COMSATS

UNIVERSITY ISLAMABAD

EEE-354: Telecommunication
Systems Engineering

Telephony Traffic Engineering

 Introductory Terminology
• Trunk Line
– High speed lines
that Connects
Switching offices.
• The arrangement of
trunks and switches
within an exchange is
called its trunking
 Tele-Traffic Engineering
• The ordinary subscriber will normally talk to only one
other subscriber at a time.
• Most subscriber do not use their telephones on a full-
time basis.
• The exact amount of common equipment required is
unpredictable because of the random nature of the
traffic.
n=50 ->transmission
lines wastage
n=1 -> congestion
leading to poor QoS
 Tele-Traffic Engineering
Objective:

Given expected traffic (+ growth assumptions), Provision

resources (trunks, frequency band etc)...
To minimize cost ...
Subject to minimum acceptable quality of service requirements.

Teletraffic theory is defined as the application of probability theory

to the solution of problems concerning planning, performance
evaluation, operation and maintenance of telecommunication
systems.
 Characterization of Telephone Traffic
Call Arrival Rate(λ):
– Average number of calls initiated per unit time (e.g. attempts
per hour)
– Each call arrival is independent of other calls (we assume)
– Call attempt arrivals are random in time
• If n calls (on average) are received during a period T seconds,
!
the calling rate is defined as, λ =
"

- (poisson formula)

The average calling rate is measured in calls per hour.

Source: Wireless Communications: Theory and Techniques By Asrar U.H. Sheikh

 Characterization of Telephone Traffic
Average holding time (h): The average duration of
occupancy of a traffic path by a call.
– For voice traffic, it is the average holding time in hours/call

– Probability of lasting time t or more is –ve exponential in

nature
h= 1/μ = average service time of a customer

– The reciprocal of the average holding time is referred to as

service/Departure rate (µ) in calls per hour is given as
!
𝜇=
"
 Characterization of Telephone Traffic
Example: On average one call arrives every 3
seconds. During a period of 12 seconds, what is
the probability that

a. No call arrives
b. One call arrives
c. More than one call arrives
 Characterization of Telephone Traffic
Example: In a telephone system, the average call
duration is 3 minutes. A call has already lasted 5
minutes. What is the probability that

• The call will last at least another 3 minutes

• The call will end within the next 3 minutes

 Traffic Measurement
Traffic=calling rate + holding time

• One measure of network capacity is the volume of traffic (traffic

density) carried over a period of time.
– Traffic volume is essentially sum of all holding times carried during the
interval.
– Formula:
– Units
• Erlang * hour
 Traffic Measurement
– A more useful measure of traffic is the traffic intensity (also called traffic
flow, carried traffic, average occupancy, traffic load).
– For a single user :
• Traffic intensity is obtained by dividing the traffic volume by the length of time during
which it is measured.
• Although traffic intensity is fundamentally dimensionless (time divided by time), it is
usually expressed in units of erlangs.
• A: traffic intensity C: number of calls arrival during time T(busy hour) , h: average
holding time
A=(C.h)/T
– Average number of calls in progress or the ratio of average arrival rate to
the average service rate.
Example:
With an arrival rate of 100 calls per hour, and each call requiring 9 minutes
(0.15 hour) of service,
Traffic volume in an 8-hour day is:
100 * 0.15 * 8 = 120 Call Hours (Ch)
Traffic intensity is 120 / 8 = 15E.
 Traffic Measurement
• Erlangs - The Erlang defines the efficiency (percent
occupancy) of a traffic resource and represents the total
time in hours to carry all calls. It is the traffic unit used
exclusively in classic traffic theory.
A single trunk can at
most can carry one
Erlang of the traffic!

