1.

Trunking theory

Cellular radio system rely on trunking to accommodate a large number of users in a limited
radio spectrum. The concept of trunking allows a large number of users to share the
relatively small number of channels in a cell by providing access to each user, on demand,
from a pool of available channels. Some important definitions are given below

a. Set-up time – The time required to allocate a radio channel to a requesting user.
b. Blocked call – A call that cannot be completed at the time of request due to congestion
(lost call).
c. Holding time ( Th ) – Average duration of a typical call.
d. Request rate (  ) – The average number of calls per unit time. It is also called rate of
arrival.
e. Traffic intensity- Measure of channel time utilization (Erlangs).
f. Load (A) – Traffic intensity across the entire radio system. It is basically offered load to
the system

GOS (grade of service) : A measure of congestion which is specified as a probability. There

are two major classes of trunked radio system namely (i) Lost call cleared (LCC) system
and (ii) Lost call delayed (LCD) system.

(i) LCC- In LCC, a user request is blocked if there is no available channel in the channel pool.
The user is free to put the call request later. The blocking probability (Pb) which is describes
by Erlang B formula is the measure of GOS in LCC system.

(ii) LCD- In LCD, queues are used to hold requests that are initially blocked. The probability of
an arriving call being delayed is the measure of GOS in the LCD system. It is described by
Erlang C formula.

Let us consider each user generates a traffic intensity of Ai Erlang is given by

Ai  iTh (1.20)
where i is the number of call request per unit time for each user and Th is the average
duration of a call. If the system contains U number of users, the total offered traffic
intensity, A , is given by

A  UAi  U iTh  Th (1.21)

where   U i . In C channel trunked system, if traffic is equally distributed among the
channels, then the traffic intensity per channel is Ac ,
A UA
Ac   i (1.22)
C C
Arrival of calls follows Poisson distribution. Let x be the number of calls arrive to the
system between the time period t and (t +T ). The probability of arrival of x calls can be
describes as
 x 
P  x  e (1.23)
x!
It is to be noted that all the arrivals are independent to each other. The probability that a call
exists more than t1 time can be represented as,
P T  t1   et1 /Tmean (1.24)
where Tmean is the mean of call duration, i.e., Tmean  Th .

Probability of an arriving call being blocked is

AC
Pb  CC ! k (1.25)
A
k 0 k !
** Pb is the measure of GOS in LCC system and it is Erlang B formula.

Prob. 1.9

In a cell with 100 MSs, on an average 30 requests are generated during an hour, with
average holding time Th = 360 sec. Find offered load.

Soln.

30
 ; Th  360 sec
3600
30
A   Th   360  3 Erlangs
3600

Prob 1.10

How many users can be supported for 0.5% blocking probability for the 5 number of
trunked channels in LCC system. Assume each user generates 0.1 Erlangs of traffic.

Soln.

From Erlang B chart we get A 1.13 for Pb  0.5% and C  5

Ai  0.1 Erlangs
A 1.13
Thus, the number of user, U    11.3  11users
Ai 0.1
Prob 1.11

An urban area has a population of 2400000. A cellular service provider of that area has 394
cells with 19 channels each. Find the number of users can be supported at 2% blocking if
each user averages two calls per hour at an average call duration of 3 minutes. Assuming
that the service provider operates at maximum capacity.

Soln.

Pb  2%; i  2 call / hr; Th  3 min s

2
Ai   3  0.1 Erlangs
60
For Pb  2% and C  19 , offered traffic intensity for a cell is A  12.33 (Erlang B chart)
A 12.33
The number of users can be supported by a cell is U    123.3  123
Ai 0.1
Thus the service provider can support 394 123  48462 users.

Prob 1.12

In a company, one call arrives every 5 minutes, on average. During a period of 10 minutes,
what is the probability that
a) One call arrives.
b) More than one call arrives.
c) A call last more than 4 minutes while Th  2 minutes.

Soln.
a)   10 / 5  2
x =1
21 2
P 1  e  0.27
1!
b)
P  x  1  1  P(0)  P(1)
 1  0.135  0.27  0.395
20 2
where P  0   e  0.135 ,
0!
c) Here, Tmean  Th  2 min
Therefore the probability that a call last more than 4 minutes is
P T  4   e4/2  0.135
 Probability of an arriving call being delayed,

AC
Pd  (1.26)
 A  C 1 Ak
A  C !1   
C

 C  k 0 k !
The probability that a delayed call will have to wait longer than t seconds is
 (C A) t 
P  delay  t | delay  0   exp    (1.27)
 Th 
The probability that a call is delayed more than t seconds,

P(delay  t )  P(delay  0) P  delay  t | delay  0  (1.28)

*** Pd  P(delay  0) and delayed calls are lost calls.

** Pd is the measure of GOS in LCD system and it is Erlang C formula.

Prob 1.13

Let us consider a cluster of 4 cells. There are 60 channels in all. Area of each cell is 5 sq.
Km. The probability call being delayed is 5%. If the traffic intensity and arrival rate per user
is 0.029 Erlang and 1call/ hr, find
a) Number of user per sq. Km. can be supported?
b) What is the probability that a delayed call will have to wait more than 10 sec.
c) What is the probability that a call is delayed for than 10 sec.

Soln.

a) N  4; Number of channels  60; Ai  0.029; i  1call / hr

Therefore the number of channels per cell = 60/4=15.
From Erlanng C chart, A  9.044  9 for Pd  5% and C  15
A 9
Hence, the number of user per cell can be supported U    310.344  310
Ai 0.029
The system can support 310/5=62 users per sq. Km.

Ai0.029
b) Here, Th    3600  104.4Sec
i 1
The probability that a delayed call will have to wait more than 10 sec
  (C A) t 
P  delay  10 | delay  0   exp   
 Th 
 (15  9)10 
 exp    0.5633
 104.4 
c) The probability that a call is delayed more than 10 sec
P(delay  10)  P(delay  0) P  delay  10 | delay  0 
 0.05  0.5633  0.0282

Here, P(delay  0)  5%  0.05

The trunking efficiency can be written as,

Traffic non  blocked 1  Pb  A

  (1.29)
Capacity C

Prob 1.14

a) Let C  100; Pb  2%; Ai  0.1 Erlang . Find efficiency.

Soln.
Erlang B chart, A  87.972


1  0.02  87.972  86.16%
100
b) Now if the number of channels are divided into two groups of 50 channel. Find trunking
efficiency.
Soln.
C  50; Pb  2%; Ai  0.1 Erlang . Find efficiency.
Erlang B chart, A  40.255

1  0.02  40.255  78.8%
50

*** One observations can be made that if the number of available channels in each cell
is splitted, trunking efficiency degrades which in turn increases blocked calls, missed
call and dropped calls.