You are on page 1of 7

Chapter 3

3.1, Prove that for a hexagonal geometry, the co-channel reuse ratio is given by Q=(3N)1/2,
where N=i2+ij+j2.
Generally, for N=i2+ij+j2, we can do the following to find the nearest co-channel neighbors of
a particular cell:
(1) move i cells along any chain of hexagons and then
(2) turn 60 degree counter-clockwise and move j cells.
From the following figure, using the cosine law, we have

D 2 = [i (2 R' )] + [ j (2 R' )] 2i (2 R' ) j (2 R' ) cos 120 0



R' =

R , therefore

D = 3i 2 .R 2 + 3 j 2 .R 2 + ij 3 R 2
= 3(i 2 + ij + j 2 ) .R = 3 N .R

Hence , Q =

= 3N

3.3, Show that the frequency reuse factor for a cellular system is given by k/S, where k is the
average number of channels per cell and S is the total number of channels available to the cellular
service provider.
Since S=kN, where N is the cluster size, we have N=S/k.
By the definition of frequency reuse factor, we have frequency reuse factor=1/N=k/S.
3.4, If 20MHz of total spectrum is allocated for a duplex wireless cellular system and each
simplex channel has 25kHz RF bandwidth, find:
(a) the number of duplex channels.
(b) the total number of channels per cell site, if N=4 cell reuse is used.
(a) 20MHz/[25kHz*2]=400channels
(b) 400/4=100
3.5, A cellular service provider decides to use a digital TDMA scheme which can tolerate a
signal-to-interference ratio of 15dB in the worst case. Find the optimal value of N for (a)

omnidirectional antennas, (b) 1200 sectoring, and (c) 600 sectoring. Should sectoring be used? If
so, which case (600 and 1200) should be used? (Assume a path loss exponent of n=4 and consider
trunking efficiency)
(a) Let i0 be the number of co-channel interfering cells, for omni-directional antennas, i0=6.

S ( 3N )n
> 15dB = 31.623 , N > 4.59
Assume n=4, we have I
Therefore , N = 7
(b) For 1200 sectoring, i0=2.

S ( 3N )n
> 31.623 , N > 2.65 , N = 3
(c) For 600 sectoring, i0=1,

S ( 3N )n
> 31.623 , N > 1.87 , N = 3
From (a), (b), and (c) we can see that using 1200 sectoring can increase the capacity by a
factor of 7/3, or 2.333. Although using 600 sectoring can also increase the capacity by the same
factor, it will decrease the trunking efficiency. Therefore we chose the 1200 sectoring.
a) Calls are not lost due to weak signal condition during handoff if:

distance travelled during handoff d min d HO

mobile speed
* d min received power at BS1 reaches Pr ,min
Pr ,min = 29 lg(d min ) d min = 10
* d HO received

Pr , min / 29

= 1083m


( 2)

power at BS1 reaches Pr ,HO

Pr ,HO = 29 lg ( d min ) d HO = 10 PHO / 29


Using ( 2),

/ 29

1083 10 r , Ho
4.5 sec onds
22.22(m / s )
Pr ,HO 86.8dBm


= Pr ,HO Pr ,min 1.2dB

( 4)
( 6)

b) If we set too large, several unnecessary handoffs will be requested and performed,
increasing the signaling traffic between the base stations and MSC. On the other hand, if

is too small, that is, Pr ,HO is only slightly greater than Pr ,min , there will not be
enough time to complete the handoff (especially for high speed MSs), and calls may be
lost due to weak signal condition.


= 400 channels
2 30kHz
= 100 channels/cell
Channels in each cell for a 4-cell reuse system=

Total available channels=

(b) The total carried traffic of 100 channels is obtained as 100 Erlangs.
90% of 100 Erlangs =90 Erlangs.

90 = U Au = U 0.1

U = 900users / cell
(c) The offered traffic is 90 Erlangs. The total channels are 100.
From Fig. 3-6, we can obtain that GOS is 0.03.
Therefore, the GOS is 3%.
(d) Each sector has 33.3 channels; Gos=3%.
From Fig.3-6, the maximum traffic provided by a sector is approximately 25
The maximum number of users that can be supported per cell is


25 / sec tor
3 = 750 users.

The number of cells is

50km 50km
= 500 cells.
5km 2

The total subscribers could be supported are 900users/cell500cells=450000users.

(f) In this cases, The total subscribers could be supported are 750users/cell
Since users are uniformly distributed over the area, each cell in the cluster is assigned the
same number of channels:

Nc =



Where, Nc=number of channels per cell, M= number of channels available in the system,
N=cluster size.
Given the number of channels per cell and the designed blocking probability Pb = 1% , we
can compute the maximum carried traffic per cell in Erlang (Cc) using the Erlang B formula and
the maximum carried traffic C in the system.

C = C c 84


Since each user offers a traffic of 0.04 Erlangs, the maximum number of users supported by

the system is

NU =



Table 2: The number of channels per cell (NC), carried traffic per cell (CC), total carried
traffic in the system (C), and maximum number of users in the system (NU), for cluster sizes N=4,
7 and 12, blocking probability 1%.

