TELETRAFFIC ENGINEERING

BASIC TELETRAFFIC
CONCEPTS

1 User Making Phone Calls

TRAFFIC is a stochastic process
BUSY 1

IDLE 0
time

How to characterize this process?

Statistical distribution of the BUSY period

statistical distribution of the IDLE period
Statistical characterization of the process memory

E.g. at a given time, does the probability that a user starts a call result
different depending on what happened in the past?
2

Traffic Characterization
Suitable For Fraffic Engineering
amount of busy time in t

t
t
average number of calls per min average call duration min

traffic intensity A i lim

probability that, at a random time t, user is in BUSY state

mean process value

All equivalent (if stationary process)

3

Traffic Intensity: Example

User makes in average 1 call every hour

