Appendix A

Solved University Question Papers of Dec. 2022

Dec. 2022 Definition : If F ( y) is che jolnt distributlen function

of a
.1 Answer any four questions two-dimensional random variable Xand Y, then the
second order partial derivative of Fo (x y) is
o1a) Explain Bayes theorem and total probabilitytheorom. exlled the
probability density function of a two dimesional
random variable (% y) or stmply the joint
(5 Marks] function of (X, Y) and denoted by Fo desity
Ans. : Bayes theorem:
(ky)
Let A1, A2 . A, be mutually exclusive and exhaustive Properties of joint denaity tunction
eyent and Bis another event associated withA, then 1. Fy (x,y) 2 0
X y

P(A¡| B) = P(A) xP(B|A) 2

i=i P(A) XP(B|A)

i= 1,2 . . N
Total probability X 0

Let A1, Ag, .. An be mutually exclusive and 4. F,(0)= (uv) du dv

exhaustive event and B is another event associated with
S,then
n
Fy() =J y(u,v) du dv
P(B) = P(A) P(B | A) y2 2
i=1

Q.1(b) Define joint distribution function. What are its

properties? [5 Marks]
Ans. : 6. ,(0)= ty xy) dy
The joint distribution function is a function that
completely characterise the probability distribution of a F,() = y(y) dx
random vector.
Find the Binomial distribution it the mean is 4 and
Distribution function of a Two dimensional Q.1(c) [5 Marks)
variance is 3.
Random variable (X, ):The probability of the joint
event (X Sx,Y <y) which is a function x and y is called Ans. : The binomial distribution is given as
joint probability distribution P(X =x) ="C, P q
3, for Binomial
We have mean = 4 and variance =
Fy (x, y) = P(XSx, Ysy)
distribution, mean= np and Variance =npq
 Random Signal Analysls (MU)
A2
Appendlx -h
np a4 and npq 3
sq and
np

The binonmial distributlon is

characteristie function of a mndom variable Xwith uniform distribution in (- 1, 1) (5 Marka)

Q.t(d) Aind the
1,1| Is
Ans. : The randomvariable xwith uniform distribution in (-
J-1sxS1
0 otherwise
1 1 sin w W#0
W

When wa0 (w) E(e)=E[1]= 1

sin wW0
W
. (w) =
1 w= 0

9,(w)

Fig. 1-Q.1(d)
2. R() iseven function
autocorrelation function and
Q.1(e) List the properties of
prove any two properties. [5 Marks] R() = E[x(t) · E(t + )]
Ans. :Properties of auto correlation : R(-t) = E[X() -E(t - )]
1. The mean square value of random process can be t-t=t iLe. t=e'+T
obtained from the autocorrelation function R() R(-) =[x(e +t)x(]
By definition : Ro() = E[x() · E(t+)]
= Ex(e)-x(e +t))
t=0 Ro (0) = E[X()· x()]
= R(r)
3. R() is maximum at t= 0 |R(C)| SR (0)
mean square value of the
process
Techasulrdg
 Random Signal Analysis (MU) A-3 Appendix -A

,/R() Is the autocorrelation of astationary random plgetting headon an odd toss)

process (x())with no perlodlc component and with p(H in first toss) +p(H in 3rd toss)
non-zero mean then
+P(H in .)
lim
tR() [E()]2 = p+q9+ qq4p

020) Thojoint pdf of FR.V. X& Yis givon as

= p[1+q' +q'...]
fxy (%. y) mce'"e,0<y <X< *
0, olsewhere
Find 1-q=p
.c
. f() &f(y) Q.2(c) Show that P(AUB) = P(A) + P(B)- P(ANB)
[10 Marks] [4 Marks]
i. (ooy) &f(ybx)
Ans. : To find constant c so that fgy(x, y) represents a
Ans. : Consider

probability density function we must have B) = P(A) +P(B) - P(AN B)

P(A U
the union of two
We just express the event AUB as
SSry) dx dy = 1
00 exclusive event Aand An B

Butce-eY dx dy = 1
00
A B

scfekdx ferY dy = 1
0

AnB

:c=1 Fig. 1-Q.2(c)

