Q1.

Cellular Mobile Communication System

a) Prove that for the hexagonal cluster geometry, the co-channel reuse ratio is give by
Q= (3N) 1/2, where, N=i2+ij +j2, i and j are small integers (Q=D/R, D-distance to the
centre of the nearest co-channel cell; R-Cell radius). Show the sketch/graph for: i=1,
j=3.

Solution:

Generally, for N=i2+ij +j2, we can do the following to find the nearest co-channel
neighbours of a particular cell:

 Move i cells along any chain of hexagons and then,

 Turn 600 counter-clockwise and move j cells.

D Do

From the figure above,

D0 0
=R cos 120
2

¿> D02 =3 R 2

D2
¿> N = 2
D0

D2
¿> N =
3 R2

D
¿> =√ 3 N
R
1
D 2
¿>Q= = √ 3 N=¿Q=( 3 N ) Proved
R
The parameter Q in the above equation is called the co-channel reuse ratio and is
related to the cluster size.

So we see that the small value of Q provides larger capacity since the cluster size N
is small.

i=1 , j=3 ;
N=i 2 +ij+ j 2=13

1
2 5
3 7 4
1 1
4 6 3
5 2

Frequency Reuse with N=13

b) A receiver in an urban cellular radio system detects a 1mW signal at d=d 0=2 meter
from the transmitter. In order to mitigate co-channel interference effects, it is
required that the signal received at any base station receiver from another base
station transmitter which operates with the same channel must be below -100 dBm. A
measurement team has determined that the average path loss exponent in the system
is n = 4 after the distanced 0. Determine the major radius of each cell if a 12 hexagonal
cell reuse pattern is used.

Solution:

Here,
P0 [ dBm ] =0 , Pr [ dBm ] =−100 , n=4 , d 0=2 meter

d
Pr [ dBm ] =P 0 [ dBm ] −10 ×n × log ( )
d0

P 0 [ dBm ] −Pr [ dBm ]

d 10 ×n
⇒ =10 =316.2277
d0

⇒ d =2× 316.2277=632.45

Now, d is the minimum distance between same frequency cells so they do not interfere
with each other, and;

d d
R= =
Q √3 N

For 12 hexagonal cell, N=12;

 d 632.45
R= = =105.4 m
√3 N 6

So, the major radius is, R = 105.4 m eter for 12 hexagonal cells reuse pattern.

c) In a mobile telephone system the traffic intensity is defined as Au =λH where λ is

the average number of calls per unit time, H is the duration of the call expressed in
the same time units. For a system of U users the total offered traffic intensity is
A=U A u. The probability of a new telephone call being blocked in a system where the
total offered traffic intensity A is being served by the number of channel C is given by
Erlang B formula. Pb is called QOS (quality of Service). Find (or derive) the Erlang B
formula and make a table of total offered traffic intensity A for the QOS = (0.01 0.005
0.002 0.001) and the number of channels C= (2, 5, 10, 20, 40, 80, 100).

Solution:

Traffic intensity, Au =λH

λ - Average number of calls per unit time, H is the duration of the call expressed in the same
time units.

For a system of U users the total offered traffic intensity is A=U A u.

The Erlang B formula determines the probability that a call is blocked, when queuing is not
used, and is given by:
AC
C!
QOS=P [ Blocking ] = C k
∑ Ak !
k=0

Here C is the number of trunked channels offered by a trunked radio system and A is the
total offered traffic.

Number of Capacity (Erlangs) for QOS

Channels C
0.01 0.005 0.002 0.001
2 0.153 0.105 0.065 0.046
5 1.36 1.13 0.900 0.760
10 4.46 3.96 3.43 3.09
 20 12.0 11.1 10.1 9.41
40 29.0 27.3 25.7 24.5
80 56.1 53.7 51.0 49.2
100 84.1 80.9 77.4 75.2
Q2. Antenna Pattern
In a cellular mobile communication system base station a linear Antenna array is used
with 5 elements with omni-directional radiation pattern. The elements are at the
distance d from each other. In order to point the antenna radiation pattern in a
desired direction an additional phase shift α is introduced between each antenna
element as shown in figure. The angle between the line connecting the array elements
and the direction of the maximum radiation of the array is θ M. The radio wave carrier
frequency is such that the wave propagates in the open space with the wave length λ.

a) If, d=λ/2 and θM=π/6, Find the direction(s) towards which there is no radio
energy radiation from the array, and find the nulls of the radiation pattern.

