INTRODUCTION TO WIRELESS AND CELLULAR COMMUNICATIONS

NPTEL JUL – NOV 2019

Assignment – 7

1. Given channel coherence time and coherence bandwidth are 2.5 us and 200 kHz
respectively. 16-QAM modulation is used for communication to achieve 1Mbps throughput.
Which of the following is true about the channel?
(a) Slow fading, Frequency selective
(b) Fast fading, frequency selective
(c) Slow fading, Frequency flat
(d) Fast fading, Frequency flat
Answer: (b)

Solution:

𝑇ℎ𝑟𝑜𝑢𝑔ℎ𝑝𝑢𝑡 106
𝐵𝑊 = = = 250𝑘𝐻𝑧
log 2 𝑀 4

1
𝑆𝑦𝑚𝑏𝑜𝑙 𝑃𝑒𝑟𝑖𝑜𝑑, 𝑇𝑠 = = 4𝜇𝑠
𝐵𝑊

𝑇𝑠 > 𝑇𝑐 ⇾ Fast fading

𝐵𝑊 > 𝐵𝑐 ⇾ Frequency selective

2. How would the above answer change if BPSK is used instead of 16-QAM. (consider 1Mbps
throughput)
(a) Slow fading, Frequency selective
(b) Fast fading, frequency selective
(c) Slow fading, Frequency flat
(d) Fast fading, Frequency flat
Answer: (a)

Solution:

BW = 1MHz , 𝑇𝑠 = 1𝜇𝑠

𝑇𝑠 < 𝑇𝑐 ⇾ Slow fading

𝐵𝑊 > 𝐵𝑐 ⇾ Frequency selective

3. Which of the following spectral density plots refer to spectrum with an LOS component?

(a)

(b)

(c)

(d)
Answer: (a)

Solution:

Jakes’ model (Bell-shaped model) represents the NLOS component (Rayleigh fading)
The impulse represents the LOS component

4. You are asked to generate a Rayleigh fading channel and are aware of a method to
generate a Rayleigh random variables and many independent random variables can be
generated following that procedure. If so, what is the need to resort to methods such as
Smith method or Clarke and Gans method?
 (a) Independent random variables do not model the time correlation.
(b) To reduce the computational complexity
(c) Both (a) and (b)
(d) None of these
Answer: (a)

Solution:

Rayleigh coefficients are correlated in time (due to doppler), which cannot be generated by
simply generating independent Rayleigh random variables

5. Given the maximum Doppler shift in a channel is 80 Hz. It is required to simulate the
channel using Jakes model with 30 oscillators. Which of the following is not an oscillator
frequency?
a) 80 Hz
b) 79.89 Hz
c) 79.74 Hz
d) 79.55 Hz
Answer: (c)

Solution:

𝑁0 + 1 = 30 ⇾ 𝑁0 = 29

Oscillator frequency,
2𝜋𝑘
𝑓𝑘 = 𝑓𝐷 𝑐𝑜𝑠 ( )
4𝑁0 + 2

𝑓0 = 80 𝐻𝑧
𝑓1 = 79.88 𝐻𝑧
𝑓2 = 79.55 𝐻𝑧

6. Which of the following is not true about the Walsh hadamard matrices?
(a) Symmetric matrix
(b) Orthogonal matrix
(c) Anti-symmetric matrix
(d) None of the above
Answer: (c)

Solution:

Walsh hadamard matrices are symmetric and orthogonal, but not anti-symmetric.

7. Which of the following is true w.r.t Smith method and Clarke and Gans?
 (a) Smith method requires less number of computations
(b) Clarke and Gans method has less computational complexity
(c) Both have same complexity
(d) None of the above
Answer: (a)

8. Consider a base station and a mobile system with two antennas. The transmission
from base station is picked up by both the antennas which experience different
channel conditions and hence different values of instantaneous SNR. The SNR for
both the antennas is below the threshold at which the signal can be properly detected.
Which of the following diversity technique(s) may lead to a proper detection of the
signal?
a) Selection diversity
b) Optimal diversity
c) Both (a) and (b)
d) None of the above
Answer: (b)

Solution:

Selection diversity helps when atleast one of the antenna experiences an SNR greater
than the threshold.
Whereas, with Optimal Diversity, even if all the receiver antennas are experiencing SNRs
below the threshold, they can be combined to push the overall SNR up.

9. In the previous question, assume that one of the antennas experiences SNR which is
above the required threshold while the other antenna does not. Which of the following
diversity technique(s) will lead to a proper detection of the signal in this case?
a) Selection diversity
b) Optimal diversity
c) Both (a) and (b)
d) None of the above
Answer: (c)

Solution:
As one of the receiver antenna is experiencing a good SNR, even Selection diversity
leads to a proper detection.

10. Selection diversity is not useful in which of the following scenarios.

(a) All the antennas are weak
(b) Antennas are correlated
 (c) The antennas are very close to each other
(d) All of the above
Answer: (d)
Solution:

Atleast one of antennas has to be reasonably strong.

