Transcript Lecture5(Interference and System Capacity)

WIRELESS COMMUNICATION
MOBILE COMMUNICATIONS
LECTURE 5
Tanvir Ahmad Niazi
[email protected]
Air University, Islamabad
1
Overview of the Previous Lecture



Comparison of Fixed Telephone versus Cellular Call
The Cellular Concept
Frequency Reuse
New Topics
Channel Allocation Schemes
Hand off Strategies
Interference and System Capacity
2
An Example of a Cellular Cluster
3
Capacity versus interference for
same size cell

Decrease N for More Capacity: If Cluster Size, N is decreased
while cell size remains fixed, more clusters are required to cover the
area (M increases). Therefore, Capacity increases.

Increase N for Less Interference: On the other hand, if N is
increased (large cluster size) means that co-channels are now
farther than before, and hence we will have less interference.

Value of N is a function of how much interference a mobile or a base
station can tolerate.
4
Frequency Reuse

Geometry of the hexagonal cells is such that to cover adjacent areas
completely, N can have only some of the values.

N should satisfy the relationship,

where i and j are positive integers. Typical values of N are 4,7,12
etc.
To draw the cell pattern given i and j:

1.
2.
Move i cells across hexagons.
Turn 60º CCW and move j cells.
5
Frequency Reuse (N=7, i=2, j=1)
6
Method of locating co-channel cells
in a cellular system
Example:
N=19
i=3
j=2
7
Channel Assignment
Strategies
8
Channel Allocation
Techniques



To satisfy the user, a channel needs to be available on
request.
Reasonable probability of call blockage.
Three types of algorithms for channel allocation:
Fixed channel allocation (FCA)
 FCA with Channel Borrowing
 Dynamic channel allocation (DCA)

9
Hand Off Strategies
10

HAND OFF:
“When a mobile moves into a different cell while a conversation is in
progress, MSC transfers the call to a new channel belonging to the new
base station”. (Hard handoff)

In modern wireless systems, there may be no change in the actual
channel; only the base station changes (Soft Hand off).

Mobile Switching Center needs to:
Identify the new base station
 Allocate new Voice and Control channels associated with the new base station.

11

Desired Behavior:





Prioritize Hand off to new call initiation
Successful transfer of call
Hand off should be as infrequent as possible
Hand off should be imperceptible (very slight)
Hand off Threshold Power Level:


There is a minimum power level of the signal at BS for acceptable
voice quality. (-90 to –100 dBm).
Hand off Threshold is defined D dBm above this level, so that the
system has time to process hand off.
12
Margin D needs to be neither too small,
nor too large



D too small  too short time to perform handoff, The call may be
dropped in the meanwhile.
D too large  unnecessary handoff burden. Hand off becomes more
frequent, and occurs even when Power is at the acceptable level.
Call will be dropped if
1.
2.

D is too small,
Delay in MSC processing because of:
High Traffic hours, New channel not available
13
Handoffs – the basics
When to Hand off? MSC needs to ensure:

When to Hand off? MSC needs to ensure:
that the power drop is not temporary: because of FADING.
 that MS is actually moving away from the current BS.


MSC measures signal strength for some time (average received signal
power). These calculations are performed to:
Avoid the unnecessary hand off,
 Complete the necessary hand off before call drops


The time available to decide hand off depends on the subscriber’s speed.

Speed of the user can be estimated from the slope of the short term
average Power. Steep curve Quick Hand off
15
First Generation Systems:






Signal Strength Measurements are made by base stations and supervised
by MSC.
Each base station constantly monitors power levels of its reverse voice
channels, which tells the relative location of the user.
Reverse Signal Strength Indicators: Power levels of all the calls in a cell.
Locator Receiver: Power levels of Users of neighboring cells are also
measured.
Both these pieces of information are passed to MSC for hand off decision.
In 1G: Hand off processing takes almost 10sec: that requires

D ~= 6-12 dB
16
Second Generation Systems: Digital
TDMA






Concept of Mobile Assisted Hand Off (MAHO)
Mobile station continually measures power levels of surrounding Base
Stations and reports these measurements to the serving Base Station.
Hand off is initiated when power level from BS of neighboring cell exceeds
that of the current cell, for certain time, or by a certain level
MAHO method ensures faster hand off since burden of MSC is shared by
Mobile Stations
MAHO is particularly suited for Micro-Cellular Systems.
In GSM: The time reduces to 1-2 seconds:  that needs
D ~= 0-6dB
17
Inter-System Hand Off

Definition: “-----When a mobile moves from one cellular system to

another, with different MSC”.
The situation occurs when
“Signal in the resident cell becomes weak, and no other cell within the
system can take the call”.