One hour(normally)of continuous use of one channel = 1 Erlang

– 1 Erlang = 1 hour (60 minutes) of traffic
• In data communications, a 1E = 64 kbps of data
• 1 Erlang is equivalent to traffic intensity that keeps:
– one circuit busy 100% of the time, or
– two circuits busy 50% of the time, or
– four circuits busy 25% of the time, etc.
 Traffic Measurement
• CCS( Centum Call Seconds) - measures the exact same traffic
intensity as the Erlangs but expresses it as the number of 100
second holding times required per hour. Traffic registers
sample stations every 100 seconds per hour to check for busies.
Since there are 36 sets of hundred seconds in an hour
– 1 Centum Call Second (CCS) is equal to 100 call seconds
– 36 CCS = 1 Erlangs
– Voice Traffic is generally measured as CCS. Depends on the equipment
and the unit of measurement it records in.
100 call × seconds
1 ccs hour = = 0.027E
1 hour × 60 min hr × 60 sec min
3600 call × seconds
36 ccs hour = = 1E
1 hour × 60 min hr × 60 sec min
 Traffic Measurement
Example:
If a group of user made 30 calls in one hour, and each call
had an average call duration of 5 minutes, then the number
of Erlangs this represents is worked out as follows:
Minutes of traffic in the hour = number of calls x duration
Minutes of traffic in the hour = 30 x 5
Minutes of traffic in the hour = 150
Hours of traffic in the hour = 150 / 60 Traffic Intensity > 1E=queuing delay
Hours of traffic in the hour = 2.5
Traffic figure = 2.5 Erlangs
Traffic Intensity<1E=can accommodate
Example: more traffic
A subscriber makes four telephone calls for 3 minutes, 3
minutes, 2 minutes and 1 minute. Calculate the subscriber’s
traffic in Centum Call Seconds (CCS).
(3+3+2+1)×60
Traffic in CCS = =5.4 CCS
!""
 Busy Hour
• Telephone traffic may fluctuate throughout the day, and may
have a “busy hour”.
• “The continuous 1-hour period lying wholly in the time interval
concerned for which the traffic or the number of call attempts is
greatest.”
• The measurements for computing normal traffic load should be
the 30 highest days in a fixed 12-month period.(Mean)
• Within the busy hour, traffic is considered to be stationary and
thus the recorded intensity is the mean value during the busy
hour.
• In order to determine an appropriate dimensioning load, the
recommendations define how the busy hour traffic shall be
measured
• ITU Recommendation E.600-ADPH,TCBH,FDMH

Source: Introduction to Telecommunications Network Engineering By Tarmo Anttalainen

Busy Hour
 Tele-traffic models
• Two phases of tele-traffic modelling:
– Modelling of the incoming traffic " traffic model
– Modelling of the system itself " system model
• The system model can be further divided in
– Loss systems (loss models)
– Waiting/queueing systems (queuing models)
• The appropriate analysis category for a
particular system depends on the system’s
treatment of overload traffic.
 Traffic Measurement
• In a loss system, there are three types of call rates:
– λoffered (arrival rate of all call attempts )
– λcarried (arrival rate of carried calls)
– λlost (arrival rate of lost calls )
• The three call rates lead to the following three traffic concepts:
– Traffic offered Aoffered = λoffered h
– Traffic carried Acarried = λcarried h
– Traffic lost Alost = λlost h
– Bc-Call Blocking Probability

λoffered = λcarried + λlost = λ

λcarried =λ(1−Bc)
λlost = λBc
 Traffic Measurement
• Offered traffic is the traffic intensity that would occur if all
traffic submitted to a group of circuits could be processed.
• Carried traffic is the traffic intensity actually handled by the
group.
• Blocked traffic is that portion of traffic that cannot be processed
by the group of circuits (I.e. offered traffic minus carried traffic).
• Blocked traffic may be rejected, retried or offered to another
group of circuits (overflow).
Circuit Utilization (r) - also called Circuit
Efficiency Acarried
T
proportion of time a circuit is busy, or r= C
average proportion of time each circuit # of Trunks
in a group is busy
 Grade of Service (GoS)
• Grade of Service is a measure of the probability that a
percentage of the offered traffic will be blocked or delayed.
• GOS is a measure of the ability of a user to access a trunked
system during the busiest hour
• The larger the value of GOS, the worse is the service given.
• The grade of service is normally specified for the traffic at
busy hour.
• For a lost-call system, the grade of service B may be defined
as: No. of calls Lost Traffic Lost
B= B=
No. of calls offered Traffic offered

