MOBILE

COMMUNICATION

M R Abidi
Lecture 3
Current Wireless Systems

„ Cellular Systems
„ Wireless LANs
„ WIMAX
„ Satellite Systems
„ Paging Systems
„ Bluetooth
„ Ultrawideband radios
„ Zigbee radios
Fundamentals of Cellular Systems
Cellular System Infrastructure
Ad Hoc and Sensor Networks
“An ad hoc network is a collection of wireless mobile
hosts forming a temporary network without the aid of any
centralized administration or standard support services
regularly available on the wide area network to which the
hosts may normally be connected.”
D B Johnson and D A Maltz
Lecture 4 & 5
Cellular Concept

„ Given a propagation environment, increasing transmitted

power will increase the service coverage area.
⇒The coverage area can be controlled by using a proper transmitted power
level.

„ In cellular systems, the total service area is divided into a

number of smaller areas, each of which is a radio cell.
⇒Advantages
„ Low transmitted power
„ Frequency reuse possible
Shape of the cell coverage area
Impact of cell shape and radius on service
characteristics
„ Hexagonal cells are popular because
‰ closest to a circle
‰ tight cellular packing
‰ perfect partitioning of the service area
Signal strength and cell parameters
Handoff
Cell cluster concept
Frequency reuse

„ Frequency reuse is limited by co-channel

interference. Cells which use the same frequency
channels are called co-channel cells.

„ Frequency is reused from cell cluster to cell

cluster. No frequency channel is reused among
cells in the same cell cluster.
⇒ Cells in each cell cluster use unique frequency
channels.
Frequency reuse
Capacity expansion by frequency reuse
Frequency reuse factor

„ A hexagonal cell has 6 equidistant neighbours

„ Lines joining the centers of any cell and each of

its neighbours are separated by multiples of 60
degress
⇒ Only certain cluster sizes and cell layouts are possible
Nearest co-channel neighbours

„ To find the nearest co-channel neighbour of a

particular cell, execute the following two steps:
- move i cells along any chain of hexagons
- turn 60 degrees counterclockwise and move j cells
where the integers i and j are parameters for
determining co-channel cells and for determining the
size of the cell cluster (N)
19-cell reuse example (N=19)
(i=3, j=2)
Geometry of hexagonal cells
Geometry of hexagonal cells

Let R: radius of the cell (from center to vertex)

D: distance from the center of the candidate cell
to the cell of the nearest cochannel cell

It can be shown that

D = 3N R
 ΔD
Frequency reuse ratio q = = 3N
R
Frequency reuse ratio and frequency reuse factor
carry the same information:
q (or N) ↑ ⇒ cochannel interference ↓
⇒ frequency reuse less often and system
capacity ↓
We should choose the minimum q (or N) subject to the
constraint on the signal to cochannel interference ratio
requirement.
Cochannel interference

Let
NI- the number of co-channel interfering cells
Ii- cochannel interference from the ith co-channel cell
S - the received power of the desired signal
The signal-to-cochannel interference ratio (S/I), also
referred to as carrier-to-co-channel interference ratio
(CIR), is
S S
=
∑
I NI
Ii
i =1
Consider only distance-dependent path loss
Pr(d) = P0(d/d0)−κ
where
Pr(d) – the received power at distance d (≥ d0)
P0 – the received power at distance d0
κ - the path loss exponent
d - the distance between the transmitter and receiver.
Consider the forward link and assume that the
transmitted power levels from all the BSs are the same,
then
Ii∝ Di -k ,
where Di is the distance from the ith cochannel cell BS
to the mobile.
When the mobile is at the cell boundary (the worst
case),
S∝ R -k
⇒

S R −k
=
∑
NI
I D i− k
i =1
If we neglect cochannel interference from the second
and other higher tiers ⇒ NI = 6.
If r = R and using Di≈ D for i=1, 2 ,….., NI

S (D / R)
=
κ
=
qκ
=
( 3N )
κ

I NI NI NI
1/ κ 1/ κ
⎛ S⎞ ⎛ S⎞
⇒ q = ⎜ NI × ⎟ = ⎜6× ⎟
⎝ I ⎠ ⎝ I ⎠
Lecture 6
CCI
CCI
CCI

