MA6451 โ RANDOM PROCESSES PART-A 1.The mean and variance of the binomial distribution are 4 and 3 respectively. Find p(X=0) Solution: Let x be a binomial random variable with parameters n and p. P(X=x)=n๐ถ๐ฅ ๐ ๐ฅ ๐ ๐โ๐ฅ ๐ฅ = 0,1,2, โฆn Mean=np, variance=npq Given mean=4 and variance=3 np=4โ (1), ๐๐๐ = 3 โ (2) (2) ๐๐๐ 3 โ = (1) ๐๐ 4 3 3 1 ๐ = ,๐ = 1 โ ๐ = 1 โ = 4 4 4 4 4 n =๐ = 1 = 16 4
1 ๐ฅ
3 16โ๐ฅ,
P(X=x)=16๐ถ๐ฅ (4) (4)
๐ฅ = 0,1,2, โฆ 16
3 16
P(X=0)=๐16 = (4)
______________________________________________________________________________ 2.If the probability that a target isdestroyed on any one short is0.5.What is the probability that it would be destroyed on the 6th attempt Solution: Here destruction is success. First Success is on the 6th attempt. Let the random variable X denote the number of trails required for the first success. Given p=0.5 so q=0.5 and x=6 1
So, X follows geometric distribution with parameter p=0.5=2
Therefore the geometric distribution is P(X=x)=๐ ๐ฅโ1 ๐, ๐ฅ = 1,2, โฆ 1 ๐ฅโ1 1
=(2)
1 ๐ฅ
.2
=(2) , ๐ฅ = 1,2, โฆ 1 6
Therefore P(X=6) =(2)
=0.0156 ______________________________________________________________________________
3.The CDF of a continuous random variable is given by F(x)= {
๐,
๐<๐
Find the ๐ โ ๐ ,๐ โค ๐ < โ โ๐ ๐
PDF and mean of x Solution: Given the CDF of the Continuous random variable X is โ๐ฅ
F(x)={1 โ ๐ 5 , ๐ฅ โฅ 0 0, ๐ฅ < 0 The PDF of X is f(x)=๐น1 (๐ฅ) 1
={
โ๐ฅ
๐ 5 ,๐ฅ โฅ 0 5
0 , ๐๐กโ๐๐๐ค๐๐ ๐ โ
โ๐ฅ
Mean of X =E(X)=โซโโ ๐ฅ๐(๐ฅ)๐๐ฅ = โซ0 โ๐ฅ
1
= 5 [๐ฅ
๐5
โ1 5
โ๐ฅ 5
โ๐ฅ
โ 1.
๐5 (
=-[๐ (๐ฅ + 5)]
โ1 2 ) 5
โ
โ
โ๐ฅ
๐ 5 ๐๐ฅ 5
] 0
0
=-[๐ โโ โ ๐ 0 . 5] =5
______________________________________________________________________________ 4. Establish the memory less property of the exponential distribution If X is exponentially distributed with parameter ๏ฌ , then for any two positive integers โsโ and โtโP[X>s+t/X>s] = P[x>t] โ๏ฌ x ,x โฅ 0 Proof: The PDF of X is f(x)={ ๏ฌ e 0, ๐๐กโ๐๐๐ค๐๐ ๐
Now P[X>k] =โซ๐ ๏ฌ eโ ๏ฌ x โ
๏ฅ
๏ฉ e ๏ญ ๏ฌx ๏น = ๏ช ๏บ ๏ซ ๏ญ ๏ฌ ๏ปk
=eโ ๏ฌ k Consider for s,tโฅ0, P[X>s+t/X>s] = = =
๐[๐>๐ +๐กโฉ๐>๐ ] ๐[๐>๐ ] ๐[๐>๐ +๐ก] ๐[๐>๐ ] ๐ โ ๏ฌ (s+t) eโ ๏ฌ s
= eโ ๏ฌ t =P[X>t] Thus P[X>s+t/X>s] = P[x>t] โ ๐ , ๐ก โฅ 0 ______________________________________________________________________________ 5. If y=x2,where X is a Normal random variable with zero mean and variance ๐๐ ,find the PDF of therandom variable Y. โ๐ฅ2
1
Solution:Given fx(x)=๐โ2๐ ๐ 2๐2 โ โ < ๐ฅ < โ Fy(y) = P[Yโคy]=P[X2โคy] = p[โโ๐ฆ โค ๐ฅ โค โ๐ฆ ] =Fx(โ๐ฆ )- Fx(-โ๐ฆ ) fy(y)=2
1
โ๐ฆ 1
=2
[๐๐ฅ โ๐ฆ + ๐๐ฅ (โโ๐ฆ)] โ๐ฆ
1
โ๐ฆ
1
[ ๐ 2๐2 + ๐โ2๐ ๐ 2๐2 ] ๐ฆ ๐โ2๐
โ 1
โ๐ฆ
fy(y) =๐โ2๐๐ฆ . ๐ 2๐2 ,y>0 ______________________________________________________________________________ 6.Find the acute angle between two lines of regression. Solution : 1โ ๐ 2
If ๐ is the angle between two regression lines , then ๐ก๐๐๐ = (
๐
)
๐๐ฅ ๐๐ฆ ๐๐ฅ 2 + ๐๐ฆ 2
.