Solution:

5
4

θ= 1800

There are 5 elements Antenna Array, so 5-1=4 Nulls.

If, d=λ/2 and θM=π/6,

M +1
2 ( n−1 ) πd
( AF )2 M +1 ( odd ) = ∑ an cos
n=1
[ λ
cos θ M ]
 2M+1 =5 => M=2

( λ2 ) cos π
2 +1
( AF )2 ×2 +1 ( odd )=∑ an cos
n=1
[ 2 ( n−1 ) π

λ 6 ]
( 2λ ) cos π + a cos 2( 2−1) π ( 2λ ) cos π + a cos 2 ( 3−1) π ( 2λ ) cos π
¿ a1 cos [ 2 ( 1−1 ) π

λ 6 ] [ 2
λ 6 ] [
3
λ 6 ]
π √3
¿ a1 cos 0+ a2 cos( )+ a3 cos( π √ 3)
2

Wave length λ
Nulls

b) Sketch an approximate radiation pattern for the case described in.

Solution:
 90

135
45

180 0

225 315

270

Radiation pattern

c) If d=λ and θM=π/6 , Find all directions of maximum radio wave radiation in
the range 0 to π (equal to the one at θM=π/6). Sketch an approximate
radiation pattern. (you may use MATLAB)

Solution:
 d=λ and θM=π/6

N=5;
2
1 π

( π
P θ M = , d= λ =
6 )1
N
|
sin

2
2
1
2
( (
sin N π cos ( θ M ) +

((
π cos ( θ M ) +
π
6
6

))
))
|
5 π 3 √ 3+ 1

¿
|1
5
sin

sin
( (
2 6
π 3 √ 3+1
2(( 6 )
)
)|
)

Is the direction of maximum radiation 0 to π.

Radiation Pattern

Q 3. Signal Propagation
Assume that a SNR of 12 dB is desired at the receiver. The GSM 900 MHz cellular
transmitter has an EIRP of 1 W (please do not confuse carrier frequency with the
 GSM bandwidth). The receiver uses a 3 dB gain antenna and the receiver has a
noise figure F=8, and by noise figure increasing the thermal noise effect. The
free space signal propagation is assumed until distance d 0 =100 m. After the
distance d0 assume propagation exponent n=4.

Furthermore assume that the thermal noise floor is given by k. B. F. T 0, where k

is the Boltzmann constant (1.38*10-23) in IS units, B is the bandwidth, F is the
noise figure and T0=3000 K.

a) Find the average signal to noise ratio level at the receiver placed 2 km far
from the transmitter.

Solution:

Noise Floor= k.B.F. T0 = 1.38 ×10−23 ×900 × 106 × 8 ×300

¿ 2.98 ×10−11 ( W )

¿−65.25 dBm

Threshold, γ =Noise Floor ( dBm ) + SNR ( dB )

¿−65.25+12

¿−53.25 dBm

Given,
c
EIRP=Pt G t =1 W ,G r=3 dB=2 , d 0=100 m, λ= =0.333 m
f

2
Pt G t G r λ
Pr ( d 0 ) =
( 4 π ) 2 d 02

1× 2× 0.3332
¿
( 4 π )2 1002

0.22
¿ 2
=0.14 ×10−6 ( W )
158× 100

¿−28.56 dBm

For, d=2 km=2000 m, n=4 ;

Pr´( d )=P r ( d 0 )−10 × n× log 10 ( dd )
0

¿−28.56−40 × log10 ( 2000

100 )

¿−28.56−52.04=−80.6 dBm

b) If we assume lognormal shadowing standard deviation σ =10 dB around

average value of the signal to noise ratio, find the percentage of time that the
desired SNR = 15dB is achieved at a receiver placed 2 km far from the
transmitter.