• Example: On average, one call in 100 will be blocked ,

B=0.01
 Grade of Service (GoS)
• It is not reasonable to dimension the network for a very small
blocking probability, since the call may be unsuccessful due to
other reasons with a much higher probability:
- Subscriber does not answer
- Subscriber busy
- One has dialed a wrong number
• Typical Values:
– in busy hour, range from 0.2% to 5% for local calls, however
– generally no more that 1%
– long distance calls often slightly higher
• There may be reattempts after unsuccessful calls.
• In the domain of traffic flow analysis, the blocking probability
analyses are referred to as congestion theory and delay
analyses are referred to as queuing theory.
• In systems with queuing, gos often defined as the probability of
delay exceeding a specific length of time

20
 Grade of Service vs Quality of Service
“The collective effect of service performance, which determine
the degree of satisfaction of a user of the service” (ITU-T E.800).
A number of traffic engineering variables to provide a measure
of adequacy of a group of resources under specified conditions
(ITU-T E.600).

Quality of service (QoS) comprises requirements on all the

aspects of a connection, such as signal-to-noise ratio, echo,
interrupts, loudness levels, and so on.
Grade of Service (GoS) requirements, which comprises aspects
of a connection relating to capacity and coverage of a network,
for example guaranteed maximum blocking probability and
outage probability.
 Handling of Lost Calls
• Lost calls held (LCH): The telephone user will immediately reattempt the call
on receipt of a congestion signal and will continue to redial.
– The call is held for Average call holding time-A system design parameter
– If the channel becomes available during the holding time, the call is
served for the remainder of average call duration otherwise it is cleared.
– Not a good model of real world behaviour (mathematical approximation
only)
– Tries to approximate call reattempt efforts

• Lost calls cleared (LCC): The user will hang up and wait some time interval
before reattempting if the user hears the congestion signal on the first
attempt. Reattempting -> new call

• Lost calls delayed (LCD): The user is automatically put in queue (a waiting
line or pool). For example, this is done when the operator is dialed.
– Held in Queue for infinite time.

Source: Wireless Communications: Theory and Techniques By Asrar U.H. Sheikh

 Blocked Calls Cleared (BCC)
2 sources
10 minutes

Source #1
Offered Traffic
1 3
Total Traffic Offered:
Source #2 TO = 0.4 E + 0.3 E
Offered Traffic TO = 0.7 E
2 4

Only one server 1st call arrives and is served

2nd call arrives but

Traffic server already busy
Carried
1 2 3 4
2nd call is cleared

3rd call arrives and is served

Total Traffic Carried:
TC = 0.5 E 4th call arrives and is served
 Blocked Calls Held (BCH)
2 sources
10 minutes

Source #1
Offered Traffic
1 3
Total Traffic Offered:
Source #2 TO = 0.4 E + 0.3 E
Offered Traffic TO = 0.7 E
2 4

1st call arrives and is served

Only one server
2nd call arrives but server busy
Traffic
Carried 2nd call is held until server free
1 2 2 3 4
2nd call is served

3rd call arrives and is served

Total Traffic Carried:
TC = 0.6 E 4th call arrives and is served
 Blocked Calls Wait (BCW)
2 sources
10 minutes

Source #1
Offered Traffic
1 3
Total Traffic Offered:
Source #2 TO = 0.4 E + 0.3 E
Offered Traffic TO = 0.7 E
2 4

1st call arrives and is served

Only one server 2nd call arrives but server busy

2nd call waits until server free

Traffic
Carried 2nd call served
1 2 2 3 4
3rd call arrives, waits, and
is served
Total Traffic Carried:
4th call arrives, waits, and
TC = 0.7 E is served
 Poisson Blocking Model
• In the United States the Poisson formula is
favored.
• The Poisson formula has the following
assumptions: • No of Trunks
• Traffic Intensity
– Infinite sources, • Grade of Service

– Equal traffic density per source, and

– Lost calls held (LCH).
N -1
Ak - A A= Mean # of
P( B) = P( N , A) = 1 - å e Busy Sources
k = 0 k!
 Example Poisson Blocking
• Note that P(N,A) assumes N is given and we calculate the blocking from A
Erlangs of offered traffic
• But in a design orientation we more naturally want to soilve the incverse
problem, i.e., what should N be to serve A at not more than a target P(B).
• Or, consider a g.o.s requirement of 1% blocking in a system with N=10
trunks
– How much offered traffic can the system handle?