______________________________________________________________________________ 7.The joint p.df of the two dimensional random variable (X,Y) is given by 8๐ฅ๐ฆ
๐(๐ฅ, ๐ฆ) = {
(i) (ii)
9 1 โค ๐ฅ โค ๐ฆ โค 2. Find 0 , ๐๐กโ๐๐๐ค๐๐ ๐ Marginal densities of X and Y The conditional density functions ๐(๐ฅ/๐ฆ) and ๐(๐ฆ/๐ฅ)
Solution : (i)
The Marginal density functions of X and Y are given by 2
๐๐ (๐ฅ) = โซ๐ฅ ๐(๐ฅ, ๐ฆ)๐๐ฆ 2 8๐ฅ๐ฆ
โซ๐ฅ
9
๐๐ฆ
8 ๐ฅ๐ฆ 2
= โ 9
2
โ
2 ๐ฅ
= =
8๐ฅ
โ
9
4๐ฅ
4 โ ๐ฅ2 2
โ
โ4 โ ๐ฅ 2 โ ,
9
1โค๐ฅโค2
๐ฆ
๐๐ (๐ฆ) = โซ1 ๐(๐ฅ, ๐ฆ)๐๐ฅ ๐ฆ 8๐ฅ๐ฆ
โซ1
9
๐๐ฅ
8 ๐ฆ๐ฅ 2
= โ = = (ii)
โ
9
2
8๐ฆ
๐ฆ2
โ
9
4๐ฆ 9
๐ฆ
1 โ1
2
โ
(๐ฆ 2 โ 1) ,
1โค๐ฆโค2
The conditional density functions are given by ๐(๐ฅ/๐ฆ) = =
๐(๐ฅ,๐ฆ) ๐๐ (๐ฅ) 4๐ฅ 9
8๐ฅ๐ฆ 9
โ4โ ๐ฅ2 โ 2๐ฆ
1โค๐ฅโค๐ฆโค2
= (๐ฆ 2 โ 1) ๐(๐ฆ/๐ฅ) = =
๐(๐ฅ,๐ฆ) ๐๐ (๐ฆ)
8๐ฅ๐ฆ 9
4๐ฆ 2 (๐ฆ โ 1) 9
2๐ฅ
1โค๐ฅโค๐ฆโค2
= โ4โ ๐ฅ 2 โ
______________________________________________________________________________ 8.If X has mean 4 and variance 9 , while Y has mean -2 and variance 5 , and the two are independent, find (a) E(XY) (b) E(X๐ 2 ) Solution: Given E(X) = 4 , V(X) = 9, E(Y) = -2, V(Y) = 5 X and Y are independent โ ๐ธ(๐๐) = ๐ธ(๐). ๐ธ(๐) = 4 ร โ2 = E(X๐ 2 ) = ๐ธ(๐).E(๐ 2 ) =4ร9
8
= 36 ______________________________________________________________________________ 9.. Let X and Y be two discrete random variables with joint probability mass function ๐(๐ = 1
๐ฅ, ๐ = ๐ฆ) = {8
(๐ฅ + ๐ฆ), ๐ฅ = 1,2 ๐๐๐ ๐ฆ = 0,1 0 , ๐๐กโ๐๐๐ค๐๐ ๐
Find the marginal probability mass functions of X and Y.
Sol:
The PMF is X Y 1 2 P(y)
0
1 0.125 0.25 0.375
0.25 0.375 0.625
P(x) 0.375 0.625 1
______________________________________________________________________________ 10.Define Wide sense stationary process A random process {X(t)} is called Wide sense stationary if its mean is constant and autocorrelation function depends only on the time difference i.e (i) E(X(t)) is always a constant (ii)E(X(t)X(t+๐))= Rxx(๐) ______________________________________________________________________________ 11.If {X(t)} is a normal process with ยต(t)=10 and C(t1,t2)=16 ๐โ|๐๐ โ๐๐| find the variance of X(10)X(6) Solution: Since {X(t)} is a Normal process, any member of {X(t)} is a normal random variable By definition C(t1,t2)= R(t1,t2)-E[X(t1)].E[X(t2)] C(t1,t2)= Var(X(t1)) C(t1,t2)= Var {X(t)} Now X (10) is a normal random variable with mean ยต(10)=10 and variance C(10,10) =16 Var(U)= Var[X(10)-X(6)] =Var[X(10)]+Var[X(6)]-2Cov[X910,X(6)] =Cov(10,10)+Cov(6,6)-2Cov(10,6) =16 ๐ โ|10โ10| + 16๐ โ|6โ6| โ 2 ร 16๐ โ|10โ6| =16+16-2 16e-4
= 31.4139 ๐๐ข=โ31.4139 =5.6048 ______________________________________________________________________________ 12.Prove that a first order stationary process has a constant mean Solution: Let X(t) be a first orderStationary process Therefore the first order density function is unaffected by shift in time origin fx(x,t)= fx(x,t+๐ฟ) for any โ (1) โ
Now E[X(t)]=โซโโ ๐ฅ๐๐ฅ (๐ฅ, ๐ก)๐๐ฅ โ
And E[X(t+ ๐ฟ)] = โซโโ ๐ฅ๐๐ฅ (๐ฅ, ๐ก + ๐ฟ)๐๐ฅ โ
=โซโโ ๐ฅ๐๐ฅ (๐ฅ, ๐ก)๐๐ฅ = ๐ธ[๐(๐ก)]๐ข๐ ๐๐๐ (1) Therefore E[X(t)]=E[X(t+ ๐ฟ)] ๐๐๐ ๐๐๐ฆ ๐ฟ Hence E[X(t)] is aconstant. Hence mean is constant ______________________________________________________________________________ 13.State the postulates of a Poisson process If X(t) represents the number of occurrences of a certain event in (0,t) then the discrete random process {X(t)} is called the Poisson Process,Provided the following Posulates are satisfied. (i) P(1 Occurrence in (t,t+โ๐ก))= ๏ฌ โ๐ก + 0(โ๐ก) (ii)P (0 Occurrence in (t,t+โ๐ก))=1- ๏ฌ โ๐ก + 0(โ๐ก) (iii)P(2 or more Occurrence in (t,t+โ๐ก)) = 0(โ๐ก) (iv) X(t) is independent of the number of occurrences of the event in any interval before and after interval (0,t) ______________________________________________________________________________ 14.Define ergodic random process A random process X(t) is said to be ergodic if its ensemble averages are equal to its appropriate time averages
15. Find the autocorrelation function of a stationary process, whose power spectral densityfunction is 2 ๏ฌ ๏ฏ๏ท , ๏ท ๏ฃ 1 ๏ฏ ๏ฎ 0, ๏ท ๏พ 1.