Solution:

σ =10 dB, SNR = 15dB,

Threshold, γ =Noise Floor ( dBm ) + SNR ( dB )

¿−65.25+15

¿−50.25 dBm

γ − Pr´( d )
Pr ( Pr ( d ) > γ ) =Q[ σ ]
−50.25+ 80.6
¿Q [ 10 ]
¿ Q [ 3.035 ] =0.00135

Percentage of time is 0.00135.

 Q 4. Viterbi Algorithm
A convolutional coder rate ½ is used for coding input binary sequence
m=101011. The following generator polynomials are used. g 1 = 1+D+D2+D3 is
characterizing path 1 of the encoder and g 2 =1+D+D3 is characterizing path 2 of
the encoder. Assume that the initial state of the encoder is S1 = (0, 0, 0)
and two trailing bits Tr= (0, 0, 0) are appended to the end of the message m.

a) Draw the encoder block diagram.

Solution:

m = 101011, coder rate= ½

PATH #1: g1 = 1+D+D2+D3

PATH #2: g2 =1+D+D3

Output
Second
First
Inputcoded
coded
data
coded
bits
bit
bits
bit

u1

m u1 , u 2
u2
t u

Time Output Time Output

t u
(Branch word) (Branch word)
001 000
 b) What is the output message from the encoder multiplexer? The multiplexer is
the two-input one-output device where bits from path 1 and path 2 of the
encoder are sent to the multiplexer output alternatively. The multiplexer is
considered to be part of the encoder.

Solution:

m= (101011)
ENCODER U

For a message sequence, m=101011;

D-domain polynomial representation, m ( D )=1. D 0+ 0. D 1 +1. D 2+ 0. D 3 +1. D 4 +1. D 5

¿ 1+ D 2 + D 4 + D 5

g1 = 1+D+D2+D3

g2 =1+D+D3

c 1 ( D )=g1 ( D ) m ( D )=( 1+ D+ D2 + D3 ) ( 1+ D2+ D 4 + D5 )

¿ 1+ D+ D8

c 2 ( D )=g2 ( D ) m ( D )=( 1+ D+ D3 ) ( 1+ D2 + D4 + D 5 )

¿ 1+ D+ D 2 + D4 + D 6 + D 7 + D 8

Therefore, the output sequences of paths #1 and #2 are as follows, respectively:

Output sequence of path #1: (110000001)

Output sequence of path #2: (111010111)

The resulting encoded sequence from the convolutional encoder:

c=(11, 11 , 01 ,00 , 01 , 00 , 01, 01 ,11)

 S N R

Noise
Stream 2’ Stream 2

Decoder 1 Decoder 2
Encoder 1 C Encoder 2

Stream 1 Stream 1’

c) Draw the encoder trellis diagram and indicate the encoder output sequence
from the multiplexer on the diagram.

Solution:

Encoder Trellis Diagram

d) Use the Viterbi Algorithm and the encoder/decoder trellis diagram to find the
original transmitted sequence (before convolutional encoder) if the received
sequence (after binary detection at the receiver) is given by: c= 11, 01, 01, 10,
01, 01, 11.

Note: (binary symbol by symbol detection/demodulator of the received

signal takes place at the receiver before the binary sequence decoding takes
place in the Viterbi Decoder)

Solution:

c = 11, 01, 01, 10, 01, 01, 11

Output sequence of path #1: (1001001)

