PUE 3123: Telecommunication Network Design

and Planning
Course Outline
Overview of telecommunication networks, Call routing,
Teletraffic for circuit-switched and packet networks;

Network Strategy and Planning:
architecture, technology, planning process,
multi-layered network model,
dimensioning;

Access Network Planning:
requirements - service and physical,
Copper access network structure and planning,
Optical fibre and Radio access systems;

Core Transport Planning;

Transport network planning:
Process,
Node locations,
Fibre connectivity,
Design trade-offs,
planning tools;

Mobile Network Planning:
Digital Radio concepts and Standards:
2G, 2.5G (GPRS, EDGE), 3G (WCDMA) and
4G (OFDM, ALL-IP,WLAN);

Architectures, Interfaces, and Protocol Structures of mobile networks,
Radio and core network planning and optimisation,
Signalling, Coverage and propagation modelling,
Traffic dimensioning, Capacity planning, Microwave Link Planning;

Cell Sites Planning,
Macrocell, Microcell, picocell and
Indoor system,

Link-budget calculation,

Forecasting Methods,
Forecasting Traffic Volume and Dispersion.
Traffic Map and Distribution Blocking Probability,
Forecasting Subscriber Distributions,
Size and Location of Switching centres,
Cell Reuses Pattern,
Individual Channel Assignment,
Cell Breathing, Sectorization and Densification,

Power Control,
Frequency hopping,
Discontinuous transmission,

Multi-layer structure (Dual Mode/Band, Overlay and Underlay) algorithms,

Field Measurement and Network Optimisation,

Network Planning and Design Exercises;

Data Network Planning:
data platforms,
Planning of data platforms:
Internet traffic,
Access to ISPs,
dimensioning,
QoS management.

Course Assessment
Continuous
Assessment Tests
Practicals Assignments End of semester
examination
Total
10%, 15% 5% 70%, 100%


Overview of Telecommunication Networks

A telecommunication network is a collection of equipment and
facilities with connectivity among users so as to enable transfer of
information among users across the network, while meeting an
acceptable quality of service.

Telecommunication Network Components
Subscriber
- Devices attached to
network
Local Loop
- Subscriber loop
- Connection to network
Exchange
- Switching centers
- End of f i ce suppor t s
subscribers
Trunks
- Branches between exchanges
- Multiplexed
The equipment
- Telephone Sets
- Computers
- Modems
- Switches
- Routers
- Bridges
- etc
The Facilities
- Wire Transmission lines
- Optical Fibre Transmission links
- Satellite links (and antennas)
- Microwave links (and antennas)
- Wireless Transmission links
- etc

Telephone Networks

Signaling Network
The Signaling network is the brain
Nodes in the signaling network are called signaling points
SSP -> Service Switching Points
STP -> Signaling Transfer Point
SCP -> Service Control Point

Telecommunication Network Architecture
Generic Architecture For PSTN, GSM, 3G and WiMAX









GSM System Architecture


GSM Network Elements
Handset
BTS: Base Transceiver Station
MSC: Mobile Switching Center
BSC: Base Station Controller
HLR/VLR: Home Location Register/Visiting Location
Register
SIM Card: Subscriber Identity Module Card
Service Centers: USSD, SMS, ...
Mobile Handset (MH)
Used by the subscriber to access the GSM network via
the air interface
Contains the hardware and software specific to the
radio interface
Need a Subscriber Identity Module (SIM) which
contains the subscriber-specific data to access GSM
network, except emergency calls
PSTN
Data Terminal
HLR/
VLR
MSC BSC
OMC
(Operation & Maintenance
Center)
Operation
Terminal
BTS
Handset
A
X.25
A-bis SS7
Network sub-system PSTN
Radio
sub-system
Mobile
station
Um
SIM
card
Base Transceiver Station (BTS)/Base Station
Responsible for communication to and from MHs via air
interface
MH communicates with the mobile system using a
radio channel to a BTS
BTS comprises radio transmission and reception
devices
Including the antennae and signaling processing
specific to the radio interface
BTS separates the speech and control signaling
associated with a MH and sends them to the BSC on
separate channels
Mobile Services Switching Center (MSC)
Responsible for all switching functions related to call
processing
Coordinate the setting up of calls to and from GSM
users
MSC is the interface between GSM network & PSTN
(Public Switched Telephone Network)

Base Station Controller (BSC)
Monitor and Control several base stations
Frequency administration, control of BTSs,
exchange functions
Responsible for all the radio interface management
Channel allocation and release, handover
management
BSC is the interface between MSC and BTS
BSC is connected on one side to several BTSs and
on the other side to the MSC


Home Location Register (HLR)
A database in charge of the management of mobile
subscribers
Contains information of network subscribers
Information on teleservices and bearer services
subscription, service restrictions, and
supplementary services
The data stored in HLR is of a semipermanent
nature and does not usually change from call to
call
Visitor Location Register (VLR)
Contains the relevant data of all MHs currently located
in a serving MSC
The permanent data is the same as data in the HLR
The temporary data includes
Temporary Subscriber Identity (TMSI)
Location Area Identity (LAI) of an MH
VLR allocates mobile subscriber roaming numbers
(MSRNs) for the incoming call setup
Subscriber Identity Module (SIM)
Subscriber Identity Module
Contains
phone number (MSISDN)
international mobile subscriber identity (IMSI)
status of SIM
service code
authentication key
PIN (personal identification code)
PUK (personal unlock code)

