Final Exam

KING ABDULLAH UNIVERSITY OF SCIENCE AND TECHNOLOGY
ELECTRICAL ENGINEERING DEPARTMENT
ECE 242 Digital Communications and Coding
Final Exam
(Dec. 14, 2020)
Q1. (30 Points)

a. (4 Points) In OFDM, what is the advantage of using a cyclic prex over a guard interval?
What is a disadvantage of using a cyclic prex or a guard interval?
b. (5 Points) Calculate the capacity of a BSC with a cross-over probability p = 0.3. An
engineer claims that he can achieve this capacity using a convolutional encoder of rate
1/3. Do you think his claim is reasonable?
c. (6 Points) Consider a linear block code with a generator matrix G of size 3 × 6 and
a parity check matrix H. We decided to interchange the 2nd and 3rd columns of G to
create a new matrix G0 . How does this change the code properties (rank(G0 ), k, n, and
dmin )? What is the corresponding parity check matrix?
d. (4 Points) A source has entropy of 5 bits/ symbol. You hired an engineer to design an
optimal code for this source and he achieved an average length of 4 bits/symbol. Do you
think he did a good job? Explain how would you improve his design?
e. (4 Points) A communication engineer claims to have designed a modulation technique
with power eciency rb = 4 and bandwidth eciency r = 4. Do you think his claim is
reasonable?
f. (6 Points) Order the following modulation techniques according to their bandwidth
eciency: 4QAM, 4PAM (DSBSC), 4PAM (SSB), 16FSK, 8PSK, 16QAM.
1
Q2. (30 Points) Consider the following syndrome table corresponding to a linear block code and
answer the following questions
s e
000 000000
001 000001
010 000010
011 000100
100 001000
101 010000
110 100000
111 100001
Table 1
a. (2 ) What is the size of this code?

Points
b. (3 ) What are the possible value(s) of dmin ?

Points
c. (3 ) What is the number of detectable error sequences?

Points
d. (6 Points) What is the probability of error for this code assuming that the code bits
pass through a BSC channel with cross over probability equal to p?
e. (6 ) Use Table 1 to construct a parity check matrix H for this code.
Points
f. (6 ) Construct a generator matrix G for this code.

Points
g. (3 ) Why do we focus on linear block codes to general block codes?

Points
2
Q3. (25 Points)
a. (4 Points) Mention one advantage of using convolutional codes over block codes. Men-
tion one advantage of block codes over convolutional codes.
b. (8 Points) consider the rate 1/3 convolutional encoder dened by the following 3 impulse
responses
g1 = [1 0 0]
g2 = [1 0 1]
g3 = [1 1 1]
Find the encoder output correspond to the following input
[x0 x1 · · · x6 ] = [1 0 0 1 0 1 1]
assuming you start from ushed registers and end with ushed registers. Use the D-
transform approach.
c. (3 Points) Now assume that the inital state of the registers is [1 1] (i.e., the register is
not initially ushed). How would you nd the output using the D-transform approach?
(Note: Don't nd the output. Just describe what you would do using the D-transform
approach.)
d. (9 Points) The coded bits are transmitted through a AWGN using BPSK modulation.
The decoder input is described by
yi = (2bi − 1) + ni
where bi is the ith transmitted bit and ni is AWGN with variance σ 2 = 0.01. We receive
the following sequence of values
−1.0 0.3 − 1.0 0.3 + 1.2 0.4 9.0 3.5 −3
Determine the input sequence based on hard decoding assuming that register is not
ushed both at the initial state and nal state.
3
Q4. (30 Points ) Consider the vector space created by the signals shown in Figure. 1.
Figure 1: Signals for Q4.
a. (5 ) Find an orthonormal basis of minimum size for this set of waveforms.

Points
b. (4 ) Find the corresponding constellation points (no need to draw them).

Points
c. (5 Points) Determine the union bound for the constellation and use that to determine
the power eciency rb .
d. (3 ) Determine the bandwidth eciency of the constellation.
Points
e. (4 ) How does the performance of this system compare to 4 QAM?

Points
f. (8 Points) We use these waveforms to send our data through a communication chan-
nel. At the receiver, we detect the waveform shown in Figure. 2. Based on the received
waveform, which of the waveforms in Figure. 1 has the highest probability of being trans-
mitted. Explain what reasons make the receive waveform dierent from the transmitted
ones? (assuming that the channel is an ultrasound channel). Assume that the waveforms
are equiprobable and the channel is AWGN with σ 2 = 0.01.
Figure 2: The detected waveform for 4f

Final Exam

Uploaded by

Copyright:

Available Formats

Final Exam

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Final Exam

Uploaded by

Copyright:

Available Formats

KING ABDULLAH UNIVERSITY OF SCIENCE AND TECHNOLOGY

ELECTRICAL ENGINEERING DEPARTMENT

ECE 242 Digital Communications and Coding

Q1. (30 Points)

a. (2 ) What is the size of this code?

b. (3 ) What are the possible value(s) of dmin ?

c. (3 ) What is the number of detectable error sequences?

f. (6 ) Construct a generator matrix G for this code.

g. (3 ) Why do we focus on linear block codes to general block codes?

Find the encoder output correspond to the following input

−1.0 0.3 − 1.0 0.3 + 1.2 0.4 9.0 3.5 −3

Figure 1: Signals for Q4.

a. (5 ) Find an orthonormal basis of minimum size for this set of waveforms.

b. (4 ) Find the corresponding constellation points (no need to draw them).

e. (4 ) How does the performance of this system compare to 4 QAM?

Figure 2: The detected waveform for 4f

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.