This preview shows page 1. Sign up to view the full content.
Unformatted text preview: ECE 603  Probability and Random Processes, Fall 2010
Final Exam
December 14th, 10:30am12:30pm, Marston 132
Overview
The exam consists of ﬁve problems for 110 points. The points for each part of each problem are given
in brackets  you should spend your two hours accordingly. The exam is closed book, but you are allowed three pagesides of notes. Calculators are not allowed.
I will provide all necessary blank paper. Testmanship
Full credit will be given only to fully justiﬁed answers. Giving the steps along the way to the answer will not only earn full credit but also maximize the
partial credit should you stumble or get stuck. If you get stuck, attempt to neatly deﬁne your approach
to the problem and why you are stuck. If part of a problem depends on a previous part that you are unable to solve, explain the method for
doing the current part, and, if possible, give the answer in terms of the quantities of the previous part
that you are unable to obtain. Start each problem on a new page. Not only will this facilitate grading but also make it easier for you
to jump back and forth between problems. If you get to the end of the problem and realize that your answer must be wrong, be sure to write “this
must be wrong because . . . ” so that I will know you recognized such a fact.
Academic dishonesty will be dealt with harshly  the minimum penalty will be an “F” for the course. ©
A'
T©
U' ©
4E¥
j
u au ©W
1i¥ V h &
h
kg
pW j
A u ©W
ei¥ V u
E¥
5 &
5s uu
W
5 '
©
5 # '
© h h
ig sfHu u
W
sv
fHu u
W
a ©
$4E¥ c
d d
e e"`
`
# E¥
# E¥
©
4E¥ ,
©
4E¥ is the Fourier Transform of sinc 7 Y Y
b ¦`
' ©W
1¢' R
E¥ S q ©
4E¥ a ©W
$ei¥ V ¥£¡
y¤¢ 'ywu u
xs
W
xs v
'ywu u
W aY
b¦ Y` & ¥ 20
P31£
' ©W
1¢'
c
d
©
c Dirp4E¥
d sinc cg
S) `C
c
fd
Fe Parseval’s Relation: If s q © W
trp1i¥ h
# X¨¥ V
W7 ©
A@ Time Function Fourier Transform
A%9¨431£
© @ 8 7¥ 20 © @¥ 20 £ © 7¥ £ ¡ 8 © @¥ £ ¡ © 7¥ 20
AE431H¨¦¤¢GF¢ED¤¢C¨4B1£ © @ 8 7¥ £ ¡
A%9¨¦¤¢ © @¥ 20 £ © 7¥ 20 £ I © @¥ £ ¡ © 7¥ £ ¡
AE431H¨431GPAE'¤¢C¨D¤¢
QAE¦¤¢H¨¦¤¢ © @¥ £ ¡ © 7¥ £ ¡ 5 ¨¦¤¢ S
7¥ £ ¡ R # A@
©
QAE431H¨431£
© @¥ 20 £ © 7¥ 20 5 ¨¦¤¢ R
7¥ £ ¡ 5 A@
©
QAE¦¤¢H¨431£
© @¥ £ ¡ © 7¥ 20 5 ¨431£ S
7¥ 20 R # A@
© # ¨¦¤¢ 7¥ £ ¡ T©
UA@ # ¨¦¤¢ 7¥ £ ¡ T©
UA@ # ¨431£
7¥ 20 T©
UA@
) ! (&
¨¦¤¢ © §¥ £ ¡ )!& # !
¢"('%$" 5
6!
¨431£
© §¥ 20 Some potentially useful information have joint probability density function:
V
v
v v ¦ v
9s 5 ©q
¨
V © V
¦ ¥ £¡
§ S¥ ¤ V
¦ ¥ £¡
§ S¥ ¤¢
$© V
¦
§ ¥
© plane where 5 a [5] (a) Shade the region in the otherwise ¦ and 1. The random variables is nonzero. [5] (b) Find .
¨ , the marginal probability density function of
©V © V ¥
¡ ¡Y ¥
¦¥ , the conditional probability density function of given . For this, write
between constants and your limits on as (potentially) functions of .
a V ¦ u V [5] (d) Find
your limits for .
a [5] (c) Find 2. Consider the following experiment: I ﬂip a fair coin. If the coin comes up “heads”, I record the
outcome (zero) in my notebook; if the coin comes up “tails”, I generate a random number from a
uniform distribution on the interval
, and record the result as the outcome in my notebook. As
has been the standard from class, denote the outcome of the experiment .
