Lecture 2

Communications Systems
Mobile cellular communications systems
Lecture Outline

 Co-channel Interference (CCI) “cont’d”

 Adjacent Channel Interference (ACI)

 Trunking and Grade of Service

 Improving Cellular System Capacity

◼ Cell splitting

2
Co-channel Interference (CCI)- Forward Link
 Cause: Frequency Reuse
◼ Many cells in a given coverage area use the same set of channel
frequencies to increase system capacity (C)
◼ Co-channel cells → cells that share the same set of frequencies
◼ Both VC & CC traffic in co-channel cells is an interfering source
to mobiles in several different cells

RL: UPLINK

FL: DOWNLINK

3
Possible Solutions?

 1) Increase base station Tx power: to improve radio signal

reception?
 this will also increase interference from co-channel cells by the
same amount
 if all cell sizes, transmit powers, and coverage patterns are
almost the same → co-channel interference is independent of Tx
power
 no improvement

✓ 2) Physically separate: co-channel cells by some minimum

distance to provide sufficient isolation due to propagation of
radio signals?
✓ but if you increase the distance between co-channel cells (D),
you also need to increase cell radius (R) to maintain the same
frequency reuse pattern

4
Co-channel Reuse Ratio (Q)
 co-channel interference depends on:
◼ R : cell radius
◼ D : distance to base station of nearest co-channel cell
✓ if (D ↑ / R ↑) then spatial separation relative to cell coverage area ↑
◼ improved isolation from co-channel R Cluster
RF energy and reduces CCI
 Q=D/R: F7 F2
co-channel reuse ratio for hexagonal geometry
F6 F1
F1 F3
(i,j) N 𝑸
𝐷 F5 F4 F7 F2
𝑄= = 3𝑁 (1,1) 3 3
𝑅
(1,2) 7 4.58 F6 F1
F1 F3
(2,2) 12 6
F5 F4
(1,3) 13 6.24

5
Quality vs. Capacity

 Fundamental tradeoff in cellular system design:

◼ 𝑄 = 𝐷/𝑅 = 3𝑁

✓ small 𝑄 = 3𝑁 → small cluster size (N)→ more frequency reuse

→ larger system capacity (𝐶) → GREAT

✓ large 𝑄 = 𝐷/𝑅 → large cell separation (D/R)→ decreased co-channel

interference (CCI) → improved transmission quality → GREAT

◼ Tradeoff: Capacity vs. Voice Quality

The higher the capacity for a given geographic area, the poorer the
quality and vice-versa. (makes sense!!)

6
Co-channel Interference Computation - forward link

First tier cochannel

Second tier cochannel Base Station
Base Station

R
D6
D5
D1

D4 Mobile Station
D2

D3

Serving (Home) Base Station

7
Co-channel Interference Computation – forward link
 Signal to Interference Ratio (SIR),

𝑆 desired signal power from 𝒉𝒐𝒎𝒆 base station

𝑆𝐼𝑅 = =
𝐼 signal power from 𝒄𝒐 − 𝒄𝒉𝒂𝒏𝒏𝒆𝒍 base stations

𝑆
◼ 𝑆𝐼𝑅 = 𝑖𝑜 (2.5)
σ𝑖=1 𝐼𝑖

◼ S : desired signal power from home BS

◼ io : number of co-channel interfering cells
◼ Ii : interference power from co-channel cell’s BS

8
 From propagation: 𝑃𝑟 = 𝑃0 𝑑/𝑑0 −𝑛 ⇒ 𝑆 ∝ 𝑅−𝑛 , 𝐼𝑖 ∝ 𝐷𝑖−𝑛 ,
𝑛: path loss exponent

Assuming base stations transmit the same power and the pathloss exponent, n, is the same

𝑅 −𝑛
 𝑆𝐼𝑅 = 𝑖
𝑜 𝐷 −𝑛
σ𝑖=1 𝑖

Path loss exponent (n)

Line of sight (LOS) 2
Urban 2:4
◼ Di : distance from ith interferer BS to mobile

◼ having the same n throughout the coverage area means radio propagation
properties are roughly the same everywhere

