Frequency reuse
1
Cellular Networks
Objectives
▪ Understand the advantage of a cellular network
topology
• Acquire basic methods for capacity planning
Outline
▪ The cellular network concept
▪ Frequency planning in cellular networks – uniform traffic
▪ Methods to increase network capacity
2
Frequency Reuse
• Bandwidth (spectrum) is scarce
– Tradeoff between transmission-capacity and reception-
– quality that best utilize a given spectrum
• If the total BW assigned is B , f is the frequency assigned
per channel.
BW =B Hz total spectrum
• channels f Hz per channel
• The number of channels a single base station
3
• can accommodate =BW/ f
Frequency Reuse (cont.)
Transmission power attenuates with distance
Reuse channel frequency is sufficiently apart
Sufficiently apart distance
4
Frequency Reuse
• Partition the service region into cells
A cell comprises the BTS coverage area
Mobiles at each cell are served by cell’s BTS
5
Frequency Reuse
• Hexagon is a geometrical shape that is most close to a circle
• that covers a region w/o overlapping
• Commonly accepted abstraction for resource and capacity
• planning in cellular networks
• Characterized by its radius R
• We need to study R
❑ For a given traffic (calls/unit area), how to select R ?
❑ How to select transmitter power as a function of R ?
❑ How to allocate frequencies (channels) among cells ?
6
Frequency Reuse
The Reuse Principle for Uniform Traffic
• Uniformly distribute all channels to a cell cluster of size N
❑ Example: A cluster with N=7
B
❑ 1/N is called “frequency reuse factor”
G C
❑ Keep co-channel cells as far apart A
F D
The feasible N are determined from: E
(3.1) N = i +i j + j ,
2 2
i, j 0 integers
j
Examples: N=3,4,7,9,12 1200
i
7
Frequency Reuse
The Reuse Principle
B
Replicate cluster keeping
G C
co-channel cells as far
A
apart
F D
Transmission capacity: E B
C=S N B G C
G C A
S – Total no. of channels
N – Cluster size A F D
F D E
Lowering N E
❑ Increases capacity C
❑ Increases co-channel interference
8
Frequency Reuse
Trading off Capacity and Interference
Measurements show that the average received power
approximately follows the “exponent decay law”
Pr (d ) = P0 (d d0 ) − n
(3.3)
Pr' (d ) = P0' −10n log(d d0 )
(Power in dB)
d0 d
P0 - Power received at reference point Fig. 3.8
n - Path loss exponent (typically between 2 – 4)
9
Frequency Reuse
Trading off Capacity and Interference
• A common measure for co-channel interference is the
• Signal Interference Ratio (SIR), namely
❑ The ratio between the received signal power and the co-channel
interference power
❑ Exercise: Verify that the worst case SIR is given by:
n
1 D
SIR =
6 R
No. of cells in the 1st tier of
co-channel interfering cells
10
Frequency Reuse
Trading off Capacity and Interference
If all cells have the same size and base stations transmit in
the same power, then
❑ SIR is independent of the transmitter power
Exercise: Verify that the “co-channel reuse ratio” satisfies
D
(3.5) = 3N
R
Hence, the “capacity” and the worst SIR become
1
(3.6) C=S N ; SIR = (3 N ) n / 2
6
11
Frequency Reuse
Trading off Capacity and Interference (cont.)
A sensible tradeoff rule is to maximize C = S N
Subject to: 1
SIR = (3 N ) n / 2 SIR (3.7)
6 target
Thus, we take the smallest N that satisfies Eq. (3.7)
N 3 4 7 9 12
n=4
SIR (dB) 11.3 13.8 18.66 20.85 23.34
❑ Example: US AMPS cellular system requires SIR of at least 18
dB
Exercise: Refine the SIR calculation ([Rappaport, Ch. 3.5.1] )
12
Frequency Reuse
Cell Radius Selection
• From Eq. the selectable N above is independent of the cell
radius R
• The radius R is determined by other considerations
❑ Base station cost drives to large R
❑ Mobile battery power however, limits R by the constraint
−n
R Ptarget
n
P0
d
P R d0
0 target or P
0
Ptarget - Minimum received power at base station
13
Frequency Reuse
The Micro-cell Concept
• A smaller R generally implies
❑ Lower transmitter power
❑ Better area coverage
❑ Higher capacity
These are the basic principles of micro-cells discussed below.
• No. of cells per unit area however, is 1 R2
❑ Thus, halving R results in 4 times more cells
14
Frequency Reuse
The Micro-cell Concept
• Traffic is not uniform in general and some spots need more
allocated channels than others
• One way to resolve it is to:
❑ Add base stations at hot spots which transmit at a lower power
❑ Without changing the frequency assignment to the macro-cells
N =3
3 micro-cells
Fig. 3.10
15
Frequency Reuse
The Micro-cell Concept
The transmit power at the micro-cells is determined by its
radius and Ptarget through the equation
= cPt1 R − n = cPt2 (R k ) , for some c
−n
(3.9) P
target
Pt1 - transmission power in the macro-cell
Pt 2 - transmission power in the micro-cell
For k=2 and n=4 we have Pt2 = Pt1 16
16
Frequency Reuse
The Micro-cell Concept
• Frequency planning
with micro-cells is
not obvious
• In a ”3-cluster cell
plan” we may
embed ”6 micro-cell N =3
clusters”
Fig. 3.11: An oversimplified example
6 micro-cells with radii R/2
17
Frequency Reuse
The Micro-cell Concept
• In a “6-cluster” cell
plan” we may
embed “6 micro-cell
clusters” N=6
An oversimplified example
6 micro-cells with radii R/2
18
Frequency Reuse
Sectoring
• Cell splitting into micro-cells increases capacity by rescaling
❑ It decreases R while keeping D/R constant.
❑ Hence packing more channels per unit area
• “Sectoring” decreases D/R while keeping R constant
❑ Base station uses several directional rather one omni-directional
antenna, which increases the receiver’s SIR
1
2 6 1 2
3 5 3
4
3 sectors per cell 6 sectors per cell
19
Frequency Reuse
Sectoring
With 3 sectors per cell and a cluster size of 7, the number of
1st tier interfering co-channel cells reduces from 6 to 2, and
with 6 sectors, to 1.
600
1200
20