AUB= AU(Ãn B)
Now, (1) G(69) =Ity (ey) dy =lo*edy
0 0
B) = P(AU(An B))
P(A U
=e-elo =ex
But the events of RHS are mutually exclusive,
Similarly (2) ,() = eY
P(A UB) = P(A) +P(A nB)
f(x, y)eeex
(3) fxly (%/y) =f) ey P(AU B) = P(A) + P(B) -P(An B)
Q.3(a) If X, Y are two independent exponential random
(4) f() eX variables with common parameter . Find the pdf of
for the first (U, V) where U = X+ Yand V= X- Y. Also findf(u)
Q.2(b) A biased coin tossed till a head appears
numbered and f (v). [10 Marks]
time. What is the probability that the
required tosses are odd? [6 Marks] Ans.: We have
lie , (x) = AeAx x>0
Ans. : Letp lie the probability of getting a head and q
toss so that
the probability of getting 0 tail in single , ) = Aey y>0
p+q=1 X Yindependent
Tech alely
 Appendix -A

Random Signal Analysis (MU) A -4 variance of Gaussian distribution

mean and
Q.3(b) Find pararmeters N(O, 1). [10 Marks)
fxy (% y) = ),) function with
Ans. :

u x+y vex-y f(3) =

Let
u+v 2x u-v=2y -o0<m<
x-ju9>y-tu-)
,X-m-z dx= rdz
du ov Now, puttíng

dy dy
du dv f(z) =
Hence p.d.f of (u, v) isgiven as
fz)dz
fuv (u, v) = |||Gy (x, y) M,(t) = E(e) = Je
2 = fez. dz
V2r
We find the marginal distribution of u by integrating
the above function w. r tov
<u-u<v<u
ie-z)
u= x+y, v=X-y |V|

Jfu, vu, v) dv 27-2tz = (z-t)?-2

f, (u) =
M(t) = jeie-o
dz -t = u dz = du
-se"=2ueu
ez
Simílarly, marginal distribution of v by integrating M(t) = du =
above function w.r. to u V2

u = xty’o and 0<|v|<u<m = 1+7t 2

Mean = Coefficient of t=0

Variance = Coefficient
22
5 Marks
Q4(a) Explain the central limit theorem.
Ans. :Central Limit Theorem
Variatewith
If x1, X2- - Xn are undependent random under
then
E(x) =H= and var()) =G, i= 1, 2, - n harmal
certain general condition Sn
variate with mean 4 and variance as n lends
infinity.
 Random Signal Analysís (MU) Appendix -A
A-5
sin (wt + )
f z the mean of sample size n, taken from the Q.4(c)Random Process is given as X () =
and w is
Where Y is uniformly distributed over (0, 2)
opulation having the mean u and varíance g L.e. a constant. Verify that X(t) is WSS
or not. [10 Marks]
Ans. :

E[x()] = E[sin (wt +y)]

sin Y
= E(sin wt cos Y + cos wt

var )-(a') - = sin(wt) E [cos Y]+cos(wt) E[sin

Y]
2T
GAD) Define SSS process and WSS process. [5 Marks] - sin ylg"=
E(cos Y) = J cos Ymdy =2 0
0
Ans.
2T
1 ylo"=0
Srict Sense Stationary Processes (SSS) :
E(sin Y) = Jsin Y dy =-lcos
A random process (x(t)} is said to be strongly
stationary process or strict sense stationary process, if E[x(t)] = 0 a constant

all íts final dímension dístributíon are invariant under Ryc () = E[x(t)· x(t + )]
translation of tíme parameter t. i.e. air invariant, if the
orlgin of tíme is changed. = E[sin (wt +)· sin {w(t + ) + Y]
Le. (xlt)} is called SSS if for all n and for every act of = E[(sin wt cos Y+ cos wt sin Y)
time Ihs tacts (t e T if i= 1, 2, ... n) -sin {w(t+) + cos Y+ cos w (t + ) sin Y}]
for order statíonary Random Process = [sin wt sin w(t + ) cos Y+ sin wt cos w(t +)
F,(%, t) = F(x, t+ e) and wt cos wt sin Y
+ cos wt sin w(t +) sin Y cos V + cos
2T 2T
F, (%, t) = F, (x)
M, (t) = E [x(t)] = M
E(cos? Y) = J cos Ydy =2n

independent
sin 0
Where m and o will be constant and 2T
of tíme. 1 sin 2y 1 cos 2y]
E(sin Y) = 2T dy = =0
WSS:Wide Sense Stationary Process:
stationary
A random process {x(t)} is called weakly Ryc (1) = sin wt sinw(t +) +,cos wt cos
if its mean is
process or wide sense stationary process,
constant and autocorrelation depends upon the time a function of z
difference Le. E[x(t)] = m, constant and E[x(t) x(t)]
Rugx (tË tz2) is function of (tz - t;) Hence the process x(t) is WSS process.
Let tË =t tz=t+T Q.5(a) The joint probablity distribution of Xand Yis given by
Thus, random process {x(t)) is WSS P(X = x, Y = y) = X+3y
24 where x =1, 2 and y= 1, 2,
Iffollowing two condition satisfied Find...
1. E[X)] ís constant i. Marginal distributions of x and y
ii. P(XS 2, Y s 1)
2. Ex(t) -x(t +)] is function of t.
ii. P(X< 1) [10 Marks)
 Random Signal Analysis (MU) A6
Appendix -A
Ans. :
Marginal probability
The joint probability
obtalning by putting x 1,distribution of (X, Y is 1