9
 Now if we consider only the first-tier (or layer) of co-channel cells
◼ assume only these provide significant interference

 And assume interfering base stations are equidistant from the desired
base station (all at distance ≈ D) then

𝑅 −𝑛 𝐷/𝑅 𝑛
 𝑆𝐼𝑅 = =
𝑖𝑜 𝐷 −𝑛 𝑖𝑜

𝑛
3𝑁 𝑄𝑛
 𝑆𝐼𝑅 = = (2.9)
𝑖𝑜 𝑖0

 Equation (2.9) relates SIR to the cluster size 𝑁, which relates to the
overall system capacity.
10
Example

 Determine the suitable cluster size (N), required to achieve 𝑺𝑰𝑹 ≥ 𝟏𝟖 𝒅𝑩

in the U.S. AMPS cellular system, assuming a path loss exponent 𝒏 = 𝟒?
𝑛
3𝑁
Solution: 𝑆𝐼𝑅 = ,
𝑖𝑜

◼ The number of first-tier interfering cells, 𝑖0 = 6

◼ Solving (2.9) for N, noting that 18 𝑑𝐵 → 63.1

2 2
𝑆𝐼𝑅 𝑖0 𝑛 63.1 × 6 𝑛
𝑁= =
3 3

𝑁 = 6.204 ⟶ 𝑁 = 7
11
`
𝑛
3𝑁
 Many assumptions are involved in deriving 𝑆𝐼𝑅 =
𝑖𝑜
◼ same Tx power
◼ hexagonal geometry
◼ n is the same throughout the coverage area
◼ Di ≈ D (all interfering cells are equidistant from the base station
receiver)
◼ optimistic results in many cases

◼ In real-life situations, in the previous example, in order to achieve

𝑆𝐼𝑅 ≥ 18 𝑑𝐵, the actual cluster size would NOT be 𝑁 = 7 → 𝑁 =
12 (subjective tests determine the required SIR)

12
I. Worst-case SIR – forward link

 Low SIR is usually in the worst-case when a mobile is at the cell edge

 Low signal power from

home base station &
high interference power
from other cells.

13
Worst-case SIR

 For pathloss 𝑛, the worst-case SIR

𝑅−𝑛
𝑆𝐼𝑅𝑤𝑐 =
2(𝐷 − 𝑅)−𝑛 + 2(𝐷 + 𝑅)−𝑛 + 2𝐷−𝑛

In terms of co-channel interference ratio:

𝑄 = 𝐷/𝑅

1
𝑆𝐼𝑅𝑤𝑐 =
2 𝑄 − 1 −𝑛 +2 𝑄 + 1 −𝑛 + 2𝑄−𝑛

◼ SIR improves as 𝑄 increases

14
Comments – design flow
𝑛
3𝑁
 Consider 𝑁 = 7, 𝑛 = 4, solving 𝑆𝐼𝑅 = yields 𝑆𝐼𝑅 = 18.6 𝑑𝐵
𝑖𝑜

 The worst-case SIR for 𝑁 = 7 (𝑄 = 4.6) is:

𝑆𝐼𝑅𝑤𝑐 = 49.56 (17 𝑑𝐵 < 18.6 𝑑𝐵)

 In order to design the system on the worst-case SIR, cluster size should be
increased to 𝑁 = 12 (𝑖 = 𝑗 = 2).

 This results in significant decrease in capacity since spectrum utilization would be

1/12 within each cell, i.e., a reduction of 7/12 in capacity would not be tolerable.
(worst-case scenario rarely occurs and it is not worth it to sacrifice capacity)

Co-channel interference (𝑪𝑪𝑰) → link performance (𝑺𝑰𝑹) → dictates

frequency reuse plan (𝑵) → determines the overall capacity of the system (𝑪)
15
Example

 If 𝑺𝑰𝑹 = 𝟏𝟓 𝒅𝑩 is required for a satisfactory forward channel performance.