SIM Cards
SIM cards are embedded with a microprocessor
Stores customer identity information and is made
to fit inside the cellular phone
Memory capacity
1K, 3K, 8K, 16K
With/Without OTAC (Over The Air Customization)
Two types of SIM card form factors
ID-1 SIM
Plug-In SIM

Application Service Centers
Responsible for GSM network add-on services
Operation and Maintenance Center (OMC)
Monitoring and control the network
Usually connect with MSC, BSC, HLR, and
other service centers
Short Message Service Center (SMSC)
provide short message services
usually connect to MSC

Geographical Network Structure
Location Units
Cell
Location Area (LA)
MSC/VLR Service Area
PLMN (Public Land Mobile Network) Service Area
GSM Service Area
Cell Planning

Location Information
GSM Service Area Hierarchy

The area in which a subscriber can access the network.

Cell Characteristics
The Basic Union In The System
defined as the area where radio coverage is given
by one base station.
Addressed by Cell Global Identity (CGI)
A cell has one or several frequencies, depending on
traffic load.
Frequencies are reused, but not used in
neighboring cells due to interference.
Cell Planning
cell
Location Area
MSC/VLR
PLMN Service Area
(one per operator)
GSM Service Area

Reusing frequencies in different geographic areas:
7 cell repeat pattern is typical, but other patterns
are also common
Cell Patterns

Frequency Utilization v.s. K
Frequency cant be reused in the same pattern
Larger K
A3
A1
A2
G3
G1
G2
C3
C1
C2
B3
B1
B2
F3
F1
F2
D3
D1
D2
E3
E1
E2
G3
G1
G2
F3
F1
F2
C3
C1
C2
A3
A1
A2
B3
B1
B2
E3
E1
E2
D3
D1
D2
7-cell reuse
pattern
Frequency
reuse
1
3
2
3
2
7
1
6
4
5
K=3
K=7
More channels in a pattern
Higher subscriber number
Worse frequency reusability
Smaller K
Less channels in a pattern, high channel distortion
Smaller subscriber number
Better frequency reusability
Numbering
Mobile Subscriber ISDN Number (MSISDN)
Mapping to Mobile Station Roaming Number
(MSRN) by HLR
International Mobile Subscriber Identify (IMSI)
Stored in SIM/HLR
Temporary Mobile Subscriber Identity (TMSI)
Assigned by VLR, Stored in VLR
International Mobile Equipment Identity (IMEI)
Unique ID to handset, used by air interface
ID-Numbers
MSISDN = CC + NDC + SN
MSISDN: Mobile Station ISDN Number
CC Country Code
NDC National Destination Code
SN Subscriber Number
IMSI = MCC + MNC + MSIN
IMSI International Mobile Subscriber Identity
MCC Mobile Country Code
MNC Mobile Network Code
MSIN Mobile Station Identification Number
IMEI = TAC + FAC + SNR + spare
IMEI Internal Mobile Equipment Identity
TAC Type Approval Code, determined by a central
GSM/PCS body
FAC Final Assembly Code, identifies the
manufacturer
SNR Serial Number, uniquely identifies all
equipment within each TAC and FAC
IMEISV = TAC + FAC + SNR + SVN
IMEISV International Mobile Equipment Identity
and Software Version Number
SVN Software Version Number
MSRN = CC + NDC + SN
MSRN Mobile Station Roaming Number
CC Country Code
NDC National Destination Code
SN Subscriber Number. In this case, the address
to the serving MSC
LAI = MCC + MNC + LAC
MCC Mobile Country Code
MNC Mobile Network Code
LAC Location Area Code
CGI = MCC + MNC + LAC + CI
CGI Cell Global Identity
CI Cell Identity
BSIC = NCC + BCC
BSIC Base Station Identity Code
NCC Network Color Code (3bits)
BCC Base Station Color Code (3bits)
LN = CC + NCD + LSP
LN Location Number
CC Country Code
NCD National Destination Code
LSP Locally Significant Part
LN = CC + NCD + LSP
LN Location Number
CC Country Code
NCD National Destination Code
LSP Locally Significant Part
RSZI = CC + NDC + ZC
RSZI Regional Subscription Zone Identity
CC Country Code
NDC National Destination Code
ZC length of the Zone code (2 octets)

Call Routing
Assisted by the Numbering plan
Numbering plan is all the telecom networks, whether
fixed, or wireless,

Information resident in MS & SIM
IMSI, TMSI, IMEI, and miscellaneous information
Routing Information used by Network
MSISDN, MSRN






Teletraffic


Telecommunication sys as a Man/Machine System
Telecommunication systems are complex man/machine systems.
The task of teletraffic theory and teletraffic engineering is to
configure optimal systems from knowledge of user requirements,
habits and behaviour
Teletraffic theory is an inductive
discipline that proceeds from
observations of real systems to
establish theoretical models.

From these models, parameters are
derived, which are then compared
with corresponding observations
from the real system.