3.15, Exercises in trunking (queuing) theory:

(a) What is the maximum system capacity (total and per channel) in Erlangs when
providing a 2% blocking probability with four channels, with 20 channels, with 40 channels?
(b) How many users can be supported with 40 channels at 2% blocking? Assume
H=105s, =1 call/hour.
(c) Using the traffic intensity calculated in part (a), fine the grade of service in a lost
call delayed system for the case of delays being greater than 20 seconds. Assume that H=105s,
and determine the GOS for four channels, for 20 channels, for 40 channels.
(d) Comparing part (a) and part (c), does a lost call delayed system with a 20 seconds
queue perform better than a system that clears blocked calls?
(a) Given GOS=2%
For c=4 channels, from the Erlang B chart,
Atotal=1.1 Erlangs, Aperchannel=Atotal/C=1.1/4=0.275Erlangs.
For c=20 channels,
Atotal=14 Erlangs, Aperchannel=Atotal/C=0.7Erlangs.
For c=40 channels,
Atotal=31 Erlangs, Aperchannel=Atotal/C=0.775Erlangs.





= 1063users

For c=4 channels, Atotal=1.1 Erlangs, H=105seconds/call.
From the Erlang C chart, we have

Pr [delay > 0] 0.03


Pr [delay > 20] = Pr [delay > 0]. exp[(c Atotal ).20 sec onds / H ]
= 0.03 exp[(4 1.1) 20 / 105] 0.017

For c=20channels, Atotal=14 Erlangs, we have

Pr [delay > 0] 0.07


Pr [delay > 20] = Pr [delay > 0]. exp[(c Atotal ).20 sec onds / H ]
= 0.07 exp[(20 14) 20 / 105] 0.019
For c=40channels, Atotal=31 Erlangs, we have

Pr [delay > 0] 0.07


Pr [delay > 20] = Pr [delay > 0]. exp[(c Atotal ).20 sec onds / H ]
= 0.07 exp[(40 31) 20 / 105] 0.013
(d) From (c) we can see that the probability that a call will be delayed for more than 20
seconds in a lost call delayed system is less than 2% for all the different channel numbers.
Thus a lost call delayed system performs better than a system that drops blocked calls.
(a) The AMPS system is duplex.
Given total bandwidth BWtotal=50MHz, total number of channel N=832, we have the
bandwidth for each channel

BW =

BWtotal 50 106
= 60kHz

This bandwidth of 60kHz for the duplex channel is split into two one-way channels, a
forward channel and a reverse channel, each with a bandwidth of 30kHz. The forward
channel is exactly 45MHz higher than the reverse channel.
For Fforward=880.560MHz Fforward=Freverse-45=835.560MHz.
Given N=832, total number of control channels N control = 42 , we have total number of
voice channels N voice = N N control = 832 42 = 790 .

N voice, A = N voice,B =








N voice 790
= 395 channels.

Number of control channels for each carrier: N control, A = N control ,B =

(d) See example 3.3.

N control 42
= 21


For 7-cell reuse, N=7,


= 3N = 21 = 4.58 ,

D = 4.85R

For 7-cell reuse, N=4,


= 3N = 12 = 3.46 ,

D = 3.46R

a) Minimum SIR
In order to compute the minimum SIR at the mobile, we need to determine the number of
interfering BSs in each possible configuration, which can be done by inspecting Figures 1
and 2. Table 1 shows the number of interfering BSs in the first tier, when 3 sectors (BW=1200)
and 6 sectors (BW=600) are used, for cluster size N=3 and 4.
Using expression (1), we determine the minimum SIR (approximation) in each configuration
(path loss exponent n=4). Results are shown in Table 2.
Therefore, cluster size N=3 cannot be used, since the minimum SIR achieved is below
SIR=18.7dB. On the other hand, both configurations using cluster size N=4 are feasible,
regarding co-channel interference (assuming that a difference of 0.1 dB is negligible).
b) Maximum carried traffic per cell
Let us now compute the carried traffic per cell when sectoring is used.
As we know, each sector is assigned a subset of channels assigned to the cell. For example,
for cluster size N=3, each cell is assigned 300/3=100 channels. If 6 sectors are employed,
each sector is assigned 100/616 channels. Using Erlang B formula, we find that each sector
carries a maximum traffic of 9.83 Erlang at a blocking probability of 0.02. therefore, the
maximum traffic carried by each cell is 9.836=58.97 Erlangs. Repeating this procedure, we
can compute the maximum carried traffic per cell for other beam widths and cluster sizes.
Table 3 presents the results.
Table1Number of interfering BSs in the first tier (i0) when 3 sectors and 6 sectors are used:

Table 2: Minimum SIR achieved when sectoring is used, for cluster sizes N=3 and 4:

Table 3: Maximum carried traffic per cell when sectoring is used, for cluster sizes N=3 and 4.
300 channels available in the system, Pb = 2 % :

Figure 1. Cluster size N=3, three & six sectors

Figure 2. Cluster size N=4, three & six sectors