Issues:
Local Call becomes long-distance call
 Compatibility between two MSCs.

18
Prioritizing Hand off



Systems differ in methods and policies of hand off:
Some systems take hand off like a new call initiation: user will be more
annoyed in case of call drop than call blocking for some time.
Various Methods of Prioritizing Hand off have been devised and
implemented.
Guard Channel Concept



Reserve some channels exclusively for hand off--- do not use them for call
initiation.
Advantage is increased probability of successful hand off. Disadvantage is
lower capacity because of less number of channels for call initiation.
No disadvantage in Dynamic Channel Allocation
19
Queuing of Hand off:




Decreases the probability of forced termination of a call due to lack of
available channels.
Trade off between the “probability of forced termination” and “total
carried traffic”. That is,
“Probability of forced termination” decreases at the cost of reduced Total
Carried Traffic.
Queuing is possible because of the time available between the Threshold
power level and the Hand off power level.
Delay Time and Queue Size are related to service area.
Queuing doesn’t guarantee zero probability of call drop.
20
Practical Hand Off Strategies:




High Speed Vehicle VS Pedestrians
In micro-cells for more capacity, MSC may be over-burdened in case high
speed users changing the micro-cells quickly.
Obtaining a new Cell Site is not always practical: Zoning, Ordinances and
Laws are barrier to new cell sites.
The UMBRELLA CELL Concept: Large and Small cells are co-located. Install
small antennas etc for low-speed users for proper radio coverage. This
guarantees large area coverage as well.
21
The Umbrella Cell Approach
22
Umbrella cell approach:




Minimizes the number of hand offs for high speed users,
Provides additional micro-cell users for pedestrians.
A High speed user converting into a low speed user may be
switched to smaller micro-cell by the Base Station, without
MSC intervention.
Speed Estimation can be performed by slope of short term
average calculations of the received power, or by some more
more sophisticated algorithms.
23
Cell Dragging

Results from Pedestrian users who provide very strong signal
to base station.

For example, in Line Of Sight (LOS) case, power does not drop
to the hand off threshold level even when the user has moved
deep into the neighboring cell.

Solution: Hand Off Threshold and Radio Coverage Parameters
need to be adjusted.
24
Faster Hand Off
Faster hand off facilitates call “rescue”, that is in need of a hand
off

Hand off in modern systems is based on calculations of
additional parameters including: co-channel and adjacent
channel interference at Base Station or Mobile Station.

In CDMA systems: Same channel is used by all the cells. MSC
decides from which BS to take the signal and pass to PSTN.
(Soft Hand Off)
25
Interference and System
Capacity
26
Interference



It is a major limiting factor in the performance of cellular radio systems.
(In comparison with wired comm. Systems, the amount and sources of
interferences in Wireless Systems are greater.)
Creates bottleneck in increasing capacity
Sources of interference are:
1. Mobile Stations
2. Neighboring Cells
3. The same frequency cells
4. Non-cellular signals in the same spectrum



Interference in Voice Channels: Cross-Talk
Interference in Control Channels: missed/blocked calls
Urban areas usually have more interference, because of:
a) Greater RF Noise Floor,
b) More Number of Mobiles
27
Types
1) Co-Channel Interference (CCI)
2) Adjacent Channel Interference (ACI)
3) Other services: like a competitor cellular service in the same area
(1)




Co-Channel Interference and System Capacity
The cells that use the same set of frequencies are called co-channel cells.
The interference between signals from these cells is called Co-Channel
Interference (CCI).
Cannot be controlled by increasing RF power. Rather, this will increase
CCI.
Depends on minimum distance between co-channels.
28
 The yellow cells use the same set of
frequency channels, and hence, interfere
with each other
 In the cellular system there are 6 firstlayer co-channels



In constant cell size and RF power, CCI is a function of Distance between
the co-channel cells(D), and the size of each cell (R).
Increasing ratio D/R, CCI decreases.
Define Channel Reuse Ratio = Q = D/R
29

For hexagonal geometry, D/R can be calculated:

Smaller Q provides larger capacity, since that would mean smaller N.
(Capacity ∝ 1/N).
Larger Q improves quality, owing to less CCI.