If there is agreement, the model is
validated. If not, the model is
elaborated further, , and so on, in a
Scientific Spiral



Scientific Spiral
Traffic Intensity The Erlang
The Erlang (symbol E) is a dimensionless unit that is used in
teletraffic as a measure of offered load or carried load on service-
providing elements such as telephone circuits or telephone
switching equipment.
In 1946 the CCITT (now ITU-T) named the international unit of
telephone traffic Erlangs in honor of the Danish mathematician
Agner Krarup Erlang.
Traffic Measurements
Carried Traffic: For carried traffic, the number of erlangs
represents the average number of concurrent calls carried by the
circuits (or other service-providing elements), where that average
is calculated over some reasonable period of time.
Offered Traffic: For offered traffic, the number of erlangs
represents the average number of concurrent calls that would
have been carried if there were an unlimited number of circuits
(that is, if the call-attempts that were made when all circuits were
in use had not been rejected).
The relationship between offered traffic and carried traffic
depends on the design of the system and user behavior. Three
common models are
(a) Blocked Calls Rejected blocked calls are rejected, and never come
back (also referred to as blocked calls cleared),
(b) Blocked Calls Delayed blocked calls try again after some time
(c) Blocked Calls Held blocked calls are held in queue(s) in the system
until a circuit becomes available.
Instantaneous Traffic
The instantaneous traffic, expressed erlangs, represents the
exact number of calls taking place at a point in time.
In this case the number is an integer.
Traffic-level-recording devices, such as moving-pen recorders,
plot instantaneous traffic.
Call Arrival Rate and Call Holding Time
Call arrival Rate: calls per unit time (e.g. calls per second)
Call Holding Time: h units of time (usually seconds)
This is the Average holding time; the average
number of calls per unit time is then the reciprocal of
the holding time

1 1
h
h

= =

Offered traffic: E erlangs
E h E

= =

This relation holds provided that h and are expressed using the
same units of time (seconds and calls per second, or minutes and
calls per minute, etc).



Busy Hour Traffic (in Erlangs): This is the average number of
concurrent calls during a given one-hour period of the day, where
that period is selected to give the highest result. (This result is
called the time-consistent busy hour traffic).
The goal of teletraffic theory is to determine exactly how many
service-providing elements should be provided in order to satisfy
users, without wasteful over-provisioning.


0 1
P P =

( )
1 2 0
2 P P P + = +

( )
2 3 1
2 3 P P P + = +

( )
3 4 2
3 4 P P P + = +

Outgoing links
Incoming
calls
Server places
the calls in
outgoing links
1
2
3
N
0 1 N N-1 2

2 3
( ) 1 N
N

( )
1 2
1
N N N
N P N P P

| |
+ = +
|
\ .

1 N N
N P P

=

Defining the system parameter
/ =
, these equations, after being
written as in the left column below, can be evaluated to give the results in
the right column

0 1
P P =

1 2
2 P P =

2 3
3 P P =

3 4
4 P P =

1 N N
P N P

=

1 0
P P =

2
2 1 0
2 2
P P P

= =

3
3 2 0
3 3 2
P P P

= =

4
4 3 0
4 4 3 2
P P P

= =



That is,
0
!
k
k
P P
k

=
for k=0,1,2,3, , N
By the Law of conservation of probability,
0
1
N
k
k
P
=
=

, it follows that
0
0
1
!
k N
k
P
k

=
=

Therefore
0
!
!
k
k
m N
m
k
P
m

=
=

Note here that if there is infinite waiting room, this becomes the well-
known Poisson distribution
!
k
k
P e
k

=


Erlang B Formula -- Blocking Probability
In teletraffic calculations, the letter A is used in place

, which is used
often in the queuing theory literature. The context identifies the
parameters.

Blocked Calls Cleared
The Erlang B formula, assumes that blocked calls are lost (lost calls cleared)

If there are N links, and a call comes when all the N links are busy, then the
call is blocked, and under Erlang-B assumptions, the call is cleared
The probability
b
P
, that this occurs,
( )
0
!
,
!
N
b m N
m
E
N
P B E N
E
m
=
= =

where:
- is the probability of blocking
- m is the number of resources such as servers or circuits in a
group
- E = h is the total amount of traffic offered in erlangs.

Blocking occurs when a new request from a source finds all the
servers (or outgoing links) already busy.
The formula assumes that blocked traffic is immediately cleared.
The Blocking Probability P
b
that a new call arriving at the circuit
group is rejected because all servers (circuits) are busy, may also
be represented by B(E, N) when E Erlang of traffic are offered to
N trunks (communication channels).

This may be expressed recursively
[4]
as follows, in a form that is
used to simplify the calculation of tables of the Erlang B formula:


Typically, instead of B(E, N) the inverse 1/B(E, N) is calculated in
numerical computation in order to ensure numerical stability:

The following is an implementation via a Visual Basic (VB)
Function

Funct i on Er l angB ( E As Doubl e, N As I nt eger ) As Doubl e
Di mI nvB As Doubl e
Di mj As I nt eger
I nvB = 1. 0
For j = 1 To N
I nvB = 1. 0 + j / E * I nvB
Next j
Er l angB = 1. 0 / I nvB
End Funct i on

The Erlang B formula applies to loss systems, such as telephone
systems on both fixed and mobile networks, which do not
provide traffic buffering, and are not intended to do so. These are
based on the blocked calls cleared model.
Although the formula is derived under the particular assumption
that the call arrivals are modeled by a Poisson process, it is
indeed valid for any statistical distribution of call holding times
with finite mean. Erlang B is a trunk sizing tool for switch to
switch traffic in a circuit-switched network.
Extended Erlang B
The Extended Erlang B is an iterative calculation, rather than a
formula, that adds an extra parameter, the Recall Factor, which
defines the recall attempts.
[6]