Output sequence of path #2: (1110111)

c 1 ( D )=g1 ( D ) m ( D )=1. D 0+ 0. D 1 +0. D2 +1. D3 +0. D4 + 0. D5 +1. D6

¿ 1+ D 3 + D 6

c 2 ( D )=g2 ( D ) m ( D )=1. D 0+ 1. D1 +1. D 2+ 0. D3 +1. D4 +1. D5+ 1. D6

¿ 1+ D+ D 2 + D 4 + D 5 + D 6

Original transmitted sequence, m=10011

PATH #1
MODULO-2 ADDER

FLIP FLIP OUTPUT

INPUT
FLOP FLOP

MODULO-2 ADDER
PATH #2
 Q 5. CDMA
A cellular mobile communication system is using a Direct Sequence Code
Division Multiple Access (CDMA) technique with Binary Shift Keying (BPSK)
modulation. The receiver is implemented as a coherent demodulator for signal
frequency down conversion with the matched filter with CDMA signal and the
threshold detector for BPSK detection. It should be noted that the matched filter
is performing signal dispreading in this case. The system is accommodating K
users and the spreading ratio G of the system is 200, i.e.

1
Pe = erfc √ SNR M
2

G=200. The probability of bit error is where SNRM represents Signal power to
the Noise power ratio after the matched filter.

a) Let us assume that in a single user system, with no other users present, the
thermal noise is the only additive white Gaussian noise present in the
receiver. If the signal to Noise ratio in the single user system after the
matched filter is SNRM =20 and/or 25[dB], calculate the signal to noise ratio
of a single user system while the signal is still spread in frequency (before
returning signal to its original bandwidth).

Solution:

1
G = 200, Pe = erfc √ SNR M , Thermal Noise Ratio=E p / N 0
2

SNR M =20 dB∧25 dB

1 z 1 z
Q ( z) =
2[ ( )]
1−erf = erfc
√2 2 ( )
√2

1 20
Pe = erfc
2 √
25

1
¿ erfc ( 0.8 )
2

1 1.13
¿ erfc
2 √2 ( )
¿ Q ( 1.13 ) ≈ 0.13567
 b) If the CDMA system consists of 40 users additional to the user of interest (the
number of users K=41) with the same power, calculate the output signal to
noise ratio (SNR) for the user of interest after dispreading (returning signal
into its original bandwidth) in the presence of interference caused by 40
interferers and the thermal noise described in (a).

Solution:

K=41, Pe =0.13567 ,
Tb
G= =200
Tc

As we know, for multiple users average probability of bit error is given by;

1
Pe =Q
(√ K −1 N 0
+
3G 2 Eb )
1 1
⇒ Q−1 ( Pe )= =
K −1 N 0 N0
√ +
3G 2 Eb √ 41−1
+ ( )
3 × 200 2 E b

1
⇒ Q −1 ( 0.13567 ) =
N0
√ 40
+( )
600 2 E b

N0
⇒ 1.1 ×
√ 40
+ ( )
600 2 E b
=1

40 N0
⇒ + ( )
600 2 E b
=0.8264

N0
⇒ ( )
2 Eb
=0.8264−0.0667=0.7597

N0
⇒ =1.5194
Eb

Output SNR:
 Eb
⇒ =0.658
N0

c) Calculate the probabilities of bit error of the user of interest for the thermal
noise levels given in (a) in the presence of interference caused by 40 users.
Use the following assumptions:

i) The multiple access interference is AWGN-like, i.e. could be approximated

by the additional additive white Gaussian noise.

Probability of bit-error is:

2 EB
PB =Q (√ ) N0

¿Q (√ 2×40200 )
¿ Q ( 3.162 )

¿ 0.00084

ii) For large, you may use approximation for the error probability

erfc ( x ) ≈ √2 e−x 2

π .x

Function (ERFC):

Solution:
∞
Q ( X )=∫
√ 2 e− x dx
2

X π .x
∞
2 1
¿ √ ∫ e− x dx
2

π X x

∞ ∞
¿
π
∫[
√2 1 e− x dx +e−x 1 dx
x X
∫
2 2

X x ]
2 1
¿√ ( −e−x ) ( 2 x ) +e− x ln x ∞
π x [ 2

X
2

]
 Error Probability: PB =Q ( X )
∞
¿∫ √2 e−x dx
2

X π.x

2 1
¿√ ( – e− x ) ( 2 x )+ e−x ln x ∞
π x [ 2 2

] X

¿ 0.45

d) If the required probability of bit error for the user of interest is P e =10-5 , find
the maximum number of CDMA users K using BPSK modulation, including the
user of interest, for the thermal noise levels given in a). You may use
approximations i) and ii) specified in c).