S R −κ
≈
I 2( D − R ) −κ + 2 D −κ + 2( D + R ) −κ
S 1
⇒ =
I 2(q − 1) −κ + 2q −κ + 2(q + 1) −κ
„ Consider a cellular system that requires an S/I ratio of
18 dB
‰ For a frequency reuse factor of 7, calculate the worst-case
S/I.
‰ Is a frequency reuse factor of 7 acceptable in terms of CCI?
If not, what would be a better choice of N?
„ Solution
For κ = 4, N = 7, q = (3N)1/2 = 4.6 ⇒ S/I = 17.3 dB
For κ = 4, N = 9, q = (3N)1/2 = 5.2 ⇒ S/I = 19.8 dB
Adjacent channel interference (ACI)
 Near-far effect:
d1 « d2 ⇒ Pr1 » Pr2 at the BS

For the signal from MS2, the adjacent channel interference ↑↑ due to the near-far effect
ACI

„ To reduce ACI
‰ use modulation schemes which have small out-of-band
radiation (e.g., MSK is better than QPSK)
‰ carefully design the receiver BPF
‰ use proper channel interleaving by assigning adjacent
channels to different cells
‰ furthermore, do not use adjacent channels in adjacent cells,
which is possible only when N is very large. For example, if
N =7, adjacent channels must be used in adjacent cells
‰ use FDD or TDD to separate the forward link and reverse
link
Trunking
Trunking allows a large number of users to share the relatively small
number of channels in a cell with full availability

Definitions
„ set-up time: the time required to allocate a trunked radio channel to a
requesting user
„ blocked call (lost call): call which cannot be completed at the time of
request, due to congestion
„ holding time (H): average duration of a typical call
„ traffic intensity (ρ): measure of channel time utilization, which is the
average channel occupancy measured in Erlangs
„ load: traffic intensity across the entire trunked radio system, measured
in Erlangs
„ request rate (λ): the average number of call requests per unit time per
user
„ The traffic intensity offered by each user is (in Erlangs)
ρu = λH
„ For a system with u users and an unspecified number of channels, the total
offered traffic intensity is (in Erlangs)
ρ = uρ u = uλH
„ In a J channel trunked system, if the traffic is equally distributed among the
channels, then the traffic intensity per channel is

ρ c = uρ u / J = uλ H / J
„ Difference between offered traffic and carried traffic
offered traffic intensity = ρ
blocking probability = PB
carried traffic intensity = ρ(1 - PB) ≤ ρ
Types of trunked systems

Blocked call cleared system

System can be modeled as M/M/J
Erlang – B formula

ρ J / J!
PB =
∑
J
ρk / k!
k =0
„ Blocked calls delayed
‰ If a channel is not available immediately, the call
request may be delayed until a channel becomes
available
‰ Other conditions (assumptions) are the same as those in
the case of blocked calls cleared
‰ Erlang – C formula: The probability of a call not
having immediate access to a channel
„ Probability of non-zero delay

„ If no channels are immediately available the call is delayed,

and the probability that the delayed call is forced to wait more
than t seconds (µ=1/H):
Average delay for all calls in a queuing system
Lecture 7 & 8
Capacity enhancement of cellular systems