The steps in the process are as follows:
1. Calculate
( )
,
b
P B E N =

as above for Erlang B.
2. Calculate the probable number of blocked calls
e b
B EP =

3. Calculate the number of recalls, assuming a Recall Factor,
f
R
:
e f
R B R =

4. Calculate the new offered traffic
1 k k
E E R
+
= +

where is the initial (baseline) level of traffic.
5. Return to step 1 and iterate until a stable value of is obtained.
Erlang C Formula
The Erlang C formula expresses the probability that an arriving
customer will need to queue (as opposed to immediately being
served).
Just as the Erlang B formula, Erlang C assumes an infinite
population of sources, which jointly offer traffic of A erlangs to N
servers.
If all the servers are busy when a request arrives, the request is
queued. An unlimited number of requests may be held in the
queue in this way simultaneously. This formula calculates the
probability of queuing offered traffic, assuming that blocked calls
stay in the system until they can be handled.
( )
( )
1
0
!
! !
N
W
k N N
k
A N
N N A
P
A A N
k N N A

=
+

where:
- A is the total traffic offered in units of erlangs
- N is the number of servers
- P
W
is the probability that a customer has to wait for service.
Engset formula
The Engset calculation is a related formula, named after its
developer, the Norwegian mathematician, Tores Olaus Engset,
used to determine the probability of congestion occurring in the
system.
The formula requires that the user knows the expected peak
traffic, the number of sources (callers) and the number of circuits
in the network.
Example
A business needs to know the minimum number of voice circuits
to and from the network. An approximate approach is to use the
Erlang-B formula.
However, if the business has a small number of extensions, then it
should instead use the more exact Engset calculation, which
reflects the fact that extensions already in use will not make
additional simultaneous calls. (For a large user population, the
Engset and the Erlang-B calculations give the same result.)
The Engset equation is similar to the Erlang-B formula, but has
one major difference:

Erlang's equation assumes an infinite source of calls, yielding a
Poisson arrival process, while Engset specifies a finite number of
callers.

Thus Engset's equation should be used when the source
population is small (say less than 200 users, extensions or
customers).
( )
( )
( )
0
, ,
N
b
k
N
k
S
A
N
P S A N
S
A
k
=
=

where
A = offered traffic intensity in erlangs, from all sources
S = number of sources of traffic
N = number of circuits in group
P
b
= probability of blocking or congestion.
In practice, like Erlang's equations, Engset's formula requires
recursion to solve for the blocking or congestion probability.
There are several recursions that could be used.
[10]
One way to
determine this probability, is to first determine an initial estimate.
This initial estimate is substituted into the equation and the
equation then is solved.
The answer to this initial calculation is then substituted back into
the equation, resulting in a new answer which is again
substituted. This iterative process continues until the equation
converges to a stable result.
Engset's equation:





Example 0. Calls arrive at an exchange and are served as soon as a
server is available. When all servers are busy, the calls are queued.
Let: n(t) = the number of calls active at time t

n
= the arrival rate (calls per sec.) when there are n calls

n
= the service rate (calls per sec.) when there are n calls
(call duration = h
n
= 1/
n
, sec. per call)
P
n
(t) = Probability that there are n calls in the system at time t



Consider an interval
[ , ] t t + A
in which A is so small that terms in
k
A
can be ignored for 2 k > .
Assuming that at the end of that interval, there are n calls active in
the system. This can happen in one of three ways:
1. There are n+1 calls at time t, with one departure in
[ , ] t t + A

2. There are n calls at time t, without an arrival and without a
departure in
[ , ] t t + A

3. There are n-1 calls at time t, with one arrival in
[ , ] t t + A


Further, the above events occur, respectively with the following
probabilities:


Event Probability
a

d
e
p
a
r
t
u
r
e
i
n


[
,
]
t
t
+
A
n
+
1

c
a
l
l
s

a
t

t
i
m
e

t

n
o
d
e
p
artu
re
in
[
,
]
t
t
+
A
n calls
at tim
e t

n
o
a
rriv
a
l
in
[
,
]
t
t
+
A

a
n
a
rriv
a
l
in
[
,
]
t
t
+
A
n
1
calls
at tim
e t

disjoint events leading to
having n calls at t + A t t + A
A
( )
n
P t + A
( )
n
P t
( )
1 n
P t

d
e
p
a
r
t
u
r
e
i
n


[
,
]
t
t
+
A
n
o

d
e
p
a
r
t
u
r
e
,

n
o

a
r
r
i
v
a
l
i
n


[
,
]
t
t
+
A
time
( )
1 n
P t
+
arrival
in [ , ] t t + A
1. There are n+1 calls at time t, with one
departure in
[ , ] t t + A

( ) ( )
1 1 n n
P t
+ +
A

2. There are n calls at time t, without an arrival
and without a departure in
[ , ] t t + A

( ) ( ) ( )
1 1
n n n
P t A A

3. There are are n-1 calls at time t, with one
arrival in
[ , ] t t + A

( ) ( )
1 1 n n
P t

A


The probability that there are n calls in the system is the sum of the
three probabilities:
( ) ( ) ( )( )( ) ( )
1 1 1 1
1 1
n n n n n n n n
P t P t P t P t
+ +
+ A = A + A A + A