given by S (๏ท ) ๏ฝ ๏ญ Sol:
R(๏ด ) ๏ฝ ๏ฝ ๏ฝ
1 2๏ฐ 1 2๏ฐ 1 2๏ฐ
๏ฅ
๏ฒ
S (๏ท ) ei๏ด๏ท d ๏ท ๏ฝ
๏ญ๏ฅ 1
๏ฒ๏ท
2
1
1 ๏ท 2 ei๏ด๏ท d๏ท ๏ฒ 2๏ฐ ๏ญ1
(cos ๏ท๏ด ๏ซ sin ๏ท๏ด ) d๏ท
๏ญ1 1
2 ๏ฒ ๏ท cos ๏ท๏ด d๏ท ๏ฝ
๏ญ1
1
1
๏ท ๏ฐ๏ฒ
2
cos ๏ท๏ด d๏ท
0
1 ๏ฆ ๏ท sin ๏ท๏ด 2๏ท cos ๏ท๏ด ๏ฝ ๏ง ๏ซ ๏ฐ๏จ ๏ด ๏ด2 1 ๏ฆ sin ๏ด 2 cos ๏ด 2sin ๏ด ๏ฝ ๏ง ๏ซ ๏ญ ๏ฐ๏จ ๏ด ๏ด2 ๏ด3 2
1
2sin ๏ท๏ด ๏ถ ๏ญ ๏ท ๏ด 3 ๏ธ0 ๏ถ ๏ท. ๏ธ
_____________________________________________________________________________________ 15. Given that the autocorrelation function for a stationary ergodic process with no periodic components is Rxx ๏จ๏ด ๏ฉ ๏ฝ 25 ๏ซ
4 . Find the mean and the variance of the process ๏ป X (t )๏ฝ . 1 ๏ซ 6๏ด 2
Sol: By the property of autocorrelation function,
๏ญ x2 ๏ฝ lim Rxx (๏ด ) ๏ด ๏ฎ๏ฅ
๏ฝ 25
๏
๏ญx ๏ฝ 5 . E ๏ป X 2 (t )๏ฝ ๏ฝ Rxx (0) ๏ฝ 25 ๏ซ 4 ๏ฝ 29
๏
Var ๏ป X (t )๏ฝ ๏ฝ E ๏ป X 2 (t )๏ฝ ๏ญ ๏จ E ๏ป X (t )๏ฝ๏ฉ ๏ฝ 29 ๏ญ 25 ๏ฝ 4. 2
_____________________________________________________________________________________ 16. Show that Spectral density function of a real random process is even function. Sol : By definition S (๏ท ) ๏ฝ
๏ฅ
๏ฒ R(๏ด ) e
๏ญ๏ฅ
๏ญ i๏ท๏ด
d๏ด
๏
S (๏ญ๏ท ) ๏ฝ
๏ฅ
๏ฒ R(๏ด ) e
i๏ท๏ด
d๏ด
๏ญ๏ฅ
Putting ๏ด ๏ฝ ๏ญu ,
S (๏ญ๏ท ) ๏ฝ
๏ฅ
๏ฒ R(๏ญu) e
๏ญ i๏ทu
du
๏ญ๏ฅ ๏ฅ
๏ฝ
๏ฒ R(u) e
๏ญ i๏ทu
du
[since R(๏ด ) is an even function]
๏ญ๏ฅ
๏ฝ S (๏ท ). Therefore, S (๏ท ) is an even function of ๏ท _____________________________________________________________________________________ 17. Define Cross Correlation function and state any two of its properties. Sol: The cross correlation of two processes X(t) and Y(t) is denoted by
RXY (t1 , t 2 ) ๏ฝ E[ X (t1 )Y (t 2 )] for any t1 and t 2 If X(t) and Y(t) are jointly WSS, then RXY (t , t ๏ซ ๏ด ) ๏ฝ R XY (๏ด ) Properties (i) RXY (๏ด ) ๏ฝ RYX (๏ญ๏ด ) (ii) R XY (๏ด ) ๏ฃ
R XX (0) RYY (0)
_____________________________________________________________________________________ 18. Define Cross-Spectral density. Sol: Let X(t) any Y(t) be two jointly stationary processes with cross correlation function R XX (๏ด ) . Then the cross power spectrum of X(t) and Y(t) is S XY (๏ท ) ๏ฝ
๏ฅ
๏ฒR
XY
(๏ด )e ๏ญi๏ท ๏ด d๏ด
๏ญ๏ฅ
_____________________________________________________________________________________ 19. Define a System. When is it called a linear system? Sol: A system is a functional relation between input x(t ) and output y (t ). A System y(t ) ๏ฝ f [ x(t )] given by the function f is said to be linear if for any two inputs x1 (t ) and x 2 (t ) the output
f [a1 x1 (t ) ๏ซ a2 x2 (t )] = a1 f [ x1 (t )] ๏ซ a2 f [ x2 (t )] for any constants a1 and a 2 . i.e., the system is linear if the principle of superposition holds good. Otherwise, the system is said to be non-linear.
