INTRODUCTION TO WIRELESS AND CELLULAR COMMUNICATIONS

NPTEL JUL – NOV 2020

Assignment – 8 Solutions

1. A receiver has 2 Antennas and performs Maximal ratio combining to improve the BER.
Both the receivers are uncorrelated and follow Rayleigh distribution. One of the
antennas has an average SNR of 4dB and the other has 6dB average SNR. Find the
average SNR after MRC.
a. 0.6504
b. 9
c. 6.493
d. 1

Answer: (c)
Solution:
γ M RC = γ 1 + γ 2
ΓM RC = E[γ 1 + γ 2 ] = E[γ 1 ] + E[γ 2 ] = 100.4 + 100.6 = 6.493

2. Given that DBPSK modulation scheme is being used, find the BER performance of the
above system.
a. 0.0286
b. 0.3059
c. 0.0684
d. 0.1529

Answer: (a)
Solution:
1 1
ψ = × = 0.05717
1+100.4 1+100.6
B ER = 0.5 × ψ = 0.0286

3. In the above question, the branch with 4 dB SNR has Nakagami-m fading with m = 2.
Find the new BER when DBPSK modulation scheme is used.
a. 0.0200
b. 0.1065
c. 0.0533
d. 0.0472

Answer: (a)
Solution:
ψ =( 1
1+0.5×100.4
)2 × 1
1+100.6
= 0.03945
B ER = 0.5 × ψ = 0.02
4. Will there be any difference in Average SNRs with conditions given in question 1 and
question 3?
a. Yes
b. No
c. Cannot be determined
d. None of the above

Answer: (b)
Average SNR in 3
ΓM RC = E[γ 1 + γ 2 ] = E[γ 1 ] + E[γ 2 ] = 100.4 + 100.6 = 6.493
This is same as Q1

5. In Question 1, if EGC is performed instead of MRC, find the average SNR after
combining
a. 3.542
b. 5.730
c. 4.246
d. 5.674

Answer: (b)
Solution:
M
γ EGC = 1
M ( ∑ √γ k ) = 0.5 × (√γ 1 + √ γ 2 ) 2 = 0.5(γ 1 + γ 2 + 2√γ 1 γ 2 )
k=1
ΓEGC = E[γ EGC ] = E [ 0.5(γ 1 + γ 2 + 2√γ 1 γ 2 )] = 0.5(Γ1 + Γ2 + 2 × E [√γ 1 ]E[√γ 2 ]
√πΓ1
√γ 1 is Rayleigh with mean​ 2
T heref ore E[√γ 1 ]E[√γ 2 ] = √ 41 2
π Γ Γ

2π √100.4 100.6
ΓEGC = 0.5(100.4 + 100.6 + 4 ) = 5.73

6.​ BER depends on which of the following?

a. Modulation scheme
b. Type of Fading
c. Antenna diversity
d. All of the above

Answer: (d)
Modulation scheme determines how many nearest neighbours exist for a symbol. More
neighbours imply larger BER.
The fading gives the instantaneous SNR and how likely a signal is to be faded during
reception. This varies the SNR.
The more the antenna diversity, less likely a signal is to be received in a faded condition.
Thus BER depends on it.

7​. Which of the following statements is correct?

a. Antenna gain is present in non-fading environment
b. Antenna gain causes a shift in the BER curve
c. Diversity gain improves the statistics of the SNR in a way that reduces the BER
d. All of the above

Answer: (d)
Solution:
Antenna gain is a property of the antenna and exists irrespective of channel
Higher antenna gain implies better signal strength and thus less fading effect
Diversity reduces chances of a deep fade event. Thus SNR is better.

8. ​For the Alamouti scheme discussed in lecture, the diversity order was found out to be
2. Consider a modified Alamouti scheme where 2 receiver antennas are present. One
can write down the matrix equations for the second receiver antenna similar to the case
in which only one receiver antenna is present and combine the matrix equations for the
two receiver antennas to get a single matrix equation for the entire system. What will be
the diversity order for this new system?
a. 2
b. 4
c. 8
d. 16

Answer: (b)
Solution:
9. Assuming i.i.d. Rayleigh fading in each branch, What is the average probability of bit error for
DPSK modulation under three-branch MRC. Given average SNRs:
γ 1 = 20 dB, γ 2 = 10dB, γ 3 = 10dB
(a) 4.09 x 10​-3
(b) 1.38 x 10​-3
(c) 4.09 x 10​-5
(d) 1.38 x 10​-9
Answer: (c)
Solution:
In linear scale,
γ 1 = 100, γ 2 = 10, γ 3 = 10
1 2
P error = 12 . 1+100
1
.( 1+10 ) = 4.09 × 10−5

10. Considering the above, what will be the error with no diversity and γ = 20dB
(a) 4.95 x 10​-3
(b) 9.98 x 10​-3
(c) 3.69 x 10​-5
(d) 9.98 x 10​-6
Answer (a)
Solution:
In linear scale,
γ = 100
P error = 12 . 1+100
1
= 4.95 × 10−3