Typical notation: Example:

P( B) = P( N , A) P(7,10)

“P” = Poisson Poisson P(B) with 10 E

“A” = Offered Traffic offered to 7 servers
“N” = No of servers

• How do we solve this equation for A that produces P(N,A) = 1%?

 Poisson Traffic Tables
P(B)=P(N,A)

A. If system with N = 10 trunks

has P(B) = 0.01:

System can handle

Offered traffic (A) = 4.14 E
 Efficiency of Large Groups
• What if there are N = 100 trunks?
– Will they serve A = 10 x 4.14 E = 41.4 E with same P(B) = 1%?
– No!
– Traffic tables will show that A = 78.2 E!
• Why will 10 times trunks serve almost 20 times traffic?
– Called efficiency of large groups:
A 4.14
For N = 10, A = 4.14 E r= = = 41.4% efficiency
N 10
A 78.2
For N = 100, A = 78.2 E r= = = 78.2% efficiency
N 100

The larger the trunk group, the greater the efficiency (at target loading)
We already know this effect qualitatively as “economy of scale”
 Erlang B Traffic Model
• Erlang B formula deals with offered traffic, which differs
from carried traffic by the number of lost calls.
• More sophisticated model than Poisson
• It is generally accepted as a standard outside the United
States.
• i.e., while Poisson is a pessimistic estmate (due to BCH
effect), Erlang B is either exact (overflow route case) or an
optimistic estimate because of its BCC assumption.

• The Erlang B formula assumes

– Infinite sources,
– Equal traffic density per source, and
– Lost calls cleared (LCC).
The formula is
 Erlang B Traffic Tables
B(N,A)=0.001

Example: In a BCC system with m=¥

sources, we can accept a 0.1% chance
of blocking in the nominal case of 40E B(N,A)=0.005
offered traffic. However, in the
extreme case of a 20% overload, we
can accept a 0.5% chance of blocking.
Q. How many outgoing trunks do
we need?
N=59 A=40 E
Nominal design: 59 trunks
Overload design: 64 trunks A»48 E

A. Requirement is: 64 trunks

N=64
 Example (2)
P(N,A)=0.01

N=32

A=20.3 E
Compare Poisson & Erlang B at High Blocking Probability
• We recognize that Poisson and Erlang B models are only approximations but
which is better?
– Compare them using a 4-trunk group offered A=10E
Erlang B Poisson

B(4,10) = 0.64666 P(4,10) = 0.98966

TC = A ´ (1 - P( B)) = 10 ´ (1 - 0.64666) TC = A ´ (1 - P( B)) = 10 ´ (1 - 0.98966)

TC = 3.533E TC = 0.103E
3.533 0.103
r= = 0.88 r= = 0.026
4 4

How can 4 trunks handle 10E offered

traffic and be busy only 2.6% of the time?
 P(N,A) & B(N,A) at High Blocking
• Obviously, the Poisson result is far off.
– 4 servers offered enough traffic to keep 10 servers busy full
time (10E) should result in much higher utilization.
• Erlang B result is more believable.
– All 4 trunks are busy most of the time.
• What if we extend the exercise by increasing A?
– à Erlang B result goes to 4E carried traffic
– à Poisson result goes to 0 E carried traffic!
• The “take-away” messages are:
– Poisson only good approximation when low blocking
– Use Erlang B if high blocking
 References
• ITU Teletraffic Engineering Handbook
• Telecommunication System Engineering Roger L. Freeman
• E.500 : Traffic intensity measurement principles
• E.600 : Terms and definitions of traffic engineering