0% found this document useful (0 votes)
35 views3 pages

Response Given by 1 5T 0 T T .: EE 251 - Fall 2009 San Jos e State University Solution of Midterm Exam # 2

This document contains the solutions to problems on a midterm exam for a digital communications systems course. Problem 1 involves calculating the impulse response of a matched filter for a given transmit filter and evaluating the filter response. Problem 2 calculates excess bandwidth for a system using raised cosine pulses and sketches the pulse spectrum. It also approximates bit error probability for the system considering inter-symbol interference from timing errors. Problem 3 compares the performance of two binary transmission systems, one using Manchester pulses and the other using NRZ, considering both average bit error rate and spectral properties on a channel with zero DC gain. The Manchester system is shown to perform better in the second case due to its power spectral density having a zero at DC.

Uploaded by

of30002000
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
0% found this document useful (0 votes)
35 views3 pages

Response Given by 1 5T 0 T T .: EE 251 - Fall 2009 San Jos e State University Solution of Midterm Exam # 2

This document contains the solutions to problems on a midterm exam for a digital communications systems course. Problem 1 involves calculating the impulse response of a matched filter for a given transmit filter and evaluating the filter response. Problem 2 calculates excess bandwidth for a system using raised cosine pulses and sketches the pulse spectrum. It also approximates bit error probability for the system considering inter-symbol interference from timing errors. Problem 3 compares the performance of two binary transmission systems, one using Manchester pulses and the other using NRZ, considering both average bit error rate and spectral properties on a channel with zero DC gain. The Manchester system is shown to perform better in the second case due to its power spectral density having a zero at DC.

Uploaded by

of30002000
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/ 3

EE 251 - Fall 2009 San José State University

Solution of midterm exam # 2

Problem 1 (35 points) A transmit filter used in a digital communication system has impulse
response given by
1
g(t) = √ e−t/5T , 0 < t ≤ T.
5T

(a) Sketch carefully the impulse response h(t) of the matched filter for g(t). (You may assume
that the constant of in matched filter is k = 1.)

Solution: The impulse response is sketched in the figure below:

g(t) h(t)=g(T-t)

(1/5T)1/2 (1/5T)1/2
e-1/5/(5T)1/2 e-1/5/(5T)1/2

t t
0 T 0 T

(b) Evaluate the response of the matched filter at t = T , yM F (T ).

Solution: At t = T , the response of the matched filter is equal to the energy Eg of the pulse:
Z T 1  2

yM F (T ) = Eg = g 2 (t) dt = 1 − e− 5 .
0 2

(c) Determine the response ycor (t) of a correlator for g(t), when the input is g(t) itself. Sketch it
and verify it using the answer from part (b).

Solution: The response of the correlator is


  
Z t  1
1 − e− 5T
2t
, 0 ≤ t ≤ T,
2
ycor (t) = g 2 (t) dt = 
− 25

0  1
1−e , t > T.
2

As expected, ycor (T ) = yM F (T ) = Eg . A sketch is shown below:

ycor(t)

(1-e—2/5)/2

t
0 T
Problem 2 (40 points) A binary communication system employs unipolar mapping and a square-
root raised cosine pulse for transmission over an ideal band limited channel. The system is subject
to AWGN of average power N0 /2. The bandwidth of the channel is 3 MHz and the bit rate 3.5
Mbps.

(a) Find the value of the excess bandwidth of this system, α.


2W 2×3 12 5
Solution: α = Rb −1 = 3.5 −1 = 7 −1= 7 or 71.4%.

(b) Sketch carefully, showing all pertinent labels, the raised cosine spectrum P (f ) used by this
system.

Solution: A sketch of the spectrum is shown in the figure below:

P(f)/T

0.5

f (MHz)
-3 -2 -1 0 1 2 3

0.5 1.75

(c) Due to a timing error, the output of the matched filter (MF) at the receiver is given by
1 1 1
Yk = √ Ak + √ Ak−1 + √ Ak−2 + Nk .
2 3 6
Approximate the average probability of a bit error, in terms of a single Gaussian Q-function
with Eb /N0 as a parameter. No details of the probability of error analysis are required, only
those of the most dominant term.

Solution: Recall that with unipolar
p mapping, 0 7→ 0 and 1 7→ 2Eb . The decision threshold
(without ISI) is set at λ = Eb /2. A list of all combinations of bits and MF outputs are
shown below:

Bk Bk−1 Bk−2 Yk − N k
0 0 0 p 0
0 0 1 p2Eb /6
0 1 0 p 2Ebp
/3
0 1 1 2Ebp
/3 + 2Eb /6
1 0 0 p 2Ebp
/2
1 0 1 p2Eb /2 + p2Eb /6
1 1 0 p 2Ebp
/2 + 2Ebp/3
1 1 1 2Eb /2 + 2Eb /3 + 2Eb /6

The worst cases occur for Bk = 0, Bk−1 = 1, Bk−2 = 0 and Bk−1 = Bk−2 = 1:
( p p ! p p p !)
1 Eb /2 − 2Eb /3 Eb /2 − 2Eb /3 − 2Eb /6
Pb ≈ Q p +Q p
8 N0 /2 N0 /2
(" s !# " s !#)
1 Eb Eb
= 1 − Q 0.97 + 1 − Q 0.16 .
8 N0 N0

Problem 3 (25 points) Two systems for binary transmission over an AWGN channel are being
evaluated: System S1 uses polar mapping with Manchester pulses. System S2 uses polar mapping
with NRZ pulses.

(a) What system has best performance? Justify carefully your answer.

Solution: Both systems use the same mapping. As a result, the performance is the same
in terms of average probability of a bit error, for the same average bit energy-to-noise ratio
Eb /N0 .

(b) Suppose that the transfer function H(f ) of the channel has zero DC gain so that H(0) = 0.
What system has best performance? (Hint: Power spectral densities.)

Solution: The power spectral density of Manchester pulses satisfies SX (0) = 0. On the other
hand, for NRZ pulses, SX (0) 6= 0 and the output of the channel will be distorted. Therefore
system S1 performs better.

You might also like