TD: Stochastic Processes
Adel Belouchrani
Ecole Nationale Polytechnique
Exercise 1.
Suppose that X1 , · · · , Xn are i.i.d Exponential random variables with parameter λ and let Mn =
max (X1 , · · · , Xn ). Show that Mn − ln(n)/λ →d V where
P (V ≤ x) = exp[− exp(−λx)]
for all x.
Exercise 2.
Let {Xn } be a sequence of random variables. Suppose that E (Xn ) → θ (where θ is finite) and
Var (Xn ) → 0. Show that Xn →p θ.
Exercise 3.
The pdf of a random variable X is p(x). A random variable Y is defined as
Y = aX + b, where a < 0. Determine the pdf of Y in terms of the pdf of X.
Exercise 4.
Suppose that X is a Gaussian random variable with zero mean and unit variance. Let Y =
aX 3 + b, with a > 0. Determine the pdf of Y .
Exercise 5.
Let Xr and Xi be statistically independent zero-mean Gaussian random variables with identical
variance. Show that a (rotational) transformation of the form
Yr + jYi = (Xr + jXi ) ejϕ results in another pair (Yr , Yi ) of Gaussian random variables that have
the same joint pdf as the pair (Xr , Xi ).
Exercise 6.
Assume that random processes x(t) and y(t) are individually and jointly stationary.
1. Determine the autocorrelation function of z(t) = x(t) + y(t).
2. Determine the autocorrelation function of z(t) when x(t) and y(t) are uncorrelated.
3. Determine the autocorrelation function of z(t) when x(t) and y(t) are uncorrelated and
have zero mean.
1
Exercise 7.
The autocorrelation function of a stochastic process X(t) is Rxx (τ ) = B2 N0 δ(τ ). Such a process
is called white noise. Suppose x(t) is the input to an ideal band pass filter having the following
frequency response characteristics
1 |f − fc | ≤ B2 and |f + fc | ≤ B2
|H(f )| =
0 elsewhere
Determine the total noise power of the output of the filter.
Exercise 8.
The covariance matrix of three random variables X1 , X2 and X3 is
µ11 0 µ13
0 µ22 0
µ31 0 µ33
The linear transformation, Y = AX is made where
1 0 0
A= 0 2 0
1 0 1
Determine the covariance matrix of Y.
Exercise 9.
Consider a band-limited zero-mean stationary stochastic process X(t) with power density spec-
trum (
1 |f | ≤ W
S(f ) =
0 |f | > W
X(t) is sampled at a rate fs = 1/T to yield a discrete time process X(n) = X(nT ).
• Determine the expression for the autocorrelation sequence of X(n).
• Determine the minimum value of T that results in a white (spectrally flat) sequence.
Exercise 10.
Consider a normal random vector x with components with zero mean that are mutually uncor-
related.
• Show that the covariance matrix Rx is diagonal.
• Show that the components of x are mutually independent.
Exercise 11.
Show that
E (X(t + τ ) − X(t))2 = 2 (Rxx (0) − Rxx (τ )) ,
x(t) ∈ R