This can be re-written as
( ) ( ) ( ) ( ) ( )
2
1 1 1 1
1
n n n n n n n n n n
P t P t P t P t
+ +
| |
+ A = A + + A + A + A
|
\ .
Subtract
( )
n
P t
from both sides, and divide both sides by A
( ) ( )
( ) ( ) ( ) ( ) ( )
1 1 1 1

n n
n n n n n n n n n n
P t P t
P t P t P t P t
+ +
+ A
= + + A +
A

Take the limit as 0 A
( ) ( ) ( ) ( ) ( )
1 1 1 1

n n n n n n n n
d
P t P t P t P t
dt

+ +
= + +

This is usually written in the form
( ) ( ) ( ) ( ) ( )
1 1 1 1

n n n n n n n n
d
P t P t P t P t
dt

+ +
+ + = +

This the rate balance equation for the system, and is represented
by the diagram

0 1 2
0

n-1
2

n
n+1
2

1 n

1 n

+
2 n

+
2 n

1 n

n
1 n

+
If one is interested in the transient behavior of the system, then time
dependence may be considered, otherwise in the steady-state, it is
taken that the time dependence vanishes, i.e.,
( ) =0
n
d
P t
dt
, and ( )
=
n n
P t P

The balance equations for the steady-state then become
( )
1 1 1 1

n n n n n n n
P P P
+ +
+ = +

A Trunk with many outgoing links
The call arrival rate is constant, and the service is proportional to the
number of calls in the system. That is
n
n
n n


=
= =

These give the balance equations we considered earlier, in considering
the Erlang B formula for the blocking probability.

Example 1.
A group of 20 trunks carries 10 Erlangs and the average call duration
is 3 minutes, calculate

A 10 Erlangs
n
h h
T
= = =



Example 1. A group of 20 trunks carries 10 Erlangs and the average
call duration is 3 minutes, calculate
(a) average number of calls in progress
(b) total number of calls originating per hour.
Given data : No. of trunks = 20
traffic intensity = 10 erlangs
holding time h = 3 minutes
observation period T = 60 minutes.

Solution
(a) Traffic intensity per trunk =
10 Erlangs
20 trunks
= 0.5 Erlangs/trunk
Average no. of calls per trunk for 1 erlang for 60 minutes = 20
For 0.5 erlang, average no. of calls in progress = 10.
(b) Traffic intensity,

A 10 Erlangs
n
h h
T
= = =

total number of calls originating per hour
60
10 200 calls per hour
3
T
n A
h
= = =

Cent call seconds (CCS) (also referred as hundred call seconds)
CCS as a measure of traffic intensity is valid only in telephone circuits.
CCS represents a call time product.
This is used as a measure of the amount of traffic expressed in units of
100 seconds. Sometimes call seconds (CS) and call minutes (CM) are
also used as a measure of traffic intensity. The relation between erlang
and CCS is given by
1E = 36 CCS = 3600 CS = 60 CM



Example 2. Consider a group of 1200 subscribers generating
600 calls during the busy hour. The average holding time is 2.2
minutes. What is the offered traffic in erlangs, CCS and CM?
Given data : n = 600
h = 2.2 minutes
T = 60 minutes.
Solution.
Traffic intensity in Erlangs

600
A 2.2 22 Erlangs
60
n
h h
T
= = = =


Traffic intensity in CCS
600 2.2 60 sec.
792 CCS
100 sec. 100 sec.
n h T
A

= = =

Traffic intensity in CM
( )
call intensity in CCS 100
792 100
1320 CM
60 60
A


= = =

Grade of Service (GoS)
For non-blocking service, it is necessary to provide as many lines as
there are subscribers. But it is not economical for the service provider.
The provider, will provide fewer lines, thereby allowing some calls to
be rejected when the lines are being used by other subscribers (the
rejected calls may be attempted later).
The grade of service refers to the proportion of unsuccessful calls
relative to the total number of calls. GOS is defined as the ratio of lost
traffic to offered traffic.
Blocked Busy Hour calls
GoS
Offered Busy Hour calls
=


0
GoS
A A
A

=


where A
0
= carried traffic (equation 8.4)
A = offered traffic
A A
0
= lost traffic.
The smaller the value of grade of service, the better is the service.
The recommended GoS is 0.002, i.e. 2 call per 1000 offered may lost.
In a system, with equal no. of servers and subscribers, GoS is equal to
zero.

GoS is applied to a terminal-to-terminal connection. But usually a
switching centre is broken into following components

(a) an internal call (subscriber to switching office)
(b) an outgoing call to the trunk network (switching office to trunk)
(c) the trunk network (trunk to trunk)
(d) a terminating call (switching office to subscriber).

The GoS calculated for each component is called component GoS. The
overall GoS is in fact approximately the sum of the component grade
of service.

Example 3. During a busy hour, 1400 calls were offered to a group of
trunks and 14 calls were lost. The average call duration has 3 minutes.
Find (a) Traffic offered
(b) Traffic carried
(c) GoS and
(d) The total duration of period of congestion.
Given data : n = 1400
h = 3 mins
T = 60 mins,
lost calls = 14
Solution



Channel Utilization
The carried load is the position of the offered load that is not lost from
the system. The carried load per line is known as the trunk occupancy.
( ) ( )
1 1 ,
b
A A
P B A N
N N

| |
= =
|
\ .