What is the co-channel reuse ratio (Q) and the cluster size (N) for maximum
capacity if the pathloss exponent is
(a) n=4, (b) n=3?
 There are 6 co-channel cells in the first tier and at the same distance from
the mobile.
Hint: for maximum capacity we need the smallest cluster size that attains the required SIR

𝐷 𝑛
Solution: 𝑄= = 3𝑁, 𝑆𝐼𝑅 = 3𝑁 /𝑖0
𝑅

(a)- 𝑛 = 4, let 𝑁 = 7 → 𝑄 = 4.583 →

4.583 4
𝑆𝐼𝑅 = = 73.5 = 18.66 𝑑𝐵 > 15 𝑑𝐵
6
cluster size 𝑵 = 𝟕, can be used.
or solve for N --> then get the next greater value of N in the table of i and j
16
(b)- 𝑛 = 3, let 𝑁 = 7 → 𝑄 = 4.583
4.583 3
→ 𝑆𝐼𝑅 = = 16.04 = 12.05 𝑑𝐵 < 15 𝑑𝐵
6
So, cluster size 𝑵 = 𝟕 can not be used and need to be
increased.

The next cluster size is 𝑁 = 12, (𝑖 = 𝑗 = 2)

the corresponding co-channel ratio 𝑄 = 6
6 3
→ 𝑆𝐼𝑅 = = 36 = 15.56 𝑑𝐵 > 15 𝑑𝐵 (OK)
6
Cluster size 𝑵 = 𝟏𝟐, SHOULD be used.

As pathloss increases, SIR increases !!

therefore, if you decrease path loss, you should increase N to maintain SIR
17
Reverse link co-channel interference

 Forward link co-channel interference considers interference at the

mobile unit (transmission from BS to mobile unit)

 What about reverse link co-channel interference at BS?

◼ LESS IMPORTANT;
 because signals from mobile antennas (near the ground) don’t
propagate as well as those from tall base station antennas

 obstructions near ground level significantly attenuate mobile

energy in direction of base station Rx

 weaker because mobile Tx power is variable → base stations

regulate transmit power of mobiles to be no larger than necessary

18
II. Adjacent Channel Interference (ACI)

 Two major types of system-generated interference:

1) Co-Channel Interference (CCI) – discussed before

2) Adjacent Channel Interference (ACI)

 Adjacent Channel Interference (ACI): adjacent in frequency

◼ Imperfect Rx filters allow energy from adjacent channels to leak
into the passband of other channels (bleed-over)

desired filter response actual filter response

19
Adjacent Channel Interference (ACI)

 This affects both forward & reverse links

 Forward Link → base station-to-mobile

◼ interference at mobile Rx from a Tx, another mobile or another
base station that is not the one the mobile is listening to, when
mobile Rx is away from the serving base station.

◼ signal from home base station is weak and others are somewhat
strong.

 Reverse Link → mobile-to-base station

◼ interference at base station Rx from nearby mobile Tx when
desired mobile Tx is far away from home base station

20
Near-Far effect
 ACI is primarily from mobiles in the same cell
◼ interfering source is near some Rx when desired source is far away

desired Not desired

MS BS MS1
2
Received signal
strength

Distance 0 Distance
MS2 d2 BS d1 MS1

21
Minimizing ACI

 Careful Channel Assignment (Frequency Planning)

◼ Within and across cells

 Use high-Q cavity filter at each base station receiver

 Proper choice of modulation schemes

 Efficient power control

22
Minimizing ACI: I-Channel Assignment (Frequency Planning)

 Careful channel assignment to users within the same cell.

A cell need not be assigned channels which are all adjacent in
frequency.
◼ don’t allocate channels within a given cell from a contiguous band
of frequencies
 for example, use channels 1, 4, 7, and 10 for a cell.
 no contiguous channels

𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦
𝟏 𝟒 𝟕 𝟏𝟎

𝒕𝒐𝒕𝒂𝒍 𝒄𝒂𝒉𝒏𝒏𝒆𝒍𝒔

23
Minimizing ACI: I- Channel Assignment (Frequency Planning)

 Maximize channel separation between neighboring cells

◼ separation of as many as N channel bandwidths at least N
◼ some schemes also seek to minimize ACI from neighboring cells by
not assigning adjacent channels in neighboring cells

y3ny:
cluster has 5 cells
cell 1 has channels A B C
cell 2 neighbouting to cell1 has A+5, B+5, C+5
by minimum --> N channels separation
(y3ny (B+5)from cell 2 - (A)from cell 1 >= 5)

