Wireless Communications

Systems (EEE464)

Dr. Guftaar Ahmad
Office: Room 412
Email: guftaarahmad@comsats.edu.pk
In 1946, the first public mobile telephone is put into
service, known as improved mobile telephone service
(IMTS)
A single, high powered transmitter mounted on a tall tower
near the center of a metropolitan area, covering distance of
50km, with a fixed number of available channels
Good coverage
Impossible to reuse the frequency.
Service poor due to call blocking.
When metropolitan area grows, higher power is needed but
more subscribers are not able to connect.

Cellular Concept
Replacing a single, high power transmitter with many low power
transmitters, each providing coverage to only a small area
Each base station is allocated a portion of the total number of
frequency channels available
to the entire system, and
nearby base stations are
assigned different groups
of channels.
Neighboring cells are
assigned different groups
of channels in order to
minimize interference
The same set of channels is
then reused at different
geographical locations.
Without increasing the
number of the channels
Coverage expand easily
Cope with increasing
demand for service easily
Cellular Concept
Cellular Concept
In a wireless cellular system the service area is divided
into a collection of geographical areas called cells
Each cell has a Base Station (BS) at the center of the
cell (usually)
Wireless (mobile) telephones are connected to the Base
Station via a wireless link
A number of base stations are connected via wire-line
to a Mobile Switching Center (MSC)
Finally, the MSCs are connected via wire-line to the
Public Switched Telephone Network (PSTN)
Cell and Cluster
Frequency Planning:
The design process of selecting and allocating channel groups for all of
the BS within a system .
Footprint:
is the actual radio coverage of a cell and is determined from field
measurements or propagation prediction models. For systematic cell
planning, a regular shape is assumed for the footprint.
Coverage contour should be circular. To avoid overlap or leaving gaps,
only three regular shapes allowed: equilateral triangle, square, hexagon.



The hexagon has been chosen due to its maximum area coverage. The
fewest number of cells can cover a region.
For hexagons model:
centre-excited cell: the cell having the base station at the cell centre.
(Omnidirectional antenna)
edge-excited cell: the cell having base stations on the three of the six cell.
(Directional antennas)
Frequency reuse :
A cellular system which has a total of duplex channels.
channels are divided among cells, with each cell uses unique and
disjoint channels.
If each cell is allocated a group of channels, then .
Cluster: the cells that collectively use the complete set of available
frequency channels. Cluster size=
S
S
S =kN
N
k
N
N
Cluster Size=7
Capacity : If a cluster is replicated times within the
system, then
If cell size does not change and we reduce , then
the capacity increases. How?
If cell size decreases and we do not change , then
the capacity increases. Why?
Frequency reuse factor:
In order to connect without gap, the allowable must
satisfy , for non-negative .

C
C =MkN = MS
N
1=N
N =i
2
+ij +j
2
N
i; j
M
N
To find the nearest
co-channel neighbors:
move cells along
any chain of
hexagons
Then turn 60
counter-clock wise
and move cells
ii
jj
Some possible
value of
NN

Example: If a total of 33.6 MHz of bandwidth is
allocated to a particular FDD cellular system which uses
two 25 kHz simplex channels to provide full duplex
voice and control channels, compute the number of
channels available per cell if a system uses (a) four-cell
reuse; (b) seven cell reuse. (c) 12 cell reuse.
Total bandwidth = 33.6 MHz
Channel bandwidth = 25 KHz 2 = 50KHz/Duplex channel
Total available channel 33600/50 =672 channels
For N=4, No of channels available per cell = 672/4 = 168
For N=7, No of channels available per cell = 672/7 = 96
For N=12, No of channels available per cell = 672/12 = 56
Honework#1-1: continued on the previous examples
If 1.5 MHz of the allocated spectrum is dedicated to control
channels, determine an equitable distribution of the control
channels and voice channels in each cell for each of the three
systems
Channel Assignment
Objectives: increasing capacity and minimizing interference
Fixed channel assignment:
Each cell is assigned a pre-determined set of frequency channels.
If all the channels in the same cell are used, the new calls are
blocked.
Variations: borrowing strategy. MSC, interference consideration
Dynamic channel assignment:
No pre-determined assignment of frequency channels is made.
When a call arrives, the base station ask the mobile switching
centre (MSC) to allocate a free channel.
MSC must take into account the co-channel interference in
channel.
Reduce blocking probability and increase capacity.
Handoff Strategies
Handoff: When a mobile station
travels from one cell to another
cell by crossing the cell
boundary, the base station that
serves this mobile station is
required to change from one to
another. This process is called
handoff.
Identifying a new base station,
allocate voice and control channel
in the new base station.
as infrequently as possible;
without causing unacceptably
perceptible nuisance to users.
The minimum usable signal level , for achieving
acceptable performance
The margin , where is the
power level for initiate handoff,
If is too large, unnecessary handoffs will burden the system.
If is too small, there will be insufficient time to complete a
handoff before a call is lost due to weak signal conditions.
Excessive delay in handoff may occur in crowded
situations because:
the system is too busy to do scheduling for many handoff requests;
no spare channel is available in the cell that the mobile enters into.