„ The capacity can be improved by

‰ cell splitting
‰ antenna sectoring
‰ dynamic channel assignment
Cell splitting

„ Subdivide a congested cell into smaller cells, each with its

own base station and a corresponding reduction in
antenna height and transmitted power
„ Cell splitting increases the capacity of a cellular system
(M ↑)
„ Reducing cell size increases handoffs, the number of base
stations needed, and may result in a difficulty in finding a
proper site for the base station.
„ Old base station should be kept in some splitting cells.
Illustrate of cell splitting from radius R to R/2 and to R/4
Let
Pt1: the transmitted power of large cell BS
Pt2: the transmitted power of small cell BS
Pr: the received power at cell boundary
Then
Pr (large cell) ∝ Pt1 · R−κ
Pr (small cell) ∝ Pt2 · (R/2)−κ
On the basis of equal received power
Pt1 · R−κ = Pt2 · (R/2)−κ
Pt1 / Pt 2 = 2 κ
10 log 10 ( Pt1 / Pt 2 ) = 10 κ log 10 2 ≅ 3κ dB
For κ = 4, Pt1 / Pt 2 = 12 dB
In general,
R → R / 2 in cell splitting
⇒ cell area → 14 cell area
⇒ capacity is increased by 3 times
Cell sectoring
Sectoring improves S/I
„ Assuming 7-cell reuse pattern, for the 3-sector case, the
number of interferers in the first tier is reduced from 6 to
2
S R −κ R −κ
= =
∑ ∑
N I −κ
I Di−κ
2
D
i =1 i i =1

Larger than the omnidirectional case where NI=6

with D i ≈ D
⎛S ⎞
× qκ
1
⎜ ⎟ =
⎝ I ⎠ omni 6
⎛S ⎞
= × qκ
1
⎜ ⎟
⎝ I ⎠ 120 degrees 2
„ Worst case scenario in 120 degree sectoring
⎛S⎞ R −κ
⎜ ⎟ =
⎝ I ⎠1200 D −κ + ( D + 0.7 R) −κ
1
=
q −κ + (q + 0.7) −κ
For κ = 4, N = 7 (S / I ) 0 = 24.5 dB > 18 dB
120
Handoff

„ Process consists of
‰ Link quality evaluation and handoff initiation
‰ Allocation of resources
„ No. of cell boundary crossings ∝1/cell size
„ Handoff rate ∝√cell density in macrocell
∝ cell density in microcell
„ More robust and reliable algorithms for smaller
cell sizes
„ Monitoring of link quality to determine when the
handoff is needed
‰ Improper strategy results in excessive switching load or
large amount of CCI
„ Link quality measurements depends on
„ BER
„ C/I
„ Distance
„ Traffic load
„ Signal strength etc.
„ HO operation involves
‰ Identifying a new BS
‰ Allocation of voice and control signals to the channels
associated with the new BS
„ HO operation
‰ Successfully
‰ As infrequently as possible
‰ Imperceptible to the users
Define:
Pr,minimum usable : min usable signal power for acceptable
voice quality at BS
Pr,handoff : HO threshold
Δ = Pr,handoff - Pr,minimum usable

Very small Δ : insufficient time to complete a HO before a

call is lost due to weak signal condition
Large Δ : unnecessary HO may burden MSC
Handoff
„ Necessary to differentiate between momentary
fading and that the MS is actually moving away
from BS
‰ BS monitor the signal level for a certain period of time
before HO
‰ Length of time used for signal strength measurement
depends on the speed of vehicle
‰ If slope of short term average received signal level in a
given time interval is steep → quick HO
Dwell time: time over which a call may be
maintained within a cell without HO
‰ governed by number of factors
„ Propagation
„ Interference
„ Distance etc.
‰ Stationary subscriber may have a finite dwell time,
why?
‰ Statistics of dwell time vary greatly, depending on
speed of MS and type of radio coverage
HO algorithms

Classification based on roles of BS and MS perform

1. Network Controlled HO (NCHO)
„ Centralized control
„ Only intercell HO
„ Several sec delay
2. Mobile Assisted HO (MAHO)
„ Both serving BS and MS measure link quality
„ Link quality meas. of alter BS are obtained by MS
„ Decision is made by serving BS along with MS
„ Both intracell and intercell HO
„ Delay 1-2 sec
3. Mobile Controlled (MCHO)
„ Decentralized strategy
„ Link quality is measured by BS and MS (with alter BS it is
done by MS)
„ Meas. done by serving BS are relayed to MS
„ HO is made by MS
„ Delay approx. 100 ms
Practical HO considerations