24
Advanced Mobile Phone System (AMPS) channel allocation
 N=7-cell reuse, 3 sectors/cell An example to minimize ACI
 ≈ 19 VC/ sector
 57 = 19 × 3 VC/ cell At least 3N=21
channel BW
 Only 1 CC/sector 3C separation within
the same sector
3A
At least N=7 3B
channel BW
4C 2C
separation
between 4A 2A
sectors within 2B
the same cell 4B 1C
1A
5C 1B 7C
5A 7A
5B 6C 7B
6A
6B

25
19

Not used

26
AMPS system channel allocation
 Originally 666 channels, then 10 MHz of spectrum (166 channels) was
added → 666+166 = 832 channels

 395 VC + 21 CC per service provider (A & B)

total number of channels 𝟖𝟑𝟐 = (𝟑𝟗𝟓 × 𝟐 = 𝟕𝟗𝟎 𝑽𝑪 + 𝟐𝟏 × 𝟐 = 𝟒𝟐 𝑪𝑪)
3*7 sectors per cluster --> each sector has a VC group
 21 VC groups with ≈ 19 channels/sector
◼ at least 21 channel separation for each sector (within the sector)

 for N = 7 → 3 VC groups (sectors)/cell

◼ For example, choose groups 1A, 1B, and 1C for a cell – so channels
1, 8, 15, 22, 29, 36, etc. are used.
◼ ∴ ≈ 19 × 3 = 57 channels/cell
◼ at least 7 channel separation between sectors within a cell.
(between a sector and another)

27
Frequency Reuse for CC
 to have high quality on control channels, 21 cell reuse is used for
CC’s
◼ instead of reusing a CC every 7 cells, as for VC’s, reuse every 21
cells
◼ greater frequency separation between control channels, so less CCI

CC reuse = total number of CC per cell

28
Minimizing ACI: II-Filters and Modulation

 Use high quality filters in base stations

◼ better filters are possible in base stations since they are not constrained
by PHYSICAL SIZE & POWER as much as in the mobile Rx

◼ makes reverse link ACI less of a concern than forward link ACI

 Proper choice of modulation schemes

◼ different modulation schemes provide less or more energy outside
their passband.

29
Minimizing ACI: III- Power Control
 base station & MSC constantly monitor mobile received signal
strength (RSS)

 mobile Tx power varied (controlled) so that smallest Tx power

necessary for a quality reverse link signal is used (lower power for
the closer the mobile is to the base station)

 also helps battery life on mobile

 dramatically improves adjacent channel SIR ratio, since mobiles in

other cells only transmit at high enough power (not at full power).

 Very important in CDMA systems where all users share the same
spectrum

30
III. Trunking & Grade of Service (GoS)

 Trunked radio system: radio system where a large number of users

share a pool of channels

 In dynamic channel assignment: channel is allocated on demand &

returned to channel pool upon call termination

31
Trunking theory

 exploits the random behavior of users so that fixed number of

channels can accommodate large number of users

 Trade-off between the number of available channels that are

provided and the likelihood of a particular user finding no channels
available during the busy hours of the day.

 efficient use of equipment resources → savings

 disadvantage is that some probability exists that mobile user will be

denied access to a channel
 BLOCKED CALL : access denied → “blocked call cleared BCC”
 DELAYED CALL : access delayed by call being put into holding
queue for specified amount of time → “blocked call delayed BCD”

32
Grade of Service (GoS)

 GoS : a measure of the ability of a user ACCESS to a trunked system

during the BUSIEST hour

◼ specified as the likelihood that a call is BLOCKED or DELAYED

◼ designed to handle the busiest hour → typically at rush hour

◼ Erlang : measure of TRAFFIC INTENSITY

 A channel kept busy for one hour is defined as one Erlang
 e.g., 0.5 erlangs = 1 channel is occupied for 30 minutes during
one hour

33
Trunking theory definitions
 Set-up time: time required to allocate a trunked radio channel to a user

 Blocked call: a call which cannot be completed at time of request.

