The sum of the probabilities over all possible values of the random
variable N must be one. Thus we calculate as follows.
1 1 1 1
c 0 + 1 + 2 + 3 =1 (2)
2 2 2 2
So c = 8/15.
8 1 12 4
P [N < 2] = P [N = 0] + P [N = 1] = 1+ = = (4)
15 2 15 5
ECE302 Spring 2006 Exam 1 February 27, 2006 2
2. Let the random variable C represent the amount the telephone company charges for calls
to directory assistance. C is a function of the number R of phone numbers requested.
Given that
$0.50 R ∈ {1, 2}
C= $0.75 R = 3 (5)
$1.00 R = 4
And
0.6 R=1
0.2 R=2
PR (r) = 0.1 R=3 (6)
0.1 R=4
0otherwise
X
E[R] = rPR (r) = 0.6(1) + 0.2(2) + 0.1(3) + 0.1(4) = 1.7 (7)
r∈SR
(c) For calls that request more than one number, what is the conditional PMF PR|R>1
of the number of numbers requested?
1/2 r = 2
PR|R>1 = 1/4 r ∈ {3, 4} (11)
0
otherwise
(d) Given that a call requests more than one number, what is the expected number of
requests E[R|R > 1]?
1 1 1
E[R|R > 1] = (2) + (3) + (4) = 2.75 (12)
2 4 4
ECE302 Spring 2006 Exam 1 February 27, 2006 3
q
σC = E[C 2 ] − (E[C])2 (14)
so we need to find E[C 2 ], which is
so
σC = 0.3563 − 0.3306 = 0.0257 (16)
ECE302 Spring 2006 Exam 1 February 27, 2006 4
3. You and a friend find a dime and a nickel. To decide who gets to keep them you flip the
coins. If a coin comes up heads you win it; tails your friend wins it.
(b) Let S be the amount of money you win. What is the PMF of S?
(
1/4 s ∈ {0, 0.05, 0.10, 0.15}
PS (s) = (18)
0 otherwise
1 1 1 1
E[S] = (0) + (0.05) + (0.10) (0.15) = $0.075 = 7.5 cents (19)
4 4 4 4
ECE302 Spring 2006 Exam 1 February 27, 2006 5
1/10
r=2
6/10 r=6
PR (r) = (21)
3/10 r = 10
0 otherwise
1 6 3
E[R] = (2) + (6) + (10) = 6.8 (24)
10 10 10
1/10 s=4
6/10 s = 36
PS (s) = (25)
3/10 s = 100
0 otherwise
1 6 3
E[S] = (4) + (36) + (100) = 52 (26)
10 10 10
ECE302 Spring 2006 Exam 1 February 27, 2006 6
5. A cellular phone call may be placed by a pedestrian or a driver. As the phone changes
location along with its owner, a call may requires no handoffs (H0), one handoff (H1) or
more than one handoff (H2) from base station to base station. We are given that the
probability that a call comes from someone in a vehicle is 0.8. The probability that a call
from a vehicle will not require a handoff is 0.5; the probability that it will require one
handoff is 0.2. The probability that a call from a pedestrian will not require a handoff is
0.4; the probability that it will require one handoff is 0.1. (Note that this is not a very
realistic model.)
(a) What are the probabilities that calls from pedestrians, respectively drivers, will re-
quire two handoffs.
(b) What is the probability that a call (of unknown origin) will not require a handoff?
(c) When there is no handoff, what is the probability that the caller is in a vehicle?
(a)
(b)
(c)
P [H0 |D]P [D] 0.5(0.8) 5
P [D|H0 ] = = = (31)
P [H0] 0.48 6
ECE302 Spring 2006 Exam 1 February 27, 2006 7
6. Suppose that a digit is selected at random from the set S = {−1, 0, 1, 2, 3, 4, 5} where the
probability of selecting n is
(
0.2 n ∈ {−1, 0, 1}
P [n] = (32)
0.1 n ∈ {2, 3, 4, 5}
(c) For the events Ai and Aj you have chosen (or in general if none exist), what is the
probability of {Ai ∪ Aj }?
7. Short Questions
(a) Suppose we are given that the PDF of the random variable X is fX (x) = e−bx + c for
x ∈ [0, 1] and zero elsewhere. What two properties of PDF’s would you use in order
to determine the acceptable values for the parameters b and c. (DON’T DO THE
MATH. Just tell me the two facts.)
Z ∞
fX (x) = 1 and fX (x) ≥ 0 ∀x (55)
−∞
(b) A source wishes to transmit data packets to a receiver over a radio link. The receiver
uses error detection to identify packets that have been corrupted by radio noise. When
a packet is received error-free, the receiver sends an acknowledgment (ACK) back to
the source. When the receiver gets a packet with errors, a negative acknowledgment
(NAK) message is sent back to the source. Each time the source receives a NAK, the
packet is retransmitted. We assume that each packet transmission is independently
corrupted by errors with probability q. Would you model the distribution of the
number of times the packet is transmitted by the source as Poisson (PX (x) below),
Exponential (fX (x) below), or (c) Neither? (Circle the correct answer below and
indicate the values of x for which the first option holds if you choose Poisson or
Exponential. If you choose neither, suggest an alternative.)
(
αx e−α
x∈
PX (x) = x!
0 otherwise
(
λe−λx x ∈
fX (x) =
0 otherwise
Neither
The point here was to realize that a random variable representing
the number of times something happens must be a discrete random
variable. I accepted either the Poisson or any other discrete
probability mass distribution that made sense. The ranges of the
values of x are x ∈ {0, 1, 2, . . .} for the Poisson distribution and x ∈
[0, ∞) for the Exponential.
(c) What are the mean and the variance of a standard Gaussian (Normal) random vari-
able?
A standard normal random variable has mean 0 and variance 1.
(d) How do you obtain a standard Gaussian (Normal) random variable from a nonstan-
dard one?
Let X be the nonstandard Gaussian random variable, µX be its mean,
2
and σX its variance. Then the random variable Z = (X − µX )/σX is
a standard normal random variable.
(e) The graph of the PDF of a Gaussian random variable X is symmetric about what
value?
the mean µX