 2 4 Blain Qt +a +26) + sin (-02)) 7118) R & in ot = Rn) anya) = 4 snot = Ry) Which verifies Eg. (7.29) Let X(0) and ¥(#) be defined by X() = Acos wot + Bsin ot (7.119) Y(®) = Boos wr —Asin ot (7.1196) where « is constant and A and B are independent random variables both having zero mean and variance 2°, Find the eross-correlation of X(f) and Y(). ‘The cross-correlation of X(0) and Y() is CHAP. 71 RANDOM PROCESSES 187 Revtti te) = EIXH)Y EI(ACOS on, + Bin of (B008 oF, ~ Asin O,)] = EVAB)(008 on,c08 cory — sin cn,sin of) = BIA? Joos ox sin cory + ETB? sin corcos oty Since FAB) = EVAJEW] = 0. 1A) = E18" = we have Rarity sin or Cos ety — c0$ wt Sin Of) sin ot) =f) (7.120) where t= hh ‘TRANSMISSION OF RANDOM PROCESSES THROUGH LINEAR SYSTEMS. 7.17. A WSS random process X(*) is applied to the input of an LTI system with impulse response A(t) = 3e"*u(0). Find the mean value of the output ¥(2) of the system if E[X(®)] = 2. By Eq, (1.53, the frequency response H() of the system is H(o) = F(h)| = — ‘Then, by Eq, (7.50), the mean value of Y(0 is y(t) = EIVOL = pyH(O) 7.18. Let ¥(0)be the output of an LTI system with impulse response hi), when X() is applied as input. Show that Rarttntad= [ADRs ts Dab (7.121) ® Ryyltyt) -[ NOR es(6, —2, 2) dx (7.122) (@ Using Bq. (7.47), we have Reyltist) = EKG) = Axe |” snxe.- nan] =f mmercenxe— onde . = [mre Dae (Sinn, RylG,t) = BYE) = Af seoxe,—naonea| = jf NEEL ~ OYE 188 7.19. 720. RANDOM PROCESSES (CHAP. 7 = [remot and Let X(¢) be a WSS random input process to an LTI system with impulse response h(t), and let ¥(0) be the corresponding output process. Show that @ Ryy(0) = A) * Rex) (7.123) @) Ryyl(2) = h(t) # Ryy(@) . (7.124) © SO) = MOS) (7.123) @ Sy(@) = H"()S¢(@) (7.126) where * denotes the convolution and "(c) is the complex conjugate of H(@). (@ IE X( is WSS, then Eg. (7.122) of Prob. 7.18 becomes Realty) [ntatn=1-as on ‘which indicates that Ryy(ty, 42) isa function ofthe time difference = r, ~ 1, only. Hence, Eq. (7.127) yields Rar) = [ MPDRx(=— dP =H) * Rex) (B) Similarly, if XC) is WSS, then Bg, (7.122) becomes ed ee or Ry J WeRle +2) = Ha) + Rel) (©) Taking the Fourier transform of both sides of Eq. (7.123) and using Egs. (7.47) and (1.28), we obtain Se(@) = H@Sy(0) (q)_ Simitarly, taking the Fourier transform of both sides of Eq. (7.124) and using Eqs. (7.36), (1.28), and (1.21), ‘we obtain HOS) Note that by combining Eqs. (7.125) and (7.126), we obtain Eq, (7.53) that is, So) Si(@) = HHS, (0) = HO) ex) Let X() and ¥(@ be the wide-sense stationary random input process and random output process, respectively, of a quadrature phase-shifting filter (—n/2 rad phase shifter of Sec. 2.6). Show that (a) Raxlt) = Rw) (7.128) & Ray(o) = Rex) (7.129) where Ryx(1) is the Hilbert transform of Rxx(7). CHAP. 71 RANDOM PROCESSES 189 TU. 7.22. (a) The Hilbert transform X(t) of X(0) was defined in See, 2.6 asthe output of a quadrature phase-shifting filter with n= 4 He)=-jseno) Since |H()|? = 1, we conclude that if X() is a WSS random signal, then ¥(¢) = X(@) and by Eq. (7.53) Srl) = IH(@)PS(@) = Sexo) Hence, Ry) = FSO = F [Sy] = Ral) (®) Using Eqs. (7.123) and (2.26), we have Rey) =H) # Re) = 4 # Rete) = Rex A.WSS random process X(0) with autocorrelation Ryx(t) = Ae Where A and a are real positive constants, is applied to the input of an LTI system with impulse response he ult) where b is a real positive constant. Find the autocorrelation of the output ¥(¢) of the system, Using Eq, (J.53), we see that the frequency response F(a) of the system is 1 Mo) = FO = 2) So Wo? = Se ‘Using Eq, (1.55) we see thatthe power spectral density of X() is Selo) = FRO = As By Eq, (7.53), the power spectral density of Yi) is Sr(@) = HOPS yo) ~ (eral) _ a ( 2» )- A ( 2a ) . CIF FP) PP or +e ‘Taking the inverse Fourier transform of both sides of the above equation and using Eq. (J.55), we obtain pot se Verify Eq. (7.38); that is, the power spectrum of any WSS process X() is real and Sx 0 for every 0. 