P
min
=P
handoff
P
min
P
handoff

False triggering: handoff due to momentary fading

rather than moving away from the base station
Solution: The base station monitors the received signal
strength for a certain period of time before triggering the
handoff procedure.
Dwell Time: The time over which a user remains within
one cell is called the dwell time.
The statistics of the dwell time are important for the practical
design of handover algorithms.
The statistics of the dwell time vary greatly, depending on the
speed of the user and the type of radio coverage.
In 1G, each BS constantly monitors the signal strengths
of all of its reverse voice channels to determine the
relative location of each mobile user with respect to the
BS. This information is forwarded to the MSC who
makes decisions regarding handover. (1G)
In 2G, Mobile assisted handover (MAHO) : The
mobile station measures the received power from
surrounding BSs and continually reports the results of
these measurements to the serving BS.
Intersystem handoff: Required if a mobile station
moves from one cellular system to a different cellular
system. Change from MSC to MSC
Prioritizing Handover: Guard channel concept
Dropped call is considered a more serious event than call
blocking. Channel assignment schemes therefore must give
priority to handover requests.
A fraction of the total available channels in a cell is reserved
only for handover requests. However, this reduces the total
carried traffic. Dynamic allocation can improve this.
Queuing of handover requests is another method to decrease
the probability of forced termination of a call due to a lack of
available channel. The time span over which a handover is
usually required leaves room for queuing handover request.
Practical handoff consideration
High speed users and low speed users have vastly different
dwell times which might cause a high number of handover
requests for high speed users. MSC burned!!!
The Umbrella Cell approach will
help to solve this problems.
Different antenna height and power
High speed users are serviced
by large (macro) cells,
while low speed users are
handled by small
(micro) cells.

`
A hard handover does break before make, i.e., The old
channel connection is broken before the new allocated channel
connection is setup. This obviously can cause call dropping.
In soft handover, we do make before break, i.e., The new
channel connection is established
before the old channel connection
is released.
Co-Channel Interference
Co-channel cell : The set of cells using the same set of
frequencies as the target cell.
Interference tier :
A set of co-channel
cells at the same
distance from the
reference cell.
Interference depends on radius and distance
Co-channel reuse ratio:
Signal to noise ratio (SNR):
improves by increasing the signal power
Signal to interference ratio (SIR):
cannot be improved by changing the signal power
Power law:
and are the power and the distance at the reference
point, is the path loss exponent.

SIR is
P
r
=I = P
r
=
P
i
0
i=1
I
i
P
r
=N
0
P
r
=P
0
(
d
d
0
)
n
P
0
d
0
n
P
r
=I = R
n
=
P
i
0
i=1
(D
i
)
n
Q =
D
R
=
p
3N Q =
D
R
=
p
3N

Example: Consider only the first tier co-channel cells
and approximated . SIR is . Assume D
i
D
(
p
3N)
n
6
n = 4
More accurate modeling
SIR =
R
n
2(DR)
n
+2(D+R)
n
+2D
n
Adjacent Channel Interference
Reasons
For signals adjacent in frequency
Resulting from imperfect receiver filters, leakage.
Near-far effect
Power control such
that powers appear at
BS equally.
Solution
Accurate filtering
channel assignment
adjacent channels
assigned to different
cells
Trunking & Grade of Service
Trunking
The idea is to accommodate a large number of users in a limited radio
spectrum
The concept is to allow the users to share the small amount of available
channels.
Trunking Theory:
exploits the statistical behavior of users so that a fixed number of channel
may accommodate a large number of users, also in the sense of
likelihood.
Grade of Service (GOS)
is a measure of the ability of the user to access a trunked system during
the busiest hour.
GOS is typically given as the likelihood that a call is blocked, or the
likelihood of a call experiencing a delay greater than a certain queuing
time.
Erlang [Danish mathematician]
The measure of traffic intensity nowadays bears his name.
1 Erlang: 1call-hour per hour or 1 call-minute per minute
Example: A radio channel that is occupied for 30 minutes during an hour
carries 0.5 Erlang of traffic.
Set-up Time: The time required to allocate a trunked radio
channel to a requesting user.
Blocked Call: Call which cannot be completed at time of request,
due to congestion.
Holding Time: Average duration of a typical call. Denoted by
(in seconds).
HH
Traffic Intensity: Measure of channel time utilization, which is
the average channel occupancy measured in Erlangs. This is a
dimensionless quantity and may be used to measure the time
utilization of single or multiple channels. Denoted by .
Load: Traffic intensity across the entire trunked radio system,
measured in Erlangs.
GOS: A measure of congestion which is specified as the
probability of a call being blocked (for Erlang B), or the
probability of a call being delayed beyond a certain amount of
time (for Erlang C).
Request Rate: The average number of call requests per unit time
from ONE user. Denoted by seconds
-1
.
AA