Solution:

For the uncoded system, the probability of block error occurs if at least one
bit is received with error;
BLER=1−( 1−P e ) 40
Where BLER = probability of block error rate,
Probability of bit error for the user of interest is Pe =10-5
40
BLER=1−( 1−10−5 ) =1−0.9996=0.0004=4 × 10−4
2 Eb
Pb=Q (√ )N0

Eb 1 −1 2
⇒ = [ Q ( Pb ) ]
N0 2

1 2
¿ [Q−1 ( 4 × 10−4 ) ] ≈ 4.2
2
 Q 6. Equalization.
A digital channel with inter-symbol interference (ISI) has the following impulse
response h = (-0.5, 0, 0.5)T. (.)TIs the matrixes transpose. The ISI is being
suppressed by MMSE equalizer with three taps w= (w0, w1, w2)T.

a) Find the equalizer taps w according to the MMSE equalization criterion if the
noise variance is σ2= 1.

Solution:
T
h=(−0.5 , 0 , 0.5 )
2
MMSE ( ξ min ) =E [ e k 2 ]=E ( I k − ^
[ I k) ]
¿ E {d 2 ( k ) }−p T R−1 p

Where, h H = Hermitian matrix of h.

❑
−1 1

( ))
2
σ MMSE =N 0 exp ∫
2 π −π
(
π ln Q ( Ω ) +
εx
N0
dΩ

2 N0 ∞
2 2
E [ e k ]=σ n = ∫ |H ( f )| df
2 −∞

σ 2=1

So,
∞
N0 2
∫ |H ( f )| df =1
2 −∞

w [ n ] =x [ n ] ⊗ h [ n ]

w0

() −1
w= w 1 = ( h H h ) h H
w2
 −0.5
¿
( )
0
0.5

b) Is the ISI entirely eliminated by equalization? If not, how many ISI terms,
besides the desired signal sample, remain?

Solution:

No, ISI not entirely eliminated by equalization, there are still 2 ISI terms remain.

Adaptive equalizers compensate for signal distortion attributed to intersymbol

interference (ISI), which is caused by multipath within time-dispersive channels.

Those are:

d ¿ ( r ) =L ( 1−r ) And x [n]

c) Show the total impulse response of the channel and the equaliser in the
cascade.

Solution:

Fixed Adjustable

ai Channel Equalizer
yi
{} {}
ui

{si}

Cascade Equalizer

w 0 −0.5

w2( )( )
w= w 1 = 0
0.5
 −0.5
h=
( ) 0
0.5

The combined impulse response { si }mapping the source sequence { a i } to the

equalizer output sequence{ y i }, as in:
y i=∑ s k ai−k is the convolution of { hi } ∧{gi }
k

L
si=∑ g k hi−k
k=0
Hence, total impulse response of the channel and the equaliser in the cascade.

REFERENCE BOOKS:

1. Principles of Mobile Communication; Stüber, Gordon L. 2nd ed., 2001, 776 p. 51

illus.
2. W.C.Y. Lee , Mobile cellular telecommunication systems. McGraw-Hill, 1989.
3. Wireless Communications: Principles and Practice by Theodore Rappaport, Prentice
Hall.
4. Wireless Digital Communications: Modulation & Spread Spectrum Applications.
Kamilo Feher. Prentice-Hall, 1995, ISBN 0-13-098617-8.
5. The Mobile Communications Handbook. Edited by Jerry D. Gibson, C R C Press, 1995,
592p., ISBN 0-8493-8573-3
6. Mobile Communication Systems. J. D. Parsons, Halsted Press, 1989 ISBN 0-470-
21213-6 292p.