 Call Request Rate (𝝀): The average number of call requests/ unit time (calls/𝒔𝒆𝒄)

 Holding time (H): average duration of a typical call (seconds)

 Traffic Intensity: measure of channel time utilization, which is the average

channel occupancy (Erlangs).

 Load: Traffic intensity across the entire trunked radio system (Erlangs)

 Grade of Service (GoS): A measure of congestion which is specified as the

probability of a call being blocked (Erlang B), or the probability of a call being
delayed for a certain a mount of time (Erlang C)
34
 Offered traffic intensity by each user 𝐴𝑢 = 𝜆𝐻 Erlangs

 λ = Average number of call requests per unit time

average arrival rate of new calls. (calls/min).
 H = average call duration (min)
 Offered: NOT NECESSARILY carried by the system
(some is blocked/ delayed)

 The total offered Traffic Intensity in a cell, A, for a system

containing U users; 𝐴 = 𝑈 𝐴𝑢 = 𝑈𝜆 𝐻 Erlangs

 In a C-channel trunked system,

Traffic intensity/ channel : 𝐴𝑐 = 𝐴/𝐶 = 𝑈 𝐴𝑢 /𝐶

 maximum possible carried traffic (C Erlangs) is the total number of

available channels that are busy all the time) 35
Types of trunked systems

 Blocked calls cleared (BCC): Erlang B formula

◼ offers no queuing for call requests (pass/ block)
◼ GoS is defined as the probability a call is blocked

 Blocked calls delayed (BCD): Erlang C formula

◼ Queue is provided to hold calls when the channel is not available
immediately.
◼ GoS is defined as the probability that a call is blocked after waiting
a specific time period in the queue.

36
Erlang B formula
◼ Calls are either admitted or blocked
𝐴𝐶 𝐶 𝐴𝑘
𝐺𝑜𝑆 = 𝑃𝑟 𝑏𝑙𝑜𝑐𝑘𝑒𝑑 𝑐𝑎𝑙𝑙 = ൘෍
𝐶! 𝑘=0 𝑘!
◼ A = total offered traffic in a cell
◼ C = total number of channels in trunking pool (e.g., a cell)
◼ AMPS is designed for GOS of 2% (2 out of 100 calls will be blocked)
◼ blocked call cleared (denied) → BCC

37
Total offered traffic (𝐴) for different GoS
 Example: for GoS=0.5%, number of channels 𝐶 = 4, the total offered traffic
𝑨 = 𝟎. 𝟕𝟎𝟏 𝑬𝒓𝒍𝒂𝒏𝒈𝒔

 Note that twice the capacity can support much more than double the load
(twice the number of Erlangs).
38
 Graphical form of Erlang B formulas

𝑮𝒐𝑺

𝑨 = 𝑼𝑨𝒖

39
Trunking Efficiency

 A measure of the number of users which can be offered a

particular GoS with a particular configuration of fixed channels.

 Number of users (𝑈)

= total offered traffic in a cell (𝐴)/ offered traffic per user (𝐴𝑢 )

 trunking efficiency in terms of number of users per channel =𝑈/𝐶

 Trunking efficiency is an issue in both FDMA/TDMA systems and in

CDMA systems (where the capacity limit is the number of possible
codes and the interference levels).

40
 the allocation of channel groups can substantially change the
number of users supported by trunked system
◼ As trunking pool size ↑, trunking efficiency ↑.
◼ 10 channels trunked together support (60%) more traffic than
two 5-channel trunks at 1% GoS

41
Example

 How many users can be supported for 0.5% blocking probability for
the following number of trunked channels in a blocked calls cleared
system?
(a) C=1, (b) C=5, (c) C=10, (d) C=20, (e) C=100
Assume each user generates 𝑨𝒖 = 𝟎. 𝟏 Erlangs of traffic?
Solution:
𝐴𝑢 = 0.1, 𝐺𝑜𝑆 = 0.005 , total number of users 𝑈 = 𝐴/𝐴𝑢
From
# Erlang 𝑮𝒐𝑺find 𝑨: A (from Table 3.4 OR
𝑨𝒖 B chart, # users 𝑼 = 𝑨/𝑨𝒖
𝑪𝒉𝒂𝒏𝒏𝒆𝒍𝒔 Erlang B chart)