„ Problem due to varying mobile speeds

‰ Microcells → increases capacity but MSCs are
burdened with high speed vehicles
‰ Difficult to find new cell sites
„ Use different antenna heights (on the same
building or tower) and different power levels
‰ Possible to provide coexisted ‘large’ and ‘small’ cells
at a single location → umbrella cell approach
‰ Large area coverage to high speed users and small area
coverage to low speed users
Umbrella cell
Cell dragging
‰ Results from pedestrian users that provide very strong
signal to BS
„ If avg. signal strength does not decay rapidly even user travels
beyond the cell coverage area → HO will not be made
⇒ Interference and traffic management problem
⇒ HO threshold and radio coverage parameters must be
adjusted carefully
Lecture 9 & 10
„ Hard HO
‰ MS can connect to only one BS at a time
‰ Absolute decision is taken about HO without making a
number of simultaneous connections among candidate
BSs
‰ Target BS selected prior to executing HO
‰ A brief interruption in the connection
„ Soft HO
‰ MS can connect to a number of BSs
‰ MS selects the best candidate BS as the target
‰ Link quality measurements are done by MS (a type of
MAHO)
‰ Complex and expensive
‰ Increased forward interference
SS based hard HO algorithm

„ A MS is reconnected to an alternate BS whenever the signal

strength estimate of the target BS exceeds that of the serving
BS by at least H dB
„ Handoff is performed between two BSs, when

where
H : hysteresis (in dB)
Y0(n) : estimated mean signal strength (in dBm) of BS0
Y1(n) : estimated mean signal strength (in dBm) of BS1
„ SS of BSi, i=0, 1,…., N – 1 can be obtained by calculating
time averages of the received squared envelope,
„ Many Variations of the algorithms exist
„ HOs are also trigerred when the measured signal strength of the
serving BS drops below a threshold
„ HO could be performed, when
„ Another variation discourages handoffs when the received
signal strength from the serving BS exceeds another threshold
ΩU
„ HO is performed if
„ Direction biased HO algorithm
„ Incorporate moving direction information into the handoff
algorithm to encourage handoffs to BSs that the MS is
approaching, and to discourage handoffs to BSs that the MS is
moving away from
„ Algorithm can be defined by grouping all the BSs being
considered as handoff candidates, including serving BS, BSs
into two sets based on their direction information
He: encouraging hypothesis
Hd : discouraging hypothesis
HO to BSj is requested if BSj ∈ R and
Zone cell concept
Channel assignment techniques

A channel allocation algorithm should yield

‰ High spectral efficiency for a specified GOS
‰ Low computational complexity
‰ Quick assignment of a channel to MS
‰ Maintain the best speech quality
Classification

Channel assignment schemes

Fixed Dynamic

Basic
Fixed Flexible
Borrowing

Hybrid Predictive

Scheduled
FCA

„ Used by 1G macrocellular systems

„ Permanent allocation of channels to cells in
advance according to traffic estimates
‰ A new call / HO arrival is blocked in a cell if all ch are
busy in the cell
„ Gives adequate performance with homogeneous
and stationary traffic
„ Becomes inefficient under nonhomogeneous,
nonstationary traffic
FCA

„ Blocking probabilities ↓ by using schemes that

borrow channels from neighbouring cells
„ Borrowed channel is prohibited to be used by all cells within
CC reuse distance
„ Under heavy traffic, worse than simple FCA
‰ Problem is solved partially by hybrid schemes (two group of
channels: one owned by the cell and the other that of borrowable
channels)
Channel assignment problem is fundamentally
different in macrocellular and microcellular
networks
DCA

„ Suitable for microcellular networks

„ Dynamic nature permits adaptation to spatial and
temporal traffic variations
„ No exclusive relationship between cells and
channels
„ Any cell can use any channel (without violation of cc reuse
constraint)
„ Outperform FCA with light nonstationary traffic
„ With heavy traffic FCA is better
DCA

„ Centralized
„ Decentralized
„ Fully decentralized
Flexible channel assignment

„ Combine the aspects of FCA and DCA

„ Each cell has a fixed set of channels
„ A pool of channels is reserved for flexible
assignment
‰ Channels assignment from this pool can be scheduled
or predictive
Borrowing schemes

„ Borrowing with channel ordering (BCO)

„ Borrowing with directional locking (BDCL)
„ Channel borrowing without locking (CBWL)
Call arrival policy for the BCO
Call departure policy for BCO
Borrowing with directional locking
Borrowing without locking