Scribd
Upload
105 views4 pages

Midterm 2011

This document provides instructions for a midterm exam in a communication systems course. The exam contains 4 problems worth a total of 80 points. It will take place on February 16th from 4-5:50pm. Students are allowed one sheet of reference notes and must show their work on separate sheets of paper. The problems cover topics like detection in noise, QAM signal constellations, autocorrelation and power spectral density of signals, and representing waveforms using different basis functions.

Uploaded by

Jose guiteerrz
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
105 views4 pages

Midterm 2011

This document provides instructions for a midterm exam in a communication systems course. The exam contains 4 problems worth a total of 80 points. It will take place on February 16th from 4-5:50pm. Students are allowed one sheet of reference notes and must show their work on separate sheets of paper. The problems cover topics like detection in noise, QAM signal constellations, autocorrelation and power spectral density of signals, and representing waveforms using different basis functions.

Uploaded by

Jose guiteerrz
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Communication Systems EE 132A

Prof. Suhas Diggavi

UCLA Winter quarter 2011/2012


Handout #24, Thursday, 16th February 2012

MIDTERM
Thursday, 16th February, 2012, 16:00-17:50
This exam has 4 problems and 80 points in total.

Instructions
You are allowed to use 1 sheet of paper for reference. No mobile phones or calculators are
allowed in the exam.
You can attempt the problems in any order as long as it is clear which problem is being
attempted and which solution to the problem you want us to grade.
If you are stuck in any part of a problem do not dwell on it, try to move on and attempt
it later.
Please solve every problem on separate paper sheets.
It is your responsibility to number the pages of your solutions and write on the first
sheet the total number of pages submitted.
Integrals:
R

eat ej2f t dt =

R0

eat ej2f t dt =

1
a+j2f
1
aj2f

Q-function Values:
Q(0) = 0.5
Q(0.5) = 0.31
Q(2/3) = 0.25
Q(1) = 0.16
Q(1.5) = 0.067
Q(2) = 0.023
Q(2.5) = 0.0062
Q(3) = 0.0013
Q(3.5) = 0.00023

Q(4) = 3.17e-5
Q(4.5) = 3.4e-6
Q(5) = 2.87e-7
Q(7.5) = 3.22e-14
Q(10) = 7.77e-24
Q(12.5) = 3.86e-36
Q(15) = 3.85e-51
Q(25) = 3.38e-138
Q() = 0

Good Luck!

Problem 1

(Detection in uniform noise (20 pts))

A one-dimensional additive noise channel, Y = X + Z, has uniform noise distribution (see also
Figure 1):
 1
L
L |z| 2
pZ (z) =
L
0 |z| > 2
where L2 is the maximum noise magnitude. The input X has binary antipodal constellation
with equally likely input values X = 1, i.e., if M = 0, X = 1 and if M = 1, X = 1; see also
Figure 2. The noise Z is independent of X.

Figure 1: Uniform noise probability density function

Figure 2: Signal constellation for transmission

(a) Design an optimum detector (showing decision regions is sufficient.)

[5pts]

(b) Find the value L0 such that for L L0 , the error probability Pe 106 .

[5pts]

(c) Find the SNR (function of L).

[5pts]

(d) Find the minimum SNR that ensures error-free transmission.

[5pts]

Problem 2

(QAM transmission(15 pts))

QAM transmission is used on an AWGN channel with N20 = 0.01. The transmitted signal
constellation points si for the QAM signal are shown in Figure 3 and are given by:
" #   

0
0
23
,
0
1
12
with each constellation point equally likely.

Figure 3: Signal constellation points


(a) Find M
P (message-set size, i.e., the number of possible messages) and the average energy
Ex = i pH (i)||si ||2 for this constellation.

[4pts]

(b) For the constellation (see Figure 3) draw the decision regions indicated for an ML detector.

[4pts]

(c) Find the minimum distance dmin for this constellation.

[3pts]

(d) Find the Nearest Neighbor Union Bound (NNUB) on Pe for the ML detector on this signal
constellation. For the NNUB you need to bound Pe maxi (Ni )Q( dmin
2 ), where Ni is the
N0
2
number of nearest neighbors of constellation point si and = 2 .

[4pts]

Problem 3

(Autocorrelation and PSD(20 pts))

Consider a zero-mean wide-sense stationary process X(t) with autocorrelation function RX ( ) =


E[X(t)X(t + )] = e2| | . Let N (t) be another zero-mean wide-sense stationary process that is
independent of X(t) with autocorrelation RN ( ) = E[N (t)N (t + )] = e3| | .
(a) Find the power spectral density SX (f ) of process X(t).

[7pts]

(b) Find the power spectral density SN (f ) of process N (t).

[3pts]

(c) Let Y (t) = X(t) + N (t). Find the autocorrelation function of Y (t).

[6pts]

(d) What is the power spectral density SY (f ) of Y (t)?

[4pts]

Hint: You may find the integrals given in the instructions useful.

Problem 4

(Waveform representations(25 pts))

Consider the following two orthogonal basis functions given in Figure 4.


(a) Use the basis functions given Figure 4 to find the modulated waveforms s0 (t) and s1 (t)
given the data symbols s0 = [1, 1] and s1 = [2, 1]. Note that the basis functions in Figure
4 are orthogonal but not orthonormal. It is sufficient to draw s0 (t) and s1 (t).
3

[7pts]

Figure 4: First basis

Figure 5: Second basis


[9pts]
(b) For the same s0 (t) and s1 (t), a different set of two orthonormal basis functions is employed, given in Figure 5. Find s0 which is the representation of s0 (t) in the new basis
functions. Find the s1 which produces s1 (t) in the new basis. Note that though the vector
representations change from si to si due to the change in basis functions, the waveforms
si (t) produced by both representations are the same.
(c) Now suppose we send the signal over a AWGN channel, i.e.,
H = i : y(t) = si (t) + N (t),
where N (t) is AWGN noise with power spectral density of 10dBm/Hz. Note that 0 dBm
corresponds to 1 mW, i.e., x dBm corresponds to power 10(x/10) /1000 in Watts. Further
suppose that both messages are equally likely, i.e., pH (0) = 12 = pH (1). Find the error
probability for this communication scheme.
Hint: Use the Q-function tables given in the Instructions.

[9pts]

You might also like