The traffic intensity from one user is

For a system containing user, the total traffic
intensity is (regardless of the number of channels):

For a channel trunked system, the traffic intensity
per channel is given by

For given and , GOS can be computed from
trunking theory and queuing theory. (omitted.).
A
u
=H A
u
=H
A =UA
u
A =UA
u
A
c
= A=C = UA
u
=C A
c
= A=C = UA
u
=C
AA CC
UU
CC
Two types of trunked system:
First type offers no queuing for call requests. If no
channels are available, the request is blocked and the
users need to try later.
This type of trunking is called blocked calls cleared.
Called Erlang B system and the probability that a call is
blocked (GOS) is


Example: AMPS system
is designed for GOS of
2% blocking.
Pr[blocking] =
A
C
C!
P
C
k=0
A
k
k!
Pr[blocking] =
A
C
C!
P
C
k=0
A
k
k!
Erlang B

Erlang B
Example 1: 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) 5
(b) 10 (c) 20 (d) 100. Assume each user generates 0.1
Erlangs of traffic.
(a) , from figure we know
. The total number of users is
(b) . The total number of users is
(c) 120
(d) 900
C =5; A
u
=0:1; GOS =0:005 C =5; A
u
=0:1; GOS =0:005
A = 4 A = 4
U = A=A
u
= 11 U = A=A
u
= 11
A = 1:1 A = 1:1
U = A=A
u
= 40 U = A=A
u
= 40
Interestingly, increases when increases, this means.
U=C U=C
CC
`
Example 2: An urban area has a population of 2 million
residents. Two competing trunked mobile networks
(systems A and B) provide cellular service in this area.
System A has 394 cells with 19 channels each, system B
has 98 cells with 57 channels each. Find the number of
users that can be supported at 2% blocking if each user
averages 2 calls per hour at an average call duration of 3
minutes. Assuming that all two trunked systems are
operated at maximum capacity, compute the percentage
market penetration of each cellular provider.
Given
Probability of blocking=0.02
Traffic intensity per user: Erlangs. A
u
= H = 2 3=60 = 0:1 A
u
= H = 2 3=60 = 0:1
System A: C=19
From Erlang B figure, the total traffic is 12 Erlangs.
The number of users per cell is then
The total number of subscribers is then 120394=47280
The market penetration is 47280/2000000=2.36%
System B: C=57
From Erlang B figure, the total traffic is
The number of users per cell is
The total number of subscribers is
The market penetration is
U = A=A
u
= 120 U = A=A
u
= 120
45098=44100.
?
U = A=A
u
= 450 U = A=A
u
= 450 ?
?
45 Erlangs.
? 44100/200000=2.205%.
Example 3: A certain city has an area of 1300 square miles and is
covered by a cellular system using a 7-cell reuse pattern. Each
cell has a radius of 4 miles and the city is allocated 40 MHz of
spectrum with a full duplex channel bandwidth of 60 kHz.
Assume a GOS of 2% for an Erlang B system. If the offered
traffic per user is 0.03 Erlangs, compute (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, (f) the
number of mobiles per unique channel, (g) the theoretical
maximum number of users that could be served at one time by
the system.
The area of the cell is . The number of the cell
is then
The number of channel per cell
From Erlang B figure, at 2%,
Maximum carried traffic=the total intensity of the system=
Erlangs
The total user supported is
The number of unique channel is . So the
number of users per unique channel is
The theoretical maximum number of users at one time is
, which is 3.4% of the supportable users.
2:6R
2
= 41:6 2:6R
2
= 41:6
1300=41:6 31 1300=41:6 31
40000=60=7 95 40000=60=7 95
A = 84 A = 84
2604=0:03 = 86800 2604=0:03 = 86800
40000=60 = 666 40000=60 = 666
86800=666 130 86800=666 130
31 95 =2945 31 95 =2945
31 84 = 2604 31 84 = 2604
The second type provides queue to hold calls when
blocked. If there is no channel available, the call request
may be delayed until a channel becomes free.
This type is called blocked calls delayed.
The GOS of this type is defined as the probability that the
call will not have immediate access to a channel. Erlang C
formula:



Pr[delay > 0] =
A
C
A
C
+C!

1
A
C
P
C1
k=0
A
k
k!

Pr[delay > 0] =
A
C
A
C
+C!