For
N=3,
N=7,
N=12,
N=13,
Q=3,
Q=4.58,
Q=6,
Q=6.24
30
31

Signal-to-interference ratio

S is the power of the signal of interest and
interference.

The signal strength at distance d from a source is

That is, received signal power is inversely related to nth power of the
distance.

where n = path loss exponent
is the power of kth
32

Then we can express the SIR in terms of distance

where the denominator represents the users in neighboring clusters using
the same channel




Let
be the distance between cell centers. Then
Note how C/I improves with the frequency reuse N.
Analog systems: U.S. AMPS required C/I ~= 18dB For n = 4, the reuse
factor for AMPS is N ≥ 6.49, so N = 7.
Now, let us consider the worst case for a cluster size of N= 7. The mobile is
at the edge of the cell. Express C/I as a function of actual distances.
33
Worst Case Design
Worst case carrier-to-interference ratio
34

Increasing N from 7 to 12, brings C/I above the 18dB level. However, the
system capacity is decreased.

Reduction of capacity by 7/12 for taking care of the worst case situation
when CIR ~= 17.4 dB is not justified because this situation will occur very
rarely.

Conclusion: Co-Channel Interference controls the link performance which
then decides Frequency Reuse Plan, and System Capacity.
35
36
Channel Planning for Wireless
Systems

Judicious assignment of the appropriate radio channels a more difficult
problem in practice. Theoretical analysis provides a guideline.

In general 5% of the available spectrum is reserved for control channels

Frequency reuse of control channels more conservative.

In CDMA N=1, however, practical difficulties forces some kind of
frequency planning.
37
Adjacent Channel Interference



Interference from channels that are adjacent in frequency,
The primary reason for that is Imperfect Receive Filters which cause the
adjacent channel energy to leak into your spectrum.
Problem is severer if the user of adjacent channel is in close proximity. 
Near-Far Effect

Near-Far Effect: The other transmitter(who may or may not be of the
same type) captures the receiver of the subscriber.

Also, when a Mobile Station close to the Base Station transmits on a
channel close to the one being used by a weaker mobile: The BS faces
difficulty in discriminating the desired mobile user from the “bleed over”
of the adjacent channel mobile.
38
Near-Far Effect: Case 1

The Mobile receiver is captured by the unintended, unknown transmitter,
instead of the desired base station
39
Near-Far Effect: Case 2

The Base Station faces difficulty in recognizing the actual mobile user,
when the adjacent channel bleed over is too high.
40
Minimization of ACI
(1) Careful Filtering ---- min. leakage or sharp transition
(2) Better Channel Assignment Strategy




Channels in a cell need not be adjacent: For channels within a cell, Keep
frequency separation as large as possible.
Sequentially assigning cells the successive frequency channels.
Also, secondary level of interference can be reduced by not assigning
adjacent channels to neighboring cells.
For tolerable ACI, we either need to increase the frequency separation or
reduce the passband BW.
41
Example 2.3






In the US APMS system the total number of channels is: 416
Out if these 395 are used as voice channels, whereas 21 are
used a control channels.
Two operators share these 416 channels eqyally.
Channel allocation according to Table 3.2 (next slide). Using
subsets iA+
For N=7, each cell in a cluster will have 1 control channel,
neighboring clusters will have remaining 14 control channels.
Control channels will be thus be reused according to N=21
pattern.
42
43
Power Control for Reducing
Interference




Power levels transmitted by every subscriber unit are under constant
control by the Base Station.
Mobile Station should transmit minimum power to maintain quality link
on reverse channel.
This has benefits of longer Battery Life at the Mobile Station Reduced
reverse SIR.
In CDMA systems, it is extremely important to control the power, as the
neighboring cells are using the same channel.
44