Assignment 9 Solutions Update1
Assignment 9 Solutions Update1
Assignment 9 Solutions
Answer: (d)
Solution: Both water-filling and channel inversion use the channel information to decide
the power allocation. So, they both need CSIT. Whereas, capacity with outage does not
need CSIT.
2) Consider a single tap channel of 15 kHz bandwidth with the following SNR
characteristics.
p( γ 1 ) = 0.25 and γ 1 =
0.13
p( γ 2 ) = 0.45 and γ 2 = 9
p( γ 3 ) = 0.30 and γ 3 =
18
Find the ergodic capacity of the channel.
a) 209.65 kbps
b) 42.20 kbps
c) 84 kbps
d) 87 kbps
Answer: (b)
Solution:
C = B ∑ p (γ i ) log 2 (1 + γ i ) = 15 × 2.813 kbps ≈ 42.20 kbps
i
P (γ i )
3) Find the optimal power allocation for each SNR state of the channel distribution
P
given in the question (2).
a) 0, 1.212, 1.288
b) 1.061, 1.212, 1.288
c) 0, 1.31, 1.37
d) 2.41, 0.143, 0.067
Answer: (c)
.
4) For the channel distribution given in question (2), find the maximum capacity when CSIT
is available.
a) 13.81 kbps
b) 45.88 kbps
c) 201.6 kbps
d) None of the above
Answer: (b)
Solution: 3 and 4
Maximum capacity with CSIT can be achieved using water-filling:
C = B ∑ (γ i /γ 0 ) p (γ i )
i
Considering all three channels to be in use, the power constraint equation becomes,
+
( ∑
i
1
γ0 − ) p (γ ) < 1
1
γi i
( 1
γ0 − 1
0.13 ) 0.25 + ( 1
γ0 − ) 0.45 + (
1
9
1
γ0 − 1
18 ) 0.3 = 1
γ 0 = 0.334
This value of γ 0 is contradicting the usage of all three channels (as γ 1 < γ 0 ). So, check for only
two channel usage.
( 1
γ0 − 1
9 ) 0.45 + ( 1
γ0 − 1
18 ) 0.3 = 1
γ 0 = 0.703
+
Pi
P = ( 1
γ0 − 1
γi )
Using the value of γ 0 calculated above, we get power allocation 0, 1.31 and 1.37
This value of γ 0 obtained above is not contradicting the usage of two channels. So the
capacity is
9 18
C = 15 × (0.45 × log 2 ( 0.703 ) + 0.3 × log 2 ( 0.703 ) = 45.88 kbps
5) For the channel distribution given in question (2), find the capacity with outage when
CSIT is not available.
(Hint: Choose different outage snr thresholds, see where you get maximum capacity.)
a) 63.72 kbps
b) 37.37 kbps
c) 19.11 kbps
d) 59.13 kbps
Answer: (b)
Solution:
When there is no CSIT, different outage can be allowed as:
Case 1. Allowing no outage, we have to send assuming the channel is always γ 1 . In this case,
C = 15 × log 2 (1 + 0.13) = 2.645 kbps.
Case 2. Allowing outage whenever the channel is γ 1 ,
C = 15 × (1 − 0.25) × log 2 (1 + 9) = 37.37 kbps.
Case 3. Allowing outage whenever the channel is γ 1 , γ 2 ,
C = 15 × (1 − 0.25 − 0.45) × log 2 (1 + 18) = 19.11 kbps.
So, best is case 2, so the capacity is 37.37 kbps.
6) For the channel distribution given in question (2), find the capacity when channel
inversion is performed.
a) 4.14 kbps
b) 5.34 kbps
c) 6.7 kbps
d) 8.81 kbps
Answer: (d)
Solution:
Using the formula for capacity with channel inversion,
C = B log 2 (1 + K )
K is obtained from:
P (γ l ) 1
∑ K( γl ) = 1 ⇒K= P (γ )
= 0.503
l ∑( γ l )
l
l
7) For the channel distribution given in question (2), find the truncated channel inversion
capacity
a) 60 kbps
b) 45 kbps
c) 30 kbps
d) 70 kbps
Answer: (b)
Solution:
Using the formula for capacity with channel inversion,
C = B log 2 (1 + K )
K is obtained from:
P (γ l ) 1
∑ K( γl ) = 1 ⇒K= P (γ )
= 15
l ∑( γ l )
l
l
C = (1 − P (γ rejected ))Blog 2 (1 + K ) = 45 kbps
8) What are the benefits of having multiple transmitters and receivers in a wireless system?
a) Diversity benefit
b) Capacity increases
c) Receiver complexity reduces
d) Both (a) and (b)
Answer: (d)
Solution:
Having multiple transmitters and multiple receivers, gives both diversity benefit and leads to
increases in capacity.