1
A
C
P
C1
k=0
A
k
k!

The probability that a call is delayed more than seconds is

given by (can also be used as GOS):



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

and the average delay for those calls which are already in
queue is
DD
D = Pr[delay > 0]
H
C A
D = Pr[delay > 0]
H
C A
H=(C A) H=(C A)
Pr[delay > t] =Pr[delay > 0]Pr[delay > tjdelay > 0]
=Pr[delay > 0] exp

(C A)t
H

Pr[delay > t] =Pr[delay > 0]Pr[delay > tjdelay > 0]

=Pr[delay > 0] exp

(C A)t
H

tt
Erlang C (delay>0)
Example 4: A hexagonal cell within a 4-cell system has
a radius of 1.387km. A total of 60 channels are used
within the entire system. If the load per user is 0.029
Erlangs, and =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 km will this system support?
(b) What is the probability that a delayed call will have to wait
for 10s?
(c) What is the probability that a call will be delayed for more
than 10s?
The area covered per cell is
The number of channels per cell is 60/4=15.
From Erlang C chart, at GOS=5%, C=15, we got A=9
Erlangs.
The number of users supported per cell is 9/0.029=310.
The number of users per square kilometer is 310/5=62.
The average holding time is
The probability for a delayed call to wait for 10 s is

The probability that a call will be delayed for more than 10 s is

2:6 1:387
2
= 5 km
2
2:6 1:387
2
= 5 km
2
exp((C A)t=H) = exp((15 9)10=104:4) = 56:29% exp((C A)t=H) = exp((15 9)10=104:4) = 56:29%
H = A
u
= = 0:029 hours = 104:4 s H = A
u
= = 0:029 hours = 104:4 s
Pr[delay > t] = 0:05 0:5629 = 2:81% Pr[delay > t] = 0:05 0:5629 = 2:81%
Some points of interest
Increase in N decrease the Frequency Reuse (Not
desired) Decrease the Capacity C=M*K*N

But Small N means Decrease in Q (Not Desired)

Service Provider: Minimum possible value of N

Service Buyer: Max possible value of Q



System Expanding Techniques
Cell Splitting
Cell splitting increases the number of BSs in order to increase
capacity. There will be a corresponding reduction in antenna
height and transmitter power.
Improving capacity by increasing the number of base stations,
or the number of times the channels are reused.
In theory, cell splitting could be repeated indefinitely. In
practice it is limited
by the cost of base stations
handover (fast and low speed traffic)
not all cells are split at the same time : practical problems of BS sites,
such as co-channel interference exist
The number of the cells increases four times. To keep the same received
power at mobile, the BS power should be reduced by a factor
2
n
In practice, not all cells are split at the same time. Special organization is
then needed
Sectoring
Instead of using one omnidirectional
antenna, a base station can use
multiple directional antennas, each
of which caters for a coverage area
at a particular direction.
Reduced co-channel interference.
For 7-cell network, if we partition a
cell into three 120 sectors, then the
number of co-channel cells are
reduced from 6 to 2 in the
first tier. Using six sectors of 60,
we have only one co-channel cell in
the first tier.

So decrease the cluster size N in
order to improve the frequency
reuse.
Handoff becomes often and smart
modern system will ask BS to do
the job.
If the number of channels in one
sector becomes smaller, then
trunking efficiency decreases.
Micro Zone Concept
Sectoring increases the
number of handoffs and
hence increases the load
of MSC.
Microcell zone concept
decreases the burden of MSC
by reducing the handoff
request.
Same channel is used when the
mobile moves from one zone
to another.


Interference Reduction Techniques
Sectorization: Directional antennas
Require more handoff
Smart Antennas/Beamforming
Require multiple Antennas
Multiuser Detection
Complexity issue
Interference Pre-subtraction
Require Channel knowledge
Frequency Hopping:
change the carrier frequency according to a unique pattern
Summary
Frequency Reuse
Channel Assignment Strategies
Handoff Strategy
Interference
Co-channel Interference
Adjacent Channel Interference
Power Control
Trunking Theory
Erlang B
Erlang C
System Expanding Techniques
Cell Splitting
Sectoring
Microcell Zone Concept

Example
Consider a cellular system in which there are a total of
301 radio channels available for handling traffic.
Suppose the area of a cell is 6 km
2
and the area of the
entire system is 2100 km
2
. The cluster size is 7.
(a) Calculate the system capacity
(b) How many times would the cluster of size 4 have
to be replicated in order to approximately cover the
entire cellular area?
Given that in a 7 cell reuse system the traffic load per
user is 0.03 Erlangs and the average number of calls per
hour per user is 1.5, for 5% probability of call blocking,
(c)Find the traffic load per cell?
(d) The mean duration of a call