The trunk occupancy is a measure of the degree of utilization of a
group of lines and is sometimes called the utilization factor. The figure
presents the output channel utilization for various blocking
probabilities and number of servers.

In designing a telephone system, it is necessary to ensure that
the system will operate satisfactorily under the moderate overload
condition.

Example 4. Consider a trunk group with an offered load 4.5
Erlangs and a blocking probability of 0.01. If the offered traffic
increased to 13 erlangs, to keep same blocking probability, find
the number of trunks needed. Also calculate the trunk
occupancies.

Given data : A = 4.5, B = 0.01
From the Figure we obtain
No. of trunks = 10
For the increase in load of 13 erlangs, from the figure
No. of trunks required = 21 for same B = 0.01 required
Hence B(10, 4.5) = B(21, 13) = 0.01
The trunk occupancies calculated as



Therefore, he group of 21 trunks is more efficient than the group
of 10 trunks.

Example 5. A group of 7 trunks is offered 4E of traffic, find (a)
the grade of service (b) the probability that only one trunk is busy
(c) the probability that only one trunk is free and (d) the
probability that at least one trunk is free.
Given data : N = 7, A = 4E

(a) B(7, 4) = 0.052 = GoS.

(b) The probability of only one trunk is busy

(c) The probability that only one trunk is free


(d) The probability that at least one trunk is free

P(k < 7) = 1 P(7) = 1 B = 1 0.052 = 0.948.


Traffic Classification
Circuit-Switched Traffic
number of on-going calls or active connections (Erlangs)
may be converted into bit rate in digital systems
e.g. a telephone call reserves 64 kbps (=8000*8 bps) in a PCM
system

Packet-Switched Traffic
bit stream (bps, kbps, Mbps, Gbps, )
packet stream (pps)
number of active flows (Erlangs)




Circuit-Switched Network
A circuit is established to connect
A and B across the circuit-switched
network

Packet-Switched Network
Each packet follows a different path from
A to B across the packet-switched
network


Circuit-Switched Network
(e.g. Telephone Traffic)
Connection oriented:
connections set up end-to-end before information transfer
resources reserved for the whole duration of connection
if resources are not available, the call is blocked and lost
Information transfer as continuous stream
traffic consists of calls
a call occupies one channel from each of the links along its route
call characterisation: holding time (in time units)

Packet-Switched Network
Connectionless:
no connection establishment
no resource reservations
Information transfer as discrete packets
Best Effort service paradigm
Network nodes (routers) forward packets as well as possible
Packets may be lost, delayed or their order may change
intelligence should be implemented at the edge nodes
or terminals to reorganize the packets,
and/or to correct for errors

Data traffic consists of packets
packet characterisation: length (in data units)

Modelling of offered traffic:
packet arrival process (at which moments new packets arrive)
packet length distribution (how long they are)

Link model: a single server queueing system
the service rate depends on the link capacity and
the average packet length
when the link is busy, new packets are buffered,
if possible, otherwise they are lost


Packet-Level Traffic Characterisation

On top of the network layer (IP) there is the transport layer
takes care of handling the IP packets in the terminals
operates end-to-end
Transport layer protocols:
TCP = Transmission Control Protocol
transmission rate adapts to traffic conditions in the network by a
congestion control mechanism
suitable for non-real time (elastic) traffic, such as transfers of
digital documents
UDP = User Datagram Protocol
transmission rate independent of traffic conditions in the network
suitable for transactions (interactive traffic with short transfers)
used also for real time (streaming) traffic with the help of upper
layer protocols, such as RTP


Routing
Shortest Path Routing
Given the lengths of cable runs, the problem is to determine the
shortest path between any two nodes in the network.

Alternatively, this problem is often encountered in different other
ways: Instead of dealing with length of cable runs, each link has a cost
assigned to it; then the problem is to determine the minimum cost
path between two nodes in the network.


Problem: For the network in the Figure above, find the shortest
route between nodes A and J

The problem can be tackled as a linear mathematical program, but it is
more efficient to use other algorithms. The simplest method is due to
Dantzig and the procedure is as follows:

a) Label the source node as 0.
b) Examine the adjacent nodes and label, each one with its
distance from the source node.
c) Examine nodes adjacent to those already labelled. When a
node has links to two or more labelled nodes, its distance from
each node is added to the label of that node. The smallest sum
is chosen and used as the label for the new node.
d) Repeat (c) until either the destination node is reached (if the
shortest route to only one node is required) or until all nodes
have been labelled (if the shortest routes to all nodes are
required).
The shortest path from node A to node J for the network is shown
in bold. In the Figure below, all steps to determine the shortest path
are illustrated, with the source node A labelled with 0.