T s s bR
5 , . .
, ﬁnd the probability that the coin came up “heads”.
'
# s
©
¥
$ %#"!
© Given Find for the experiment. (If you have a hard time ﬁnding
for partial credit.)
§ ¥ [10] (b) Let © [10] (a) Find a nontrivial space
give the probability density function of ¥ (&
' [10] (c) Let
. Determine whether the sequence
converges, and, if so, to what
and in what ways? Consider almost sure convergence, mean square convergence, convergence in
probability, and convergence in distribution. You can do them in any order that you want, and you can
use one type of convergence to imply another when appropriate.
'&
$ )¢! 3 ’s, each with mean #s 1
2a 3. For a ﬁxed , and independent and identically distributed sequence of
and variance , we know that, if
, then:
A 91 7
1
5a B@8 1
T 5a R 3
E0 4
0 1
T 5a R 4 9C
D1
A 9C 1
T 5a B1
A R 4 T R 0 1
T 5a R 6 4 4 Answer each of the parts below independently.
, but now
. 15a AB91 7 6
1
T Ga R 4 and variance
. Again, let s R 3
[5] (a) Suppose that the
are still identically distributed with
they are correlated such that
, for some
in terms of , , , and .
Find
(& 1 I
R 6
6© 1
Fa R
P H
a 1Qa ¥ ¢I¨ 6 3 T 0 R 4 [10] (b) Now, go back to assuming an independent and identically distributed sequence of ’s, each
with mean
and variance . But now suppose that
is also random. That is, suppose
1a 0 6 1
T Sa R 4 3 I have a nonnegative discrete random variable ; I generate
and then sum together the ﬁrst
elements of the sequence
. Call the result . Find the expected value of in terms of
,
the variance
of , and . (Hint: Write the joint probability density function of and
as
, and use the deﬁnition of the expectation of
to get started. You can get to the
answer in just a few steps if you head the right direction.)
T 0 0 4 0R 0 0 6
$1 a! 3 0 © with mean zero and autocorrelation function ¥ sfu u
sv
fu u ¥ ¡ ©W
1i¥ ¡ ©W
ei¥ ¥
5s uu §
¨¡ q© 7 B1
A9 ©¥ 96 ¥
e© 9 ¦¥ u Y ¥ 9 4. Consider a stationary Gaussian random process ¥£
¦¥ ¤¢
[10] (a) Consider the sequence
formed by averaging samples of
taken
one second apart. Determine whether the sequence
converges, and, if so, to what and in what
ways? Consider almost sure convergence, mean square convergence, convergence in probability, and
convergence in distribution. You can do them in any order that you want, and you can use one type of
convergence to imply another when appropriate. ¡ formed by averaging samples of ©W
1i¥ ©
'x §
¨¥ ¡
B1
A9 & 7
9 9& 9& Find the probability density function of &
$ 9 ¢! 9 [20] (b) Consider the sequence
onehalf second apart. taken . (Note: Starting from the probability density function of
is probably not the easiest way to
solve this part.) Determine whether the sequence
converges, and, if so, to what and in
what ways? Consider only mean square convergence, convergence in probability, and convergence in distribution. You can do them in any order that you want, and you can use one type of
convergence to imply another when appropriate.
©W
1i¥ ¡ & &
$ 9 ¢! 5. Suppose I have a zeromean stationary Gaussian random process
with autocorrelation function
. Suppose
is input to a linear timeinvariant ﬁlter with impulse response
. You measure the power at the output of the ﬁlter to be 3 milliWatts.
1i¥
©W
¨
¨ © ©W
1i¥
©W
ei¥ © © ¡ ©W
1i¥ © ¥
¦¥ ©ei¥
W
2
30 £ 30 £
2 ¡ ©W
1i¥ ©W
1i¥ ¡ ©W
ei¥ 30 £
2 [8] (b) Suppose
is input to a linear timeinvariant ﬁlter with impulse response
. Find the power at the output of the ﬁlter.
¤¨ [7] (a) Suppose
is input to a linear timeinvariant ﬁlter with impulse response
. Find the power at the output of the ﬁlter. © s 5 i¥ 30 £
W2 ...
View Full
Document
 Fall '10
 DennisGoeckel

Click to edit the document details