9) In the lecture it was shown that using DS-SS increases the bandwidth of the signal. What is
the consequence of this increase in bandwidth?
a) Time resolution becomes smaller and as a result many multipath components can be
received which can be combined using MRC diversity combining
b) Time resolution becomes larger and as a result many multipath components can be
received which can be combined using MRC diversity combining
c) Time resolution becomes smaller and as a result very few multipath components can be
received which can be combined using MRC diversity combining
d) Time resolution becomes larger and as a result very few multipath components can be
received which can be combined using MRC diversity combining
Answer: (a)
Solution:
Since BW increases, time resolution becomes smaller. As time resolution becomes smaller,
different multipath components can be distinguished and hence they can be combined using
MRC.
Answer: (c)
Solution:
DS-SS increases bandwidth and decreases the magnitude. So the answer is “Statement 1 is
true, Statement 2 is false.”
11)In CDMA multiple users can occupy the same frequency bands at the same time.
(a) True
(b) False
Answer: True
Answer: (d)
Answer:(d)
Solution:
Waterfilling maximizes throughput by transmitting only on good channels. This means that data
is delayed when waiting for fading effects to disappear.
14. You have to allocate power efficiently for a fast fading frequency selective channel to
maximize throughput over a long period of time. All the while, you are heavily constrained by
power. Which of the following will you choose:
(a) Waterfilling over time.
(b) Waterfilling over frequency.
(c) Both (a) and (b)
(d) Neither (a), nor (b)
Answer: (c)
Solution:
Channel is fast fading. Hence instantaneous SNR changes very rapidly. Thus waterfilling over
time is required.
Channel is frequency selective. Thus every frequency is attenuated differently. A waterfilling
solution is required to select the better channels at any point of time over which data is to be
transmitted.
15. Spread spectrum communication uses chips for transmission and reception of signals.
You have a modulator with chip duration 2 µs. You are transmitting a signal with symbol rate
31.25 Kbaud. What is the spreading factor?
(a) 16
(b) 32
(c) 125
(d) None of the above
Answer:(a)
Solution:
106
31.25 Kbaud ⇒ 31250 symbols/sec ⇒ S ymbol duration = 31250 μs = 32 μs
Symbol duration 32
S preading f actor = Chip duration = 2 = 16
16. Given below is the Spectrum of the pulse before spreading by DS-SS technique.
(The x- axis in all plots is in MHz)
Identify the spectrum after it is spread with chips of spreading factor 8.
(a)
(b)
(c)
(d)
Answer : (b)
Solution:
The spectrum is spread by 8. Therefore, the first zero crossing point is at 8x where x is
the zero crossing point of the unspread pulse spectra.
Given: zero crossing point is at 0.02
Thus, for spread pulse, zero crossing point is at 8x0.02 = 0.16
Also, the amplitude attenuates by a factor of Q2
This is option (b)
17. Consider a system with 1 transmit antenna and L receive antennas. Independent
C N (0, N 0 ) corrupts the signal at each of the receive antennas. The signal has a transmit power
constraint of P. Suppose the gain between the transmit antenna and receiver antennas is
constant, equal to 1. What is the capacity of the channel?
a) log 2 1 +( P
N0 )
b) log 2 (1 + LP
N0 )
c) log 2 (1 + P
LN 0 )
d) log 2 (1 + L2 P
N0 )
Answer: (b)
y k = x + nk for all 1 ≤ k ≤ L
L L
︿
y = ∑ ( x + nk ) = L x + ∑ nk
k=1 k=1
This looks like a single AWGN channel with transmit signal power L2 P and noise
variance LN 0 . Therefore, the capacity is given by
C = log 2 1 + ( L2 P
LN 0 ) = log (1 + ) 2
LP
N0
18. Consider a time-invariant frequency-selective block fading channel that has three
subchannels of Bandwidth B = 0.5 MHz each. The frequency responses associated with each
subchannel are H 1 = 1 , H 2 = 2 , and H 3 = 3 , respectively. The transmit power constraint is
P = 10 mW and the noise PSD N 0 /2 has N 0 = 10−9 W /Hz . Find the Shannon Capacity of this
channel and the optimal power allocation that achieves this capacity.
e) 5.465 Mbps
f) 3.415 Mbps
g) 6.831 Mbps
h) 10.93 Mbps
Answer: (c)
Substituting the values, we get γ 1 = 20 , γ 2 = 90 , and γ 3 = 180 . Considering all the three
channels are being utilised, using water-filling algorithm,
3
∑
j=1
( 1
γ0 − 1
γj )=1
3 1 1 1
γ0 =1+ 20 + 80 + 180
γ 0 = 2.81 < γ j for j =1, 2 and 3
19. Consider two users simultaneously transmitting to a single receiver in an AWGN channel.
This is a typical scenario in a cellular system with multiple users sending signals to a base
station. Assume the users have equal received power of 10 mW and total noise at the receiver
in the bandwidth of interest of 0.1 mW. The channel bandwidth for each user is 20 MHz.
Suppose that the receiver decodes user 1’s signal first. In this decoding, user 2’s signal acts as
noise (assume it has the same statistics as AWGN). What is the capacity of user 1’s channel
with this additional interference noise?
i) 19.86 Mbps
j) 133.16 Mbps
k) 11.65 Mbps
l) 20 Mbps
Answer: (a)
P = 10 mW
I = 10 mW
N = 0.1 mW
SINR = P / (I + N) = 0.99