Trunking efficiency
1 0.1 0.005 0.005 𝑈 = 0.005/0.1 ≈ 1 user
5 0.1 0.005 1.13 𝑈 = 1.13/0.1 ≈ 11 users
10 0.1 0.005 3.96 𝑈 = 3.96/0.1 ≈ 39 𝑢𝑠𝑒𝑟𝑠
20 0.1 0.005 11.1 𝑈 = 11.1/0.1 ≈ 110 𝑢𝑠𝑒𝑟𝑠

100 0.1 0.005 80.9 𝑈 = 80.9/0.1 ≈ 809 𝑢𝑠𝑒𝑟𝑠

42
Example
 Three competing trunked mobile networks (systems A, B, and C)
provide cellular service in this area. System A has 394 cells with 19
channels each, system B has 98 cells with 57 channels each, and
system C has 49 cells with 100 each. Find the number of users that
can be supported at 2% blocking if each user averages 2 calls/ hour at
an average call duration of 3 minutes.
Solution: GoS=0.02,
3
traffic intensity/ user 𝑨𝒖 = 𝝀𝑯 = 2 × = 0.1 𝐸𝑟𝑙𝑎𝑛𝑔𝑠.
60
From Erlang B chart, get the total offered traffic 𝑨 for each system
system # cells # channels/ cell Total offered Number of users / Total number of

Trunking efficiency
(C) traffic 𝑨 = 𝑼𝑨𝒖 cell 𝑼 = 𝑨/𝑨𝒖 subscribers
A 394 19 12 Erlangs 120 120 × 394
= 47280
B 98 57 45 Erlangs 450 450 × 98
= 44100

C 49 100 88 Erlangs 880 880 × 49

= 43120
43
Example

 A city has an area of 1300 square miles and is covered using a N=7-cell
reuse pattern. Each cell has a radius of 4 miles. The city is allocated 40
MHz of spectrum with a full duplex channel BW=60 kHz.
Assume a GoS of 2% for an Erlang B system.
If the offered traffic per user 𝑨𝒖 = 𝟎. 𝟎𝟑 Erlangs. Determine:

(a)- the number of cells in the service area

(b)- the number of channels per cell
(c)- traffic intensity of each cell
(d)- the maximum carried traffic
(e)- the total number of users that can be served for 2% GoS.

44
Solution
coverage area = 1300 sq mi, cell radius 𝑅 = 4 miles, cell area = 2.598 × 𝑅2 = 41.57 sq mi.

𝑐𝑜𝑣𝑒𝑟𝑎𝑔𝑒 𝑎𝑟𝑒𝑎 1300

(a)- Total number of cells 𝑁𝑐 = = 41.57 = 𝟑𝟏 𝒄𝒆𝒍𝒍𝒔.
𝑐𝑒𝑙𝑙 𝑎𝑟𝑒𝑎

(b)- number of channels/ cell

allocated spectrum 40×106
𝐶= = 60×103×7 = 𝟗𝟓 channels/cell
channel BW×reuse factor

(c)- from Erlang B chart, given 𝐶 = 95, 𝐺𝑜𝑆 = 0.02 →

traffic intensity/cell 𝐴 = 𝟖𝟒 𝑬𝒓𝒍𝒂𝒏𝒈𝒔/cell

(d)- maximum carried traffic = number of cells × traffic intensity per cell
= 31 × 84 = 𝟐𝟔𝟎𝟒 Erlangs

(e)- number of users = traffic per cell × number of cells/ traffic per user
𝐴 84×31
𝑈 = 𝐴 × 31 = = 𝟖𝟔𝟖𝟎𝟎 𝒖𝒔𝒆𝒓𝒔
𝑢 0.03

45
Erlang C formula (reading only)
 blocked calls are delayed → BCD → put into holding queue

𝐴 𝐶−1 𝐴𝑘
Pr 𝑑𝑒𝑙𝑎𝑦 > 0 = 𝐴𝐶 / 𝐴𝐶 + 𝐶! 1 − ෍
𝐶 𝑘=0 𝑘!