_____________________________________________________________________________________
20. Define Band-Limited white noise. Sol: Noise having a non-zero and constant power spectral density over a finite frequency band and zero elsewhere is called band-limited white noise. ๏ The PSD of the band-limited white noise is given by
๏ฌ N0 ๏ฏ S NN (๏ท ) ๏ฝ ๏ญ 2 ๏ฏ ๏ฎ 0
, for ๏ท ๏ฃ WB , elsewhere
_____________________________________________________________________________________ 21. Define White noise process. Sol: A sample function X(t) of a WSS noise random process {X(t)} is called white noise if the power spectral density of {X(t)} is a constant at all frequencies. We denote the power spectral density of white noise w(t) as S w ( f ) ๏ฝ
N0 2
_____________________________________________________________________________________ 22. Define Linear time invariant system. Sol: A linear system given by y(t ) ๏ฝ f [ x(t )] is said to be time-invariant if y(t ๏ซ h) ๏ฝ f [ x(t ๏ซ h)] for any
h ๏ (๏ญ๏ฅ, ๏ฅ) i.e., any time shift h in the input results in the same shift in time of the output. _____________________________________________________________________________________
23. .State the Convolution form of the output of a linear time invariant system. Sol: If X (t ) is the input and h(t ) be the system weighting function and Y (t ) is the output, then ๏ฅ
Y (t ) ๏ฝ X (t ) * h(t ) ๏ฒ h(u ) X (t ๏ญ u )du ๏ฝ๏ฅ
_____________________________________________________________________________________ 24.
PART B 1. (A) The density function of random variable X is given by f(x)={
๐ฒ๐(๐ โ ๐), ๐ โค ๐ โค ๐ ๐, ๐๐๐๐๐๐๐๐๐
Find the value of K, the mean, variance and rth moment ๐พ๐ฅ(2 โ ๐ฅ), 0 โค ๐ฅ โค 2 Solution: Given the pdf of Continuous random variable X is f(x)={ 0, ๐๐๐ ๐๐คโ๐๐๐ โ
โซ f(x)dx = 1 โโ 2
โซ ๐๐ฅ(2 โ ๐ฅ)๐๐ฅ = 1 0 2
๐ โซ(2๐ฅ โ ๐ฅ 2 )๐๐ฅ = 1 0 2
2๐ฅ 2 ๐ฅ 3 ๐[ โ ] =1 2 3 0 23 ๐ [2 โ ] = 1 3 2
4๐ =1 3 3 =๐พ 4 3/4๐ฅ(2 โ ๐ฅ), 0 โค ๐ฅ โค 2 f(x)={ 0, ๐๐๐ ๐๐คโ๐๐๐ โ
๐๐1
= ๐ธ(๐
๐)
= โซ ๐ฅ ๐ ๐(๐ฅ)๐๐ฅ โโ
2
3 = โซ ๐ฅ ๐ ๐ฅ(2 โ ๐ฅ)๐๐ฅ 4 0
2
3
= 4 โซ0 (2๐ฅ ๐+1 โ ๐ฅ ๐+2 )๐๐ฅ 2
3 ๐ฅ ๐+2 ๐ฅ ๐+3 = [2 โ ] 4 ๐+2 ๐+3 0 2 2.2๐+2
= 3[ =
๐+2
โ
2๐+3 ๐+3
]
3 ๐+3 ๐ + 3 โ (๐ + 2) 2 [ ] (๐ + 2)(๐ + 3) 4
3 2๐+3 โ (1) 4 (๐ + 2)(๐ + 3)
๐๐1 = Substituting r=1, 2 in (1) ๐11
3 24 3 25 6 1 = = 1, ๐2 = = 4 3.4 4 4.5 5 6
1
Mean of X is E(X) =๐11 = 1 and Var(X)=๐21 โ (๐11 )2= 5 โ 1 = 5 1. (B) X is distributing normally, with mean 16 and standard deviation 3. Find(i)P(Xโฅ19),(ii)P(12.5
Put Z= (i)
๐
=
๐โ16 3
19โ16
When X=19, Z=
3
=1
P(Xโฅ19)=P(Zโฅ1) =05-P(0
P((12.5
When X=12.5,Z=
3 19โ16
When X=19 ,Z=
(iii)
3
= โ1.16
=1
Therefore P(12.5
When X=10,Z=
3 25โ16
When X=10,Z=
3
= โ2 =3
P(10X<25)=P(-21/2. Solution: (i)โ7๐ฅ=0 ๐(๐ฅ) = 1 K+2K+2K+3K+K2+2k2+7k2+K=1 10k2+9K-1=0
4 3k
5 K2
6 2k2
7 7k2+k
5 1 100
6 2 100
7 17 100
1
K=10 or K=-1 But since P(x) cannot de negative ,K=-1 is rejected 1
Hence K=10 X
0 0
1 2 1 2 P(x) 10 10 (ii) P(X<6)=P(X=0)+P(X=1)+โฆ.+P(X=5) 1
2
2
3
1
3 2 10
4 3 10
81
=10+10 + 10 + 10 + 100 = 100 81
19
P(Xโฅ6)=1-P(X<6)=1-100=100 P(0
(iii) P(Xโค3)= 2
8
8 10
1
P(Xโค4) =10 > 2 ______________________________________________________________________________ 3. If the p.d.f of a two dimensional random variable (X,Y) is given by ๐(๐ฅ, ๐ฆ) = ๐ฅ + ๐ฆ , 0 โค ๐ฅ, ๐ฆ โค 1. Find the p.d.f of U=XY. ๐ข
Solution: ๐ข = ๐ฅ๐ฆ , ๐ฃ = ๐ฆ. Hence ๐ฅ = ๐ฃ and ๐ฆ = ๐ฃ. ๐ฝ=
๐๐ฅ
๐๐ง
|๐๐ข ๐๐ฆ
๐๐ฃ | ๐๐ฆ
๐๐ข
๐๐ฃ
1 ๐ฃ = |โ๐ข ๐ฃ2
0 |= 1
1 ๐ฃ
The joint p.d.f. of ๐ข and ๐ฃ is given by ๐(๐ข, ๐ฃ) = ๐(๐ฅ, ๐ฆ)|๐ฝ| = (๐ฅ + ๐ฆ)
1 ๐ข 1 ๐ข ๐ข = ( + ๐ฃ) = 2 + 1 = 1 + 2 ๐ฃ ๐ฃ ๐ฃ ๐ฃ ๐ฃ
Since , 0 โค ๐ฆ โค 1 , 0 โค ๐ฃ โค 1, , 0 โค ๐ฅ โค 1 โ , 0 โค ๐ข โค ๐ฃ ๐ฃ varies from ๐ฃ = ๐ข to ๐ฃ = 1. Hence the p.d.f of U is given by โ
๐(๐ข) = โซ ๐(๐ข, ๐ฃ)๐๐ฃ 1
โโ
๐ข ] ๐๐ฃ ๐ฃ2 ๐ข ๐ข1 = [๐ฃ]1๐ข โ [ ] = 1 โ ๐ข โ ๐ข + 1 ๐ฃ ๐ข = โซ [1 +
๐(๐ข) = 2(1 โ ๐ข), 0 โค ๐ฅ โค 1. ______________________________________________________________________________
4.