The adjacent downstream nodes to A are B and C.
For those nodes the distances are found by adding the label of A with
the distance of nodes from A.
This gives for B 0+ 6 = 6,
and for C 0+10 = 10.
These values are used as labels for B and C respectively.
The next step is to find the adjacent nodes to B and C and then the
labels to these nodes.
For B, there are two adjacent downstream nodes (D and E).
for node D, the label is 11 = (6+5),
and for node E, the label is 12 = (6+6).
Node C has three adjacent downstream nodes (i.e. nodes D,E,F)
To node D is assigned the label 20 = (10+10).
But the node D is also reached through B.
The smallest distance is kept, which is 11, via B and the longer
distance D(20) is eliminated.
This procedure is continued for C for the remaining adjacent
downstream nodes (E, F). The procedure keeps F(15) and
eliminates F(17).
Continuing this way, the procedure stops when all the concerned
nodes have been examined.
In the first Figure, the shortest path from A to J is drawn in bold.
This procedure can be repeated for any node pair in the network, to
establish the required shortest path.



Network Planning
Telecommunication network architectures are changing to
accommodate the requirements of a growing number of
applications (broadband, IP, mobile, multimedia, streaming,
multicasting, etc.)

New-generation technology is to be introduced in the networks,
speeding up the convergence process, obliging planners to apply
different specialized up-to-date planning tools.


Generic Planning Process

Capture of Traffic Characteristics and Market
Demand
Traffic Characteristics
One of the key challenges is to accurately capture the service by its
traffic characteristics related to bandwidth requirements, taking
burstyness of traffic and QoS into consideration. Accurate capture of
service characteristics is necessary in determining the effective
bandwidth for the various services aggregated based on customer
demand is required to size the network appropriately.

Market Demand
Beyond capturing service characteristics, it very important to capture
market demand so that the network capacity is built based on
projected market need rather than just empirical historical trends and
assumptions.
The market needs to be captured based on customer demographic
type
as different customers may subscribe to different
services
estimated number of customers, and
growth or decline over the planning period.

Customer types and population density vary by location, so it is
essential for the planning system to capture the market demand along
with its geographic location.
The population density of a customer type could be distributed over a
geographic region, or there could be aggregated traffic demand at
a specific location such as an enterprise customer or a multiple-
dwelling unit.
For cases where historical data is available, the planning system
should forecast the customer growth or customer decline based on
the historical data from performance monitoring, billing, or network
traffic data systems.
Planning of New and Existing Networks
The planning process should support the planning of new networks as
well as enhancements to existing networks.
In order to plan augmentations to the existing network, it is very
critical to capture the current network topology, including network
elements, configuration, and intra-site and inter-site physical and
logical connectivity.
The equipment model must handle the variety of network elements
from a multitude of vendors spanning all technologies.

I ntra- Domain, I nter- Domain, I nter- Layer Planning
Typically end users employ various applications such as
e-mail, chat,
gaming, and
VoD (video on demand).
These applications are carried as end-user service demand and could
be grouped or categorized into various service types.
For example,
e-mail,
chat, and
IM (Instant messaging)
could be grouped as best-effort
data services,
while
gaming and
VoD
could be categorized as
interactive video services.

These services are typically carried over logical connections across the
network.
The logical connections are carried over the physical network
infrastructure such as network elements and fibre.


Cellular Network Planning

Multi-Layer Network Planning Levels

Dimensions of Network Planning
By Time

Short, Medium, Long term

By Level of Detail

Strategic/Business/Network/NE
By Network Layer

Service layer up to Infrastructure

By Network Segment

Access, Metro, Core, end-to-end
By Technology

PSTN,SDH, Radio, ATM, IP, etc

By Service

POTS, Data, LL, Video, etc

By Solution

Combination of the previous

for a Network Planning Project







Example of Network Architectures and Solutions
Initial status
Network with low deployment level
Heterogeneous areas with varying customer densities
Demand of PSTN and data services in Metro and suburban
Basically POTS demand in rural areas

Target
Network infrastructure growth at high rate
Most economical solutions per scenario
Optimized architecture per area


Wireless Access


Fixed Wireless Access(FWA), Optical in Sub-Urban, and Metropolitan


Wired Access, With Optical in Sub-Urban

The Cell Planning Process





Dimensioning Networks
1. Circuit switched networks
Dimension the capacities (trunks/link speed) such that the total
cost of the networks is minimized with the constraint that the
end-to-end blocking probabilities may not exceed a given limit.
Moes principle

2. Packet switched networks
Square root method
Dimension the link capacities such that the mean delay of a
packet is minimized given an upper bound to the total cost of the
system.

Dimensioning Circuit switched networks
For a single link, Moes principle defines an incremental approach by
looking at the highest gain/cost ratio.
Consider first a single link with capacity N (trunks) and the offered
traffic intensity A.
If one more trunk is added, the intensity of the carried traffic is
increased by the amount
( ) ( )
, 1, A A B N A B N A
| |
A = +
|
\ .

where
( )
, B N A
is Erlang B blocking formula

With the aid of this, it is possible to assess whether the gain obtained
by the additional trunk is big enough to compensate for the costs.
The number of trunks can be increased as far as the benefit exceeds
the costs.
With increasing number of trunks the marginal gain for each added
trunk diminishes.
Example
Given A = 15 Erlangs
Initially N = 11
What is the gain in the Grade of Service when the number of trunks is
increased by 1 ?
What is the corresponding increase in the carried traffic resulting from
this increment ?