 GOS is probability that a call will still be blocked even if it spends

time in a queue and waits for up to t seconds
Pr 𝑑𝑒𝑙𝑎𝑦 > 𝑡 = Pr 𝑑𝑒𝑙𝑎𝑦 > 0 𝑃𝑟 𝑑𝑒𝑙𝑎𝑦 > 𝑡|𝑑𝑒𝑙𝑎𝑦 > 0
Pr 𝑑𝑒𝑙𝑎𝑦 > 𝑡 = Pr 𝑑𝑒𝑙𝑎𝑦 > 0 exp(− C − A t/H)

The average delay 𝑫 for all calls in a queued system is given by

𝐷 = Pr 𝑑𝑒𝑙𝑎𝑦 > 0 𝐻/(𝐶 − 𝐴)

46
 Graphical form of Erlang C formulas (reading only)

𝑨 = 𝑼𝑨𝒖

47
Example (reading only)
A hexagonal cell within a 4-cell system has a radius 𝑹 = 𝟏. 𝟑𝟖𝟕 𝒌𝒎. A total of 60
channels are used within the entire system. If the load per user is 𝑨𝒖 = 𝟎. 𝟎𝟐𝟗
Erlangs, and 𝝀 = 𝟏 call/hour. Compute the following for an Erlang C system that
has a 5% probability of a delayed call:
(a)- how many users per square kilometer will this system support?
(b)- what is the probability that a delayed call will have to wait for more than 10
sec?
(c)- what is the probability that a call will be delayed for more than 10 seconds?
Solution:
Area covered per cell = 2.598𝑅2 = 2.598 × 1.387 2 = 5 𝑘𝑚2 .
𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑐ℎ𝑎𝑛𝑛𝑒𝑙𝑠 60
Number of channels per cell (C) = = = 15 𝑐ℎ𝑎𝑛𝑛𝑒𝑙𝑠.
𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑐𝑒𝑙𝑙𝑠 𝑝𝑒𝑟 𝑐𝑙𝑢𝑠𝑡𝑒𝑟 4
(a)- from Erlang C chart, for 5% probability of delay with 𝐶 = 15 →
𝐴 9
traffic intensity 𝐴 = 9 𝐸𝑟𝑙𝑎𝑛𝑔𝑠 → number of users = = = 310 users.
𝐴𝑢 0.029
number of users 310 𝑢𝑠𝑒𝑟𝑠
Number of users per square km = = = 62
covered area 5 𝑠𝑞 𝑘𝑚

48
(reading only)
(b)- given arrival rate 𝜆 = 1 𝑐𝑎𝑙𝑙/𝐻𝑜𝑢𝑟
𝐴𝑢
Holding time 𝐻 = = 0.029 hour = 104.4 seconds.
𝜆
The probability that a delayed call will have to wait for more than 10 sec is

Pr delay > t delay > 0 = exp(− C − A t/H)

15 − 9 10
= exp − = 56.29 %
104.4

(c)- Given Pr delay > 0 = 5% = 0.05

Probability that a call is delayed more than 10 seconds

Pr 𝑑𝑒𝑙𝑎𝑦 > 10 = Pr 𝑑𝑒𝑙𝑎𝑦 > 0 Pr[𝑑𝑒𝑙𝑎𝑦 > 𝑡|𝑑𝑒𝑙𝑎𝑦]

= 0.05 × 0.5629 = 2.81%

49
IV. Improving Cellular System Capacity

 As the demand on wireless service increases, the number of

channels eventually becomes insufficient.

