ELEN90054 Probability and Random Models

Practice Questions on Multivariate Probability Models

April 6, 2016

1. Let X and Y be RV’s with joint PDF


 kxy 0<y<x<1
fX,Y (x, y) = 
0 else

a. Find the value for the constant k so that the PDF is valid .
b. Find fY (y).
c. Find fX (x).
d. Find P [X ≥ 12 , Y ≥ 14 ].

2. The joint PDF of random variables X and Y has constant value 16 in two rectangular
regions in the x−y plane. These are {0 < x < 1, 0 < y < 2} and {2 < x < 4, 0 <
y < 2}. Outside of these regions, the joint PDF is zero.

a. Find FX,Y (x, y).

b. Find P [Y − X > 0].
c. Find fY (y).
d. Find fX (x).
e. Find the conditional PDF, fY |X (y| x). Be sure to specify where this conditional
PDF is defined.
f. Find the conditional PDF, fX|Y (x| y), again specifying where it is defined.

3. Consider random variables X and Y with joint pdf

(
2 exp−(x+y) , 0 ≤ x ≤ y < ∞
fX,Y (x, y) =
0, else

a. Draw the x-y plane and indicate on it the region where fX,Y (x, y) is nonzero.
b. Find the marginal pdf’s of X and Y . Are they independent?
c. Define Z = Y − 3X. Find P [Z > 0].
d. Find E[Z | Y = 0].
e. Define W = X − Y . Find P [W > 0].

1
4. Let X and Y be RV’s with joint CDF

1 2


 4
y x 0 ≤ y < 2, 0 ≤ x < 1


x y ≥ 2, 0 ≤ x < 1






1 2
FX,Y (x, y) =  4
y x ≥ 1, 0 ≤ y < 2

x ≥ 1, y ≥ 2




 1


0 x < 0 or y < 0



a. Find the joint PDF of X and Y .

b. Find the marginal PDF of X.
c. Find the marginal PDF of Y .
d. Find P [X ≤ 12 , Y > 1]
e. Find P [X < Y ]

5. Let Z = X + Y where X and Y are independent random variables such that X is

exponentially distributed with parameter λ > 0, and Y is uniformly distributed on
the interval [0, 1].

a. Compute the cumulative distribution function FZ (z).

b. Compute the probability density function fZ (z).

6. Let the output of a communication channel be a vector (X, Y ) of two independent,

zero mean Gaussian RVs with the same variance. If a signal ‘0 is present then this
common variance is σ02 , but if a signal ‘1 is present then it increases to σ12 . Further
suppose that the probabilities of a 0 or 1 being transmitted are respectively p and
1 − p. Let the total (instantaneous) output power be

R2 = X 2 + Y 2 .

a. Find the conditional pdf’s of the output power given ‘0 and ‘1.
b. Hence obtain obtain the unconditional pdf.
c. Suppose that the receiver operates by picking a decision threshold T > 0 and
deciding that a ‘1 (0) was transmitted if the received power ≥ T (< T ). Find
an expression for the probability of error.
d. Find the value of T that minimises the error probability.

2
7. Let X, Y, Z be independent Gaussian random variables each with zero mean and vari-
ance 1. Define rv’s U, V, W as
U = X, V = X + Y, W = X + Y + Z.
a. The rv’s U , V and W are the result of an affine transformation
   
U X
 V  = A Y .
   

W Z
Specify the matrix A.
b. Use part (a) to find the parameters of the random vector pdf fU,V,W (u, v, w).
8. Let X and Y be correlated random variables with means µX = 2 and µY = 5, and
2
variances σX = 9 and σY2 = 4. Let the covariance between them be Cov(X, Y ) = 5.
Define the random variables:
U = 2X + 1.5Y + 1 V = 2X − 2.5Y
a. The rv’s U and V are the result of an affine transformation
" # " #
U X
=A + b.
V Y
Specify the matrix A and the vector b.
b. Find the means and variances of U and V , and the correlation coefficient ρU,V
by using part a).
9. X and Y are zero-mean jointly Gaussian random variables each with variance σ 2 . The
correlation coefficient between X and Y is given to be ρ.
a. Write down the joint probability density function of X and Y .
b. Random variables Z and W are given by:
Z = X cos(θ) + Y sin(θ)
W = −X sin(θ) + Y cos(θ)
i. What are the means and variances of Z and W ?
iii. What is the covariance between Z and W ?
iv. Write the joint probability density function of Z and W .
" #
X1
10. Consider the Gaussian random vector X = . The mean vector and correlation
X2
matrix are given to be:
" # " #
1 5 0
E[X] = RX =
−2 0 10
(a) Find the covariance matrix KX .

Now consider the transformation

" # " #
Y1 2 1
Y= = X
Y2 3 7

(b) Find E[Y].

3
11. Let V1 and V2 be jointly Gaussian noise voltages with each with mean 0.2 and variance
1. The covariance between them is 0.5. Define the column-vector V = [V1 V2 ]T .

a. Write down KV , the covariance matrix of V (also written as cov(V)).

b. Suppose that V is transformed by a linear 2-port circuit to give the output vector
W = [W1 W2 ]T . Write the mean vector and covariance matrix of W given
the transformation is:
" # " #" #
W1 1 2 V1
W= =
W2 0 1 V2

c. Calculate the autocorrelation matrix of W.

d. Using your answers to (b), write the parameters of the probability density func-
tion of W.
e. Calculate KV,W , the cross-covariance matrix between V and W.
" #
X1
12. Consider the Gaussian random vector X = . The mean vector and covariance
X2
matrix are given to be:
" # " #
3 2 1
E[X] = KX =
−1 1 4

(a) Find the correlation matrix RX .

Now consider the transformation

" # " #
Y1 2 1
Y= = X
Y2 3 −1

(b) Find E[Y] and KY .

" #
W1
Now consider the zero-mean Gaussian random vector W = with covariance
W2
matrix: " #
3 −1
KW =
−1 2

The cross-covariance matrix between X and W is given to be:

 
1 0.5
KX,W =  
−1 −1

(c) Define the new random vector Z = X + W . Find E[Z] and KZ .

13. Bob loses every pen he uses after T weeks, where T is an exponential r.v. with mean
1 week. If the semester is 12 weeks long, how many pens should Bob buy at the
beginning of the semester so that the probability that he runs out of pens before
the semester finishes is no greater than 0.01?

4
14. A binary transmission channel introduces bit errors with probability 0.15. Estimate
the probability that there are no more than 20 errors in 100 bit transmissions.

15. There are 100 students in a lecture theatre. If the probability that any student wears
a blue top is 0.25, find an approximate probability that at least 30 students wear a
blue shirt to the lecture.

End of Questions