The Probability distribution of the process{X(t)} is given by (๐๐)๐โ๐ , ๐ = ๐, ๐, โฆ ๐+๐ ๐ท(๐ฟ(๐) = ๐) (๐ + ๐๐) ๐๐ ๐=๐ { ๐ + ๐๐ , Show that {X(t)} is not stationary. Solution: Given
X(t)=n P{X(t)}=p(xn)
0
1
๐๐ก 1 + ๐๐ก
1 (1 + ๐๐ก)2
2
๐๐ก (1 + ๐๐ก)3
Mean E{X(t)}=โโ ๐=0 ๐๐(๐ฅ๐ ) ๐๐ก
1
(๐๐ก)2
๐๐ก
=0.1+๐๐ก +1.(1+๐๐ก)2 +2.(1+๐๐ก)3 + 3. (1+๐๐ก)4 + โฏ 1
๐๐ก
๐๐ก
2
=(1+๐๐ก)2 [1 + 2. 1+๐๐ก + 3. (1+๐๐ก) + โฏ ]
3
(๐๐ก)2 (1 + ๐๐ก)4
... ...
1
๐๐ก
= (1+๐๐ก)2 [1 โ 1+๐๐ก] 1
= (1+๐๐ก)2 ร [ 1
โ2
1+๐๐กโ๐๐ก โ2 1+๐๐ก 1
]
=(1+๐๐ก)2 ร (1+๐๐ก)โ2 1
= (1+๐๐ก)2 (1 + ๐๐ก)2 =1, which is a constant Now 2 E(X2(t)) =โโ ๐=0 ๐ ๐(๐ฅ๐ ) = โโ ๐=0[๐(๐ + 1) โ ๐]๐(๐ฅ๐ ) โ =โโ ๐=0 ๐(๐ + 1)๐(๐ฅ๐ ) โ โ๐=0 ๐๐(๐ฅ๐ ) 1
(๐๐ก)2
๐๐ก
={0 + 1.2 (1+๐๐ก)2 + 2.3 (1+๐๐ก)3 + 3.4 (1+๐๐ก)4 + โฏ }-E{X(t)} 2
๐๐ก
๐๐ก
2
=(1+๐๐ก)2 {1 + 3. (1+๐๐ก) + 6. (1+๐๐ก) + โฏ } โ 1 2
๐๐ก
=(1+๐๐ก)2 [1 โ 1+๐๐ก] =
2 (1+๐๐ก)2
โ3
โ1
. (1 + ๐๐ก)3 โ 1
=2(1+at)-1 = 2+2at-1 =1+2at Var{X(t)}= E(X2(t))-[E(x(t))]2 =1+2at-1 =2at Which is dependent on t {x(t)} is not stationary ______________________________________________________________________________ 5. (A) If the Wss process {X(t)} is given by X(t)=10cos(100t+๐ฝ),where ๐ฝ is uniformly distributed over (-ฯ,ฯ).Prove that {X(t)} is correlation ergodic Solution: We Know that Rxx(๐)= E(X(t).X(t+ ๐)) =E(100cos(100t+๐)cos(100t+100 ๐ + ๐)) =100E(cos(100t+๐+100t+100 ๐ + ๐)+cos(100t+๐-100t-100 ๐ โ ๐)) 100
=
2
E(cos(200t+2๐+100 ๐) +cos(-100 ๐))
50 E(cos(200t+2๐+100 ๐) +cos100 ๐) = 50cos100 ๐+50E(cos(200t+2๐+100 ๐)) โ (1) 1
๐
Now E(cos(200t+2๐+100 ๐)) =2๐ โซโ๐ cos(200t + 2๐ + 100 ๐)๐๐ 1
๐
= ๐ โซ0 cos(200t + 2๐ + 100 ๐)๐๐
๐ ๐๐(200t + 2๐ + 100 ๐) ๐ =[ ] 2 0 1
= 2๐ [ ๐ ๐๐(200t + 2๐ + 100 ๐)- ๐ ๐๐(200t + 100 ๐)] =0 Substituting in Equation (1),we have E(X(t).X(t+ ๐))=50cos100 ๐ โ (2) 1
๐
Therefore ZT= 2๐ โซโ๐ ๐(๐ก). ๐(๐ก + ๐)๐๐ก 1
๐
50
๐
25
๐
=2๐ โซโ๐ 100 cos(100๐ก + ๐) cos(100๐ก + 100๐ + ๐) ๐๐ก =2๐ โซโ๐[๐ถ๐๐๐ (200๐ก + 2๐ + 100๐) + cos(โ100๐) ๐๐ก = ๐ โซโ๐ cos(100๐) ๐๐ก + 25
๐
25 ๐
๐