In both option b) and c), antennas will not be independent. So if one of the antenna is weak, the
others will also be reasonably weak, as they are correlated

11. Consider a three antenna BPSK system. If only one of the antennas is used, it has a
performance of 𝑃𝑒 = 3.87 × 10−6 in an AWGN channel. Now, if all the three antennas are
used with selection diversity, what will be 𝑃𝑒 if it is experiencing Rayleigh fading? It is known
that the minimum SNR required for successful reception is 4dB. Given 𝑄(√20) =
3.87 × 10−6 . Given the probability of bit error for coherent BPSK in AWGN channel is given
by Pe,BPSK = Q(√2𝛾) where 𝛾 is the SNR.
(a) 0.013
(b) 0.036
(c) 0.011
(d) 0.222
Answer: (c)
Solution:

In AWGN channel, 𝑃𝑒 = Q(√2Γ) ⇾ Γ = 10

Note: In an AWGN channel, instantaneous SNR (𝛾) = Avg. SNR (Γ)

In Rayleigh Fading, with 3 antennas for selection diversity,

𝛾𝑡ℎ 3
𝑃𝑒 = (1 − 𝑒 − Γ ) = 0.011

12. In an AWGN channel, which of the following diversity methods will be effective.
(a) Selection diversity
(b) Optimal diversity
(c) Co phasing
(d) Both (b) and (c)
Answer: (d)
Solution:

Selection Diversity is not effective as all the antennas experience same instantaneous SNR (in
an AWGN channel).
13. In Co-phasing diversity:
Statement 1: The scaling constant Gk is chosen such that the weak antennas are given
lower weightage.
Statement 2: The scaling constant Gk is chosen such that the phase dependence on SNR is
eliminated.
Which of the following is correct?
(a) Statement 1 is False; Statement 2 is True
(b) Statement 1 is True; Statement 2 is True
(c) Statement 1 is False; Statement 2 is False
(d) Statement 1 is True; Statement 2 is False
Answer: (a)
Solution:

Co-phasing only eliminates the phase dependence by multiplying 𝑒 −𝑗𝜙 . But doesn’t scale the
magnitude based on the antenna strength.

14. Co phasing was discussed in the lecture. Consider the case where cophasing is
implemented using 4 receiver antennas. There are four signals 𝑟1 (𝑡) , 𝑟2 (𝑡), 𝑟3 (𝑡) and 𝑟4 (𝑡)
that will be combined here. Assume that the average SNR of each antenna is the same and
is given by 𝛤. What is the average SNR i.e. E[𝛾co-phasing] for the combination of the 4
antennas in cophasing?
𝜋
(a) 𝛤(1 + )
4
𝜋
(b) 𝛤(1 + )
3
𝜋
(c) 𝛤(1 + 2)
3𝜋
(d) 𝛤(1 + )
4
Answer: (d)
Solution:

(𝑁𝑟 − 1)𝜋
E[𝛾𝑐𝑜−𝑝ℎ𝑎𝑠𝑖𝑛𝑔 ] = Γ (1 + )
4

15. Which selection diversity will you implement in a high noise system and why? Choose the
option with the most appropriate reason.
(a) Post Detection Selection Diversity, because all the SNRs are present and we can
select the maximum.
(b) Pre Detection Selection Diversity, because only one Receiver is required compared
to Post Detection Selection Diversity where many Receivers are required.
(c) Post Detection Selection Diversity, because even though more receivers are
required we can be sure about the SNR value. RSSI is not a good measure of SNR
in high noise system.
(d) Both Post Detection and Pre Detection Selection Diversity can be implemented as
there is no difference between the two in a high noise system.
Answer: (c)
Solution:

In a high noise system, RSSI is not a good measure of SNR.

16. For the different diversity schemes

𝛤MRC : Average SNR for MRC with diversity 2
𝛤cophasing : Average SNR for Co-phasing with diversity 3
𝛤SC : Average SNR for Selection diversity with diversity 4
Arrange 𝛤MRC , 𝛤cophasing and 𝛤SC assuming the average SNR of each antenna is same and
is given by 𝛤.
(a) 𝛤MRC > 𝛤cophasing > 𝛤SC
(b) 𝛤SC > 𝛤MRC > 𝛤cophasing
(c) 𝛤MRC > 𝛤SC > 𝛤cophasing
(d) 𝛤cophasing > 𝛤SC > 𝛤MRC
Answer: (d)
Solution:

Γ𝑀𝑅𝐶 = 2Γ

(𝑁𝑟 − 1)𝜋 𝜋
Γ𝑐𝑜𝑝ℎ𝑎𝑠𝑖𝑛𝑔 = Γ (1 + ) = Γ (1 + ) = 2.57𝜋
4 2

1 1 1
Γ𝑆𝐶 = Γ (1 + + + ) = 2.083Γ
2 3 4