2 and y 1, 2 in the ,0) - Se-x-y) dx=-y 0<y<1

probability function P(x. y). 0
1
Y’ 1 2 Total Now.E(6) S*h(69-x)dx
1. 7 11
24 24 1

2. 5 3 1 13 E) = Jy yo) dy
24 24 3 24
1

Total 15 1
24

The marginal probability distribution

X P(y) 5
P(x)
= 43
1.
9 1 11
24 11
24
JJxy fy (%, y) dx dy
i) E(xy) = 00
15 2 13
24 24
11

P(xS2, ys1) = P(x= 1, y =1) = xy(2-x-y) dx dy

00
1 1

11 00
JS(2xy -x'y -xy') dx dy
P(xS1) = 24
variables (X, Y) has the dx
Q.5(b) Two dimensional random 2
following distribution
0sys1
fxy (x,y) = 2-x-y,0 Sxs1,
1
x3|1
0, elsewhere
0
-
Find
I. E(XY)
. Cov (X, Y) [10 Marks]

Ans. : i) cov(xy) = E(xy) - E(x) E(y)

density function x
The marginalprobability 1 5 5
6 1212
1

,() - Je-x-y) dy
0
-1
144
ifthe
Q.6(a) Prove that for a linear time invariant systern,
also WSS
input is a WSS process, then output is
2-x--2-x 0<x<1 Process.
[10Marks}
Technledg
 ndom Sgnal Analysis (MU) Appendix - A
Ans. :

Letthetime invariant liniar system be

YRO (o) x(t- u) du fh(ug) du, Jh(u2) duz E[X(E - u) X(t2 - u2)
EYO) h(u) E[x(t -u)] du x(t) is WSS, its autocorrelation is a function of
time difference Le. difference between t - u and tz -
U2
t) is WSs, E[x(t- u)] is constant. E[X(t - u) X(t2- uz)] = R [(ty -u) (2 - u2)]
Let E[x(t- u)] =
= R (t;,t)
EVO) = Jh(u) , du = R [(t1 - t2)-(uj- u2)l
Ry(t1tz) = J Jh(u) h(u2) R,[(t; - t)-(uj-u)]du du
- HJh(u) du (independent of t)
= Ry ()
Further,
Thus the mean of Y(t) is
the autocorrelation ofy(t) is given by independent oft and the
autocorrelation function of Y(O is function of time
Rt t2) = EY(t) -Y(t)] difference
Hence, Y(t) is also WSS.
Q.6(b) From the following data, obtain the twO
regression equations.
Sales 91 97
[10 Marks]
108 121 67 124 51 73 111 57
Purchases 71 75 69 97 70 91 39
Ans. : 61 80 47

In the given data

X
X-X (X- X? Y-Y
91
(Y-)² (x-X)(Y-)
1 1 71 1
1
97 7
49 75 5 25 35
108 18
324 65 -1 11 -18
121 31
961 97 27 729 837
67
-23 529 70 0 0
124 34
51
1156 91 21 441 714
-39 1521 39 -31 961
73 1209
-17 289 61 -9 81 153

Tech autedye
 Random Signal Anatysis (MU)
AppendixA
X
(x-) Y-Y (V-y (x-X)(V-)
111 21 441 80 10 100 210

S7 33 -23 S29 759

1089 47

Sx-900 Sx-6360 X70 ZY-)2868 3900

Mean Y 700
10 -= 70
The coefficient of
regression of XonY:
baxy = -00-) 3900 2868 1.36

The coefficient of
regression of Yon Xis,
byx = Yx-0(y -n_ 3900 6360 0.61

The equation of regression line of

sales on purchase (X on ) is given by,
(X- X) = bxy (Y - Y
(X-90) = 1.36x (Y- 70)
X- 90 = 1.36 Y-95.2
:.X = 1.36 Y- 5.2 is the
required line of regression of xon y.
The equation of regression line of
purchase on sales (Y on X) is given by,
(Y-) = byx (X - X)
(Y-70) = 0.61 (X-90)
Y- 70 = 0.61 X- 54.9
Y= 0.61 Y- 15.1 is the
required line of regression of yonx.