◼ Need to provide more channels per unit coverage area

◼ Would like to have orderly growth
◼ Would like to upgrade the system instead of rebuild
◼ Would like to use existing towers as much as possible

 Techniques to improve system capacity

◼ A- Cell splitting
◼ B- Sectoring
◼ C- Zone microcell concept

50
A - Cell splitting - downscaling
 subdivide congested cells into several smaller cells
 increases number of times channels are reused in a given area
 must decrease antenna height & Tx power
 smaller coverage per cell results
 co-channel interference level is held constant
 Depending on traffic patterns the smaller cells may be
activated/deactivated in order to efficiently use cell resources

Large cell (low density)

small cell (high density)

microcell (very high density)

51
 Capacity is improved by down-scaling the system. By decreasing
cell radius R, and keeping the co-channel reuse ratio Q=D/R
unchanged, cell-splitting increases the number of channels/ unit
area, hence increases capacity
 each smaller cell keeps the same number of channels as the larger
cell, since each new smaller cell uses the same number of
frequencies
◼ this means that we keep the same cluster size

 capacity increases because channel reuse increases per unit area

(the whole coverage area remains unchanged)

 smaller cells are called “microcells”

52
Illustration for cell splitting – towers at corners

◼ Assume the area served by base station A is congested – high blocking

probability
◼ Base station A is surrounded by 6 new microcell base stations with the
original frequency reuse plan is preserved
◼ The radius of each new microcell is half of the original (R/2).

53
Microcell Power Adjustment
 The received power at the new and the old cell boundaries should
be equal in order to keep the frequency reuse plan unchanged.

 If the small cell has half the radius (R/2) of the large cell
 the transmit power of the large & small cells are 𝑃𝑡1 & 𝑃𝑡2

𝑃𝑟 𝑎𝑡 𝑜𝑙𝑑 𝑐𝑒𝑙𝑙 𝑏𝑜𝑢𝑛𝑑𝑎𝑟𝑖𝑒𝑠 ∝ 𝑃𝑡1 𝑅−𝑛

𝑅 −𝑛
𝑃𝑟 𝑎𝑡 𝑛𝑒𝑤 𝑐𝑒𝑙𝑙 𝑏𝑜𝑢𝑛𝑑𝑎𝑟𝑖𝑒𝑠 ∝ 𝑃𝑡2
2
𝑃𝑡1
⇒ 𝑷𝒕𝟐 = (𝑷𝒕𝟏 /𝟐𝒏 ) For path-loss 𝑛 = 4, we have: 𝑝𝑡2 =
16

The transmit power must be reduced by 12 dB in order to fill in the

coverage area using micro cells while maintaining the SIR requirements,

54
Advantages of cell-splitting

 Allows the system to grow without upsetting the channel allocation

scheme required to maintain the min co-channel reuse ratio 𝑸.

 system capacity can gradually expand as demand increases.

 only needed for cells that reach max. capacity → not all cells

 implemented when Pr[blocked call] > acceptable GOS

 Do not suffer the trunking inefficiencies.

55
Disadvantages of cell-splitting
◼ Number of handoffs/unit area increases (increased number of cells)
◼ more base stations → higher cost for real estate, towers, etc.

 What about NEIGHBOURING existing base stations?

 If kept at the same power, they would overpower
new microcells.
 If reduced in power, they would not cover their own cells.
 Solution 1:
◼ Use antenna down-tilting for neighboring existing BSs
 Solution 2:
◼ Use separate groups of channels.
◼ One group at the original power and another group at the lower power
◼ New microcells only use low power channels.
◼ Larger cell is usually dedicated to high-speed traffic ‘umbrella cell’
 As load growth continues, more and more channels are moved to
lower power.

56
Example

Assume each BS uses 60 channels regardless of cell size. If each original cell
has a radius of 1 km and each microcell has a radius of 0.5 km. Find the
number of channels contained in a 3 km by 3 km square centered around A
under the following conditions:
(a)- without the use of microcells
(b)- when the lettered microcell as shown in the figure
(c)- consider all microcells in the square

Assume cells at the edge of the square to be

contained within the square?

57
Solution

(a)- without microcells, the number of BS inside the square is five.

the total number of channels is 5 × 60 = 𝟑𝟎𝟎 channels

(b)- when the BS A is surrounded by six microcells, the total number of BS

inside the square is 5 + 6 = 11 BS
the total number of channels is 11 × 60 = 𝟔𝟔𝟎 channels.

(c)- when all BS are microcells, the number of BS inside the square is
5 + 12 = 17 BS.
the total number of channels is 17 × 60 = 𝟏𝟎𝟐𝟎 channels.

The capacity in this case has increased by 1020/300=3.4 times compared to (a)

58