190 123. RANDOM PROCESSES CHAP. 7 ‘The realness of the power spectrum of X(j) was shown in Prob. 7.9. Consider an ideal bandpass filter with frequency response (Fig. 7-12) af! a (A) 7.140) Substituting Eqs. (7.144) und (7.146) into Eq, (7.58), we obtain Eq. (7.141). Let X; = X(t) and X, = X(t) be jointly gaussian random variables, each with a zero mean and a variance o*, Show that the joint bivariate gaussian density function is given by 1 [et] fix 2) 7.147) ex ee i=p | 2 @-p*) where p is the correlation coefficient of X; and X, given by p = Cyz/o? (Eq. (6.83) 2ni Sutiming Cy = Con nd Cia = C= po int Ba 7.8, we ave e-[2 @]-[) 4] det cP 4 f=) Since x= 0, x— =x, and roxsaiiely T]s] - a 2pxix, +33) ‘Substituting these results in Eq, (7.58), we obtain Bq, (7.147). ‘Note that Eq. (7.147) can be obtained from Eq, (6.107) (Prob, 6.28) by setting X = X,;¥ = Xa, ie = say = 0, and oy = oy = 0, ‘The relation between the input X(t) and output ¥() of a diode is expressed as Ye) = 2 (7.148) Let X(0 be a zero-mean stationary gaussian random process with autocorrelation Rex Find the output mean ip(#), the output autocorrelation Ryy(1), and the output power spectral density Sy). a>0 Hxl0) = EU) — BX} — RyxlO) 149) Ryrlty te) = ELMO VON = EPG 196 RANDOM PROCESSES (CHAP. 7 ‘Since X(¢,) and X(¢) are zero-mean jointly gaussian random variables, by the result of Prob. 6.71 eon] = EPP ey] e)] + ERO KOE (7.150) Since X() is stationary EX" (t,)] = EIX*()] = Rex) and ELK(0) X¢@)] = Rexltn— A) = Ria) Hence, Retin) = Rw) = ROP + AReeOP (7.151) and using Eqs. (.4/) and (1.29), we have So) = FUR) = Pala OP6(0) ++ SL) * Sx) (7.132) Now, forthe given input autocorrelation, by Eqs. (7.149) and (7.151), Hy = Ry) = 1 and Riya) — 1 4 2 By using Eqs. (747) and (1.55), the output power spectral density is 8a Siro) = FIR (0) = 225(0) ++ Sas 7.32. The input X(0) to the RC filter shown in Fig. 7-18 is a white noise process. (@) Determine the power spectrum of the output process Y(0). (6) Determine the autocorrelation and the mean-square value of ¥(1). it [| xo ck Yo | | Fig. 7-18 RC filter From Prob. 2.6 the frequency response of the RC filter is 1 HO) = TjoRe (@)_ From Eqs. (7.62) and (7.53) Sux(o) 14 Smo) = HOPS) = Re (7.153) CHAP. 7) RANDOM PROCESSES 197 (b)_ Rewriting Bq, (7.153) as 1 21/RO)) DERE + 1 ROE So) and using the Fourier transform pair Eq, (1.55), we obtain (7.154) Finally, from Eq. (7.154) (7.453) 7.33, The input X(0) to an ideal bandpass filter having the frequency response characteristic shown in Fig, 7-19 is a white noise process. Determine the total noise power at the output of the filter. me ” m4 = ° . Fig. 719 =! Sa(o) =4 Spo) = H(@)PSya) = 310)? ‘The total noise power atthe output of the filter is BP Ol=z) _Srodo=ZF[" lado (7.156) where B = Wp/(2x) (in Hz), 7.34, The equivalent noise bandwidth of a system is defined as 1 Jo IH(@)Pdeo Sdn HOM max |H(@)/ Hz (7.157) where |H(co){>, (@) Determine the equivalent noise bandwidth of the ideal bandpass filter shown in Fig, 7-19. (b) Determine the equivalent noise bandwidth