โซ๐ ๐ถ๐๐ (200๐ก + 100๐ + 2๐)๐๐ก
=50cos(100 ๐)+ ๐ โซ๐ ๐ถ๐๐ (200๐ก + 100๐ + 2๐)๐๐ก 25
1
= 50cos(100 ๐)+ ๐ [200 (sin(200๐ก + 2๐ + 100๐) โ (๐ ๐๐200๐ก + 2๐))] 1
=50cos(100 ๐)+4๐ [sin(200๐ก + 2๐ + 100๐) โ (๐ ๐๐200๐ก + 2๐)] Now lim (๐๐ ) = 50cos(100 ๐) ๐โโ
=R(๐) Therefore {X(t)} is correlation ergodic ______________________________________________________________________________ 5. (B) If customers arrive at a counter in accordance with poisson process with a mean rate of 2 per minute, find the probability that the interval between 2 consecutive arrivals is (a)more than 1 minute (b) between 1 minute and 2 minute and (c) 4 minute or less Solution: Let T be the random variable denoting inter arrival time. By Property (3) fT(t)= ๏ฌe ๏ญ ๏ฌt and P(T>t)= e ๏ญ ๏ฌt Here ๏ฌ =2/minute Therefore fT(t)=2 e
๏ญ2 t
โ
(a) P(T>1)=โซ1 2๐ โ2๐ก ๐๐ก = 0.1353 2
(b) P(1โคTโค2)=โซ1 2๐ โ2๐ก ๐๐ก = 0.117 4
(c) P(Tโค4) =โซ0 2๐ โ2๐ก ๐๐ก = 0.99967
6. State and prove Wiener-Khintchine theorem Statement: Let X(t) be a real WSS process with power density spectrum S XX (๏ท ) . Let X T (t ) be a portion of the
๏ฌ X (t ) ,๏ญT ๏ผ t ๏ผ T X T (t ) ๏ฝ ๏ญ , elsewhere ๏ฎ 0 process X(t) in time interval โT to T. i.e.,
Let X T (๏ท ) be the Fourier
transform of X T (t ) , then
S XX (๏ท ) ๏ฝ
lim
1
T ๏ฎ ๏ฅ 2T
๏ป
E X T (๏ท )
2
๏ฝ
Proof: Given X T (๏ท ) is the Fourier transform of X T (t )
๏ X T (๏ท ) ๏ฝ
๏ฅ
๏ฒX
T
(t )e ๏ญi๏ทt dt
๏ญ๏ฅ T
๏ฝ
๏ฒX
T
(t )e ๏ญi๏ทt dt
๏ญT T
๏ฝ
๏ฒ X (t )e
๏ญi๏ทt
dt
๏ญT
X T (๏ท ) ๏ฝ X T* (๏ท ) X T (๏ท ) [where * denotes complex conjugate] 2
T
T
๏ญT
๏ญT
i๏ทt ๏ญi๏ทt ๏ฒ X (t )e dt. ๏ฒ X (t )e dt
๏ฝ
[๏ X (t ) is real]
T T i๏ทt ๏ญ i๏ทt 2 dt ๏ฝ ๏ฒ X (t )e 1 dt . ๏ฒ X (t )e 1 1 2 2 ๏ญT ๏ญT T T
๏ฝ
๏ฒ ๏ฒ X (t ) X (t 1
2
)e ๏ญi๏ท ( t2 ๏ญt2 ) dt1dt 2
๏ญT ๏ญT
T T
๏
lim T ๏ฎ๏ฅ
๏
E X T (๏ท )
2
๏ ๏ฝ T ๏ฎ ๏ฅ 2T lim
1
๏ฒ ๏ฒ X (t ) X (t 1
2
)e ๏ญi๏ท ( t2 ๏ญt2 ) dt1dt 2
๏ญT ๏ญT
But E๏X ((t1 )(t 2 )๏ ๏ฝ RXX (t1 , t 2 ) if ๏ญ T ๏ผ t1 , t 2 ๏ผ T T T
๏
lim T ๏ฎ๏ฅ
๏
E X T (๏ท )
2
๏ ๏ฝ T ๏ฎ ๏ฅ 2T lim
1