11. The average probability of symbol error for MPSK in iid fading is: (Given: g = sin2 ( Mπ ) and
M γ (s) is the common MGF for the branch SNRs)
(a)
(M −4)π
M
Pe = 1
π ∫ (M γ (− sing ϕ ))M dϕ
2
0
(b)
(M −1)π
M
Pe = 1
π ∫ (M γ (− sing ϕ ))M −1 dϕ
2
0
(c)
π
M
Pe = 1
π ∫ (M γ (− sing ϕ ))M dϕ
2
0

(d)
(M −1)π
M
Pe = 1
π ∫ (M γ (− sing ϕ ))M dϕ
2
0

Answer(d)

Refer Andrea Goldsmith Equation 7.52

12. Consider a channel with SNR = 15 dB. What is the capacity of the channel per unit
bandwidth?
a. 5.027 bits/sec/Hz
b. 1.279 bits/sec/Hz
c. 3.76 bits/sec/Hz
d. 0.8136 bits/sec/Hz

Answer: (a)
Solution:
C
B = log 2 (1 + S N R) = log 2 (1 + 101.5 ) = 5.027 bits/s/Hz

13. For a signal with unit average signal power, the capacity of the channel depends on
a. The modulation scheme
b. Receiver sensitivity
c. Symbol rate
d. All of the above

Answer: c)
Ps Ps
C = B log 2 (1 + S N R) = B log 2 (1 + Pn) = B log 2 (1 + N 0B )

If P​s​ is constant, B is the only variable and it depends on Symbol Rate

14.​ Consider a channel with Free-Space path loss. What is the normalised capacity of the
channel (bits/sec/Hz) when the transmitter and receiver are separated by a distance of 150m?
Assume that SNR w/o fading = 90 dB, Carrier frequency = 900 MHz.
a. 6.34
b. 5.01
c. 2.87
d. 2.04

Answer: (b)
Solution:
SNR with fading = SNR without fading - Path Loss
4πdf 2 6 2
Path Loss (dB) ​= 10 log 10 ( c ) = 10log 10 ( 4π×150×900×10
3×108
) = 75.048 dB
SNR with fading = 90 - 75.048 dB= 14.952 dB
C
B = log 2 (1 + S N Rf ading ) = log 2 (1 + 101.4952 ) = 5.01 bits/sec/Hz

(​Consider for questions 15-19​) Consider a fading channel with instantaneous SNR given by
α2 Γ , where Γ = 30 dB and α is the instantaneous fading coefficient that varies with time.
BW = 20 kHz. The probability distribution of α is as follows:
pα (0.25) = 0.125 , pα (0.5) = 0.25 , pα (1) = 0.5, pα (1.5) = 0.125 .

15.​ What is the average SNR of the channel?

a. 20 dB
b. 25.2 dB
c. 29.30 dB
d. 33 dB

Answer: (c)
16.​ Can you find an upper bound on the Channel capacity using Jensen’s inequality?
a. 175.14 kbps
b. 64.28 kbps
c. 194.69 kbps
d. 58.05 kbps

Answer: (c)

17.​ What is the Ergodic capacity of the channel?

a. 64.28 kbps
b. 182.34 Kbps
c. 58.05 kbps
d. 192.84 kbps

Answer: (b)

18​. What should be the SNR of a flat-fading channel that has the same capacity as the Ergodic
capacity of the fading channel discussed above?
a. 27.44 dB
b. 34.16 dB
c. 19.3 dB
d. 56.28 dB

Answer: (a)

19.​ What is the Ergodic capacity of the channel discussed above, if there is a 6% chance of
outage at any time?
a. 183.2 kbps
b. 8.76 kbps
c. 9.642 kbps
d. 171.4 Kbps

Answer: (d)

Solutions 15-19:

𝛼 0.25 0.5 1 1.5

P(𝛼) 0.125 0.25 0.5 0.125

SNR𝛼 62.5 250 1000 2250

E [α2 ] = 0.252 × 0.125 + 0.52 × 0.25 + 12 × 0.5 + 1.52 × 0.125 = 0.8516

15.
Average SNR = E [α2 Γ] = Γ E[α2 ] = 103 × 0.8516 = 851.6 = 10log 10 (851.6) dB = 29.30

16
Channel capacity:
C = E[B log 2 (1 + S N R)] ≤ B log 2 (1 + E [SN R]) = 20 × 103 × log 2 (1 + 102.93 )
Therefore: C ≤ 194.70 Kbps

17
Ergodic capacity:
C = E[B log 2 (1 + S N R)] = B E α [log 2 (1 + S N Rα )]
C = 20 × 103 × [0.125log 2 (1 + 62.5) + 0.25log 2 (1 + 250) + 0.5log 2 (1 + 1000) +
0.125log 2 (1 + 2250)]
C = 20 × 103 × 9.117 = 182.34 Kbps
Option is not present in question paper

18
F rom above, Blog 2 (1 + S N R) = 182.34 Kbps
log 2 (1 + S N R) = 9.117
1 + S N R = 29.117 ⇒ S N R = 554.27` = 27.44 dB
Option is not present in question paper
19
Ergodic Capacity with outage = Ergodic capacity without outage x (1-P(outage))
= 182.34 × (1 − 0.6) Kbps ≈ 171.4 Kbps
Option is not present in question paper