of the low-pass RC filter shown in Fig. 7-18, and compare the result with the 3-dB bandwidth obtained in Prob. 2.9. (@) For the ideal bandpass filter shown in Fig. 7-19, we have max |(a)I? = 1 and 198 RANDOM PROCESSES (CHAP. 7 Wa, baad ° (odo BH 7.188 na = ag), Modo = 5. assy (D) For the low-pass RC filter shown in Fig. 7-18 we have 1 Woo! = oReF and max |H(@)P? = |H(O)P = 1 Thus a f£ do Lip? de Be Jo THORCF 222). TF wRCE ae HM (7.159) 1 1 Bus = 5 Wass = sen He M8 9g NM SRC Thus, Bag = Bap = 157B 39 7.35. A narrowband white noise X(0) has the power spectrum shown in Fig. 7-20(a). Represent X(t) in terms of quadrature components. Derive the power spectra of X,(t) and X,(t), and show that ELX2()] = BIX:(0] = EXO] (7.160) From Eq, (7.67) X() =X. eos wt ~X(O sin ot From Eq, (7.70) and Fig. 7-20 (a), we have Sea(0)=Syu(0)=f8 ISRO 2B = » oO Pz = ° * 7 By Sxuxle = Sux) “wow > w Fig. 7-20 CHAP. 7) RANDOM PROCESSES 199 736. 731. 7.238. 7.29. which is plotted in Fig. 7-20(b). From Fig. 7-20(a) awol=zfSuordo oy Joe From Fig. 7-20(6) soe! 1 austen eto ze) [ yom Zena) = 298 Hence, FO) X20] = ELC) nb (7.161) Supplementary Problems Consider a random process X(¢) defined by X() = cos OF where Q is a random variable uniformly distributed over [0, ap]. Determine whether X() is stationary. Ans, Nonstationary Hint: Examine specific sample functions of X(0) for different frequencies, say, Q = 1/2, x, and 2r. Consider the random process X(t) defined by X() = Acos ot Where © is a constant and A is a random variable uniformly distributed over (0, 1]. Find the autocorrelation and autocovariance of X(0) Ans, Ryxltsts) = feos 4008 Galt fa) = 4998 008 Let X( be a WSS random process with autocorrelation Ryle) = Ae Find the second moment of the random variable ¥ = X(5) ~ X(2). Ans, 2A(1 Let X() be a zero-mean WSS random process with autocorrelation Ryy(). A random variable ¥ is formed by integrating XC: 1 2, y [roe FFind the mean and variance of ¥. Lf aaca( Ans. jy = Oe) 200 RANDOM PROCESSES (CHAP. 7 740, A sample function of a random telegraph signal X(#) is shown in Fig. 7-21. This signal makes independent random shifts between two equally likely Valves, A and 0. ‘The number of shifts per unit tlme is governed by the Poisson distribution with parameter 2. | (a) Find the autocorrelation and the power spectrum of X(), (&) Find the rms value of X(9. x0 il Fig. 7-21 Ans) Regi) = 20 +e" Seco) ) A 2 7.41. Suppose that X(#) is a gaussian process with B= 2 Ryxle) = Se Find the probability that X(4) < 1 Ans. 0.159 7A2. The output of a filter is given by YQ =XC+T)-XG-1) where X(t) is a WSS process with power spectrum Syx(0o) and Tis a constant. Find the power spectrum of H(t). Ans, Sy) = 4sin 7@TSex(0) 743. Let X@) is the Hilbert transform of a WSS process X(0). Show that Ryx(O) ol Exes Hint, Use relation (b) of Prob. 7.20 and definition (2.26). TAS, When a metallic resistor Ris at temperature T; random electron motion produces @ noise voltage Vis) at tho open circuited terminals. This voltage V() is known as the thermal noise. Its povser spectrum Sjy(») is practically constant for f = 10"? Hz and is given by [ — | ; | Svvta | ig. 7-22 CHAP. 7) RANDOM PROCESSES 201 Syy(o) = 2KTR where k= Boltzmann constant = 1.37(10"%), joules per kelvin (J/K) T = absolute temperature, Kelvins (K) R= resistance, ohms () Calculate the thermal noise voltage (rms value) across the simple RC circuit shown in Fig. 7-22 with R = 1 kilohm (4M), C= 1 microfarad (uF), at T= 27° C. f stage eon

You might also like