๏ฒ ๏ฒR
๏ญT ๏ญT
XX
(t1 , t 2 )e ๏ญi๏ท ( t2 ๏ญt2 ) dt1dt 2
We shall now make a change of variables as below Put t1 ๏ฝ t and t 2 ๏ญ t1 ๏ฝ ๏ด ๏ t 2 ๏ฝ ๏ด ๏ซ t ๏ the jacobian of transformation is
๏ถt1 J ๏ฝ ๏ถt ๏ถt 2 ๏ถt
1 0 ๏ถt1 ๏ฝ ๏ฝ1 1 1 ๏ถ๏ด ๏ถt 2 ๏ถ๏ด
๏ dt1dt 2 ๏ฝ J dtd๏ด The limits of t and โT and T When t 2 ๏ฝ ๏ญT ,๏ด ๏ฝ ๏ญT ๏ญ t and t 2 ๏ฝ T ,๏ด ๏ฝ T ๏ญ t
๏
lim
1
T ๏ฎ ๏ฅ 2T
๏
E X T (๏ท )
2
1 ๏ ๏ฝ T lim ๏ฎ ๏ฅ 2T
T ๏ญt
T
๏ฒ ๏ฒR
XX
(t , t ๏ซ ๏ด )e ๏ญi๏ท ๏ด dtd๏ด
XX
(๏ด )e ๏ญi๏ท ๏ด dtd๏ด
๏ญT ๏ญt ๏ญT
Since X(t) is WSS Process, RXX (t , t ๏ซ ๏ด ) ๏ฝ RXX (๏ด )
๏
lim
1
T ๏ฎ ๏ฅ 2T
๏
E X T (๏ท )
2
1 ๏ ๏ฝ T lim ๏ฎ ๏ฅ 2T
T ๏ญt
T
๏ฒ ๏ฒR
๏ญT ๏ญt ๏ญT
๏ฝ
๏ฝ
lim
1
T ๏ญt
T
๏ญT ๏ญt
๏ญT
T ๏ฎ ๏ฅ 2T
lim
๏ญi๏ท ๏ด ๏ฒ RXX (๏ด )e d๏ด . ๏ฒ dt
1
T ๏ญt
T ๏ฎ ๏ฅ 2T
๏ฒR
XX
(๏ด )e ๏ญi๏ท ๏ด d๏ด .2T
๏ญT ๏ญt
๏ฝ
lim T ๏ฎ๏ฅ
T ๏ญt
๏ฒR
XX
(๏ด )e ๏ญi๏ท ๏ด d๏ด
๏ญT ๏ญt
๏ฅ
๏ฝ
๏ฒR
XX
(๏ด )e ๏ญi๏ท ๏ด d๏ด ๏ฝ S XX (๏ท )
๏ญ๏ฅ
๏
๏ S XX (๏ท ) ๏ฝ lim E X T (๏ท ) T ๏ฎ๏ฅ 2T
2
๏
, by definition.
Hence proved. _____________________________________________________________________________________
7. If X(t) is the input voltage to a circuit (system) and Y(t) is the output voltage.{X(t)} is a stationary process with ๏ญ x ๏ฝ 0 and Rxx (๏ด ) ๏ฝ e function is H (๏ท ) ๏ฝ
๏ญ๏ก ๏ด
. Find ๏ญ y , S yy and Ryy (๏ด ) , if the power transfer
R . R ๏ซ iL๏ท ๏ฅ
๏ฒ h(u) X (t ๏ญ u)du
Y (t ) ๏ฝ Sol: (i) We know that
๏ E๏Y (t )๏ ๏ฝ
๏ญ๏ฅ
๏ฅ
๏ฒ h(u) E[ X (t ๏ญ u)]du
๏ญ๏ฅ
Since X(t) is stationary with mean 0, E[ X (t )] ๏ฝ 0 for all t ๏ E[ X (t ๏ญ u)] ๏ฝ 0 ๏ E[Y (t )] ๏ฝ 0
S XX (๏ท ) ๏ฝ
๏ฅ
๏ฒR
XX
(๏ด )e ๏ญi๏ท ๏ด d๏ด
๏ญ๏ฅ
(ii)We know that
๏ฅ
๏ฝ
๏ฒe
๏ญ๏ก ๏ด
e ๏ญi๏ท ๏ด d๏ด
๏ญ๏ฅ
๏ฅ
0
๏ฝ ๏ฒ e๏ก๏ด e ๏ญi๏ท ๏ด d๏ด ๏ซ ๏ฒ e ๏ญ๏ก๏ด e ๏ญi๏ท ๏ด d๏ด ๏ญ๏ฅ
0 ๏ฅ
0
๏ฝ ๏ฒ e (๏ก ๏ญi๏ท )๏ด d๏ด ๏ซ ๏ฒ e (๏ก ๏ซi๏ท )๏ด d๏ด ๏ญ๏ฅ
0 0
๏ฅ
๏ฉ e (๏ก ๏ญi๏ท )๏ด ๏น ๏ฉ e ๏ญ (๏ก ๏ซi๏ท )๏ด ๏น ๏ฝ๏ช ๏ซ ๏บ ๏ช ๏บ ๏ซ ๏ก ๏ญ i๏ท ๏ป ๏ญ๏ฅ ๏ซ ๏ญ (๏ก ๏ซ i๏ท ) ๏ป 0 1 1 ๏ฝ [1 ๏ญ 0] ๏ญ [0 ๏ญ 1] ๏ก ๏ญ i๏ท ๏ก ๏ซ i๏ท 1 1 2๏ก ๏ก ๏ซ i๏ท ๏ซ ๏ก ๏ญ i๏ท ๏ฝ ๏ซ ๏ฝ 2 ๏ฝ ๏ก ๏ญ i๏ท ๏ก ๏ซ i๏ท (๏ก ๏ญ i๏ท )(๏ก ๏ซ i๏ท ) ๏ก ๏ซ ๏ท 2 Given H (๏ท ) ๏ฝ