Solution
N 10 11 12 13 14 15 16 17 18
B(A,N) 0.410341 0.358792 0.309626 0.263222 0.219983 0.180316 0.144602 0.113153 0.086169
B A


0.051549 0.049166 0.046404 0.043239 0.039667 0.035714 0.031449 0.026984
A A


0.773235 0.737489 0.696059 0.648581 0.594998 0.535714 0.471738 0.404761
N = 11 P
b
= 0.358792
N = 12 P
b
= 0.309626
0.049166 GoS B A = A =

Increase in carried traffic =
0.737489 A A B A = A =

_____________________________________

Routing Plan
The routing plan describes the design of the bearer network and its
use for traffic;
Stipulates rules for:
The structure of the network
node hierarchy, how the switching nodes interwork to
establish connections
Traffic routing
which network paths should be used as first choice,
second choice etc.
Gateways for switching national and international traffic

Two Types of Routing
Static Routing
Fixed routing established by network administrator
Same path followed irrespective of load
inefficient when traffic is dynamic and unpredictable

Dynamic Routing
Routing choices change based on changes in network traffic
Takes advantage of stored program control switches and
common channel signaling (CCS)

Routing Protocols: RIP Routing Information Protocol
OSPF Open Shortest Path First
BGP Border Gateway Protocol
____________________________________
Routing Revisited - Example

Example Network
- Nodes and Links with link costs
(metrics, weights)

- Routing is done based on the
shortest path to the destination


A
D
C
E
B
80
10
10
20
100
10
20
From A

Routing Table
at A

From B

Routing Table
at B

From C

Routing Table
at C

From D


Routing Table
at D


From E

Routing Table
at E

D (20)
C(30)
E(30)
B(50)
A(0)
D (30)
C(20)
E(30)
B(0)
A(50)
D (10)
C(0)
E(10)
B(20)
A(30)
D (0)
C(10)
E(10)
B(30)
A(20)
D (10)
C(10)
E(0)
B(30)
A(30)
Routing Algorithms
- Dijkstras Algorithm
- Bellman-Ford Algorithm

Dijkstras Algorithm


Example:
Compute the shortest paths from A to all other nodes







Homework
1. Use the lost calls held trunking (also called the Molina lost calls held
trunking) to calculate the blocking probability (GoS) when the total
offered traffic is 2 Erlangs and the number of available transmission
channels in the network is 5.

2. Using lost calls cleared, draw two curves for GoS levels of 1% and
10%, interpreting GoS as GoS = P
b
(where P
b
is the blocking
probability). Use the vertical axis as the ratio A/N from 1% to 100%
and the horizontal axis as a number of circuits N from 1 to 20. What
can you say about network utilization when the number of circuits N is
small? How does the utilization of the circuits depend on the allowed
probability of blocking? (channels = circuits = trunks).

3. Using lost calls cleared, what is the approximate capacity of a network
(i.e., how many channels are available) if there are 100 subscribers
and each of whom generates offered traffic of 40 mErlang
(1mErlang=0.001 Erlang) when the probability of blocking is (a) 20%
and (b) 1%? (channels = circuits = trunks).

Access Network Planning:
Requirements - Service and Physical,
Copper Access Network Structure and Planning,
Optical Fibre and Radio Access Systems;

Access Network
The access network provides the connection between the user equipmwnt
customers pre and the local exchange (the local switching centre).
- Copper: two-wires as a subscriber loop
Ordinary telephone and
ISDN subscribers use two wires,
- Optical Fiber
business customers
- Microwave radio
business customers
- Mobile Radio

Access Network Options





























Network Strategy and Planning:
Architecture, Technology, Planning Process,
Multi-layered Network Model,
Dimensioning;

Access Network Planning:
Requirements - Service and Physical,
Copper Access Network Structure and Planning,
Optical Fibre and Radio Access Systems;
Core Transport Planning;

Transport network planning:
Process,
Node Locations,
Fibre Connectivity,
Design Trade-offs,
Planning Tools;

Mobile Network Planning:
Digital Radio Concepts and Standards:
2G, 2.5G (GPRS, EDGE), 3G (WCDMA) and
4G (OFDM, ALL-IP,WLAN);

Architectures, Interfaces, and Protocol Structures of mobile networks,
Radio and core network planning and optimisation,
Signalling, Coverage and propagation modelling,
Traffic dimensioning, Capacity planning, Microwave Link Planning;
Cell Sites Planning,
Macrocell, Microcell, picocell and
Indoor system,

Link-budget calculation,
Forecasting Methods,
Forecasting Traffic Volume and Dispersion.
Traffic Map and Distribution Blocking Probability,
Forecasting Subscriber Distributions,
Size and Location of Switching centres,
Cell Reuses Pattern,
Individual Channel Assignment,
Cell Breathing, Sectorization and Densification,
Power Control,
Frequency hopping,
Discontinuous transmission,

Multi-layer structure (Dual Mode/Band, Overlay and Underlay)
algorithms,
Field Measurement and Network Optimisation,

Network Planning and Design Exercises;

Data Network Planning:
data platforms,
Planning of data platforms:
Internet traffic,
Access to ISPs,
dimensioning,
QoS management.

-------------------------------------------------------------------------
QoS Grade of Service
Congestion
- Queuing system
- Lost-call Systems
Traffic Offered by the network <Actual traffic offered to the network (Overload Traffic is
rejected) - (GOS)
GOS =Ratio of Lost Traffic to offered traffic
Offered Traffic = Avg. No. Calls Generated By Users X
Avg. Holding Time per Call