R . R ๏ซ iL๏ท
We know that S yy (๏ท ) ๏ฝ H (๏ท ) S xx (๏ท ) 2
2
R2 2๏ก ๏ฝ 2 R ๏ซ iL ๏ท ๏ก ๏ซ ๏ท 2
๏ฝ
R2 2๏ก . 2 2 2 2 R ๏ซ L ๏ท ๏ก ๏ซ๏ท2
(iii) The autocorrelation function of Y(t) is
RYY (๏ด ) ๏ฝ
1 2๏ฐ
1 ๏ฝ 2๏ฐ ๏ฝ
๏ฅ
๏ฒS
YY
(๏ท )e i๏ท ๏ด d๏ท
๏ญ๏ฅ ๏ฅ
2๏กR 2 e i๏ท ๏ด ๏ฒ 2 2 2 . 2 2 d๏ท ๏ญ๏ฅ ( R ๏ซ L ๏ท ) ๏ก ๏ซ ๏ท
๏กR 2 ๏ฐ
๏ฅ
1 e i๏ท ๏ด . ๏ฒ๏ญ๏ฅ ( R 2 ๏ซ L2๏ท 2 ) ๏ก 2 ๏ซ ๏ท 2 d๏ท
1 2 2 First we shall write ( R ๏ซ L ๏ท )(๏ก ๏ซ ๏ท ) as partial fraction, treating 2
2
2
ฯ2 as u. We shall write the special partial fraction as
1 A B ๏ฝ 2 ๏ซ 2 2 2 2 2 ( R ๏ซ L ๏ท )(๏ก ๏ซ ๏ท ) R ๏ซ L ๏ท ๏ก ๏ซ๏ท2 2
2
2
R2 L2 we get A ๏ฝ L2 ๏ก 2 L2 ๏ญ R 2 1 Put u ๏ฝ ๏ญ๏ก 2 we get B ๏ฝ 2 R ๏ญ L2๏ก 2 L2 1 2 2 2 2 1 ๏ก L ๏ญ R ๏ซ R ๏ญ L2๏ก 2 ๏ 2 ๏ฝ ( R ๏ซ L2๏ท 2 )(๏ก 2 ๏ซ ๏ท 2 ) R 2 ๏ซ L2๏ท 2 ๏ก 2 ๏ซ๏ท2 Put u ๏ฝ ๏ญ
L2 1 1 1 ๏ซ 2 2 2 2 2 2 2 2 2 2 ๏ก L ๏ญR R ๏ซL๏ท R ๏ญ L ๏ก ๏ก ๏ซ๏ท2
๏ฝ
๏ฅ ๏ฅ ๏กR 2 L2 1 ๏กR 2 1 i๏ด๏ท ๏ RYY (๏ด ) ๏ฝ .e d๏ท ๏ญ .e i๏ด๏ท d๏ท 2 2 2 2 2 2 2 2 2 2 ๏ฒ ๏ฒ ๏ฐ (๏ก L ๏ญ R ) ๏ญ๏ฅ ( R ๏ซ L ๏ท ) ๏ฐ (๏ก L ๏ญ R ) ๏ญ๏ฅ๏ก ๏ซ ๏ท 2
๏ฝ
๏กR 2 L2 1 2 2 2 ๏ฐ (๏ก L ๏ญ R ) L2
๏ฅ
1
๏ฒ
๏ญ๏ฅ
2
(
R ๏ซ๏ท2) 2 L
By Contour integration, we know that ๏ฅ
e imz ๏ฐ ๏ญma ๏ฒ๏ญ๏ฅ z 2 ๏ซ a 2 dz ๏ฝ a e , m ๏พ 0
.e i๏ด๏ท d๏ท ๏ญ
๏ฅ
๏กR 2 ๏ฐ (๏ก L ๏ญ R 2
2
2
) ๏ฒ๏ก ๏ญ๏ฅ
2
1 .e i๏ด๏ท d๏ท 2 ๏ซ๏ท
๏ RYY (๏ด ) ๏ฝ
๏กR 2 L2 ๏ฐ ๏ญ๏ด ๏กR 2 ๏ฐ ๏ญ๏ก ๏ด e ๏ญ e 2 2 2 2 2 2 ๏ฐ (๏ก L ๏ญ R ) ๏ฆ R ๏ถ ๏ฐ (๏ก L ๏ญ R ) ๏ก ๏ง ๏ท ๏จL๏ธ
๏ฝ
๏ฝ
๏กLR ๏ก 2 L2 ๏ญ R 2
e
๏ญ๏ด
๏กLR R2 L2 (๏ก 2 ๏ญ 2 ) L
e
R2 ๏ฆR๏ถ ๏ญ๏ก ๏ด e ๏ง ๏ท๏ญ 2 2 2 L ๏ก L ๏ญ R ๏จ ๏ธ ๏ญ๏ด
๏ฆR๏ถ ๏ง ๏ท๏ญ ๏จL๏ธ
R2 R2 L2 (๏ก 2 ๏ญ 2 ) L
e
๏ญ๏ก ๏ด
2
๏ฆR๏ถ ๏ฆR๏ถ ๏ง ๏ท ๏ญ๏ก ๏ด ๏จ L ๏ธ e๏ญ ๏ด ๏ฆ R ๏ถ ๏ญ ๏จ L ๏ธ ๏ฝ e ๏ง ๏ท 2 2 ๏จL๏ธ ๏ฆR๏ถ ๏ฆR๏ถ ๏ก2 ๏ญ๏ง ๏ท ๏ก2 ๏ญ๏ง ๏ท ๏จL๏ธ ๏จL๏ธ
๏ก๏ง ๏ท
๏ฆR๏ถ ๏ง ๏ท ๏จL๏ธ ๏ฝ 2 ๏ฆR๏ถ 2 ๏ก ๏ญ๏ง ๏ท ๏จL๏ธ
๏ฉ ๏ญ๏ฆ๏ง R ๏ถ๏ท ๏ด ๏ฆ R ๏ถ ๏ญ๏ก ๏ด ๏ช๏กe ๏จ L ๏ธ ๏ญ ๏ง ๏ทe ๏จL๏ธ ๏ช๏ซ
๏น ๏บ ๏บ๏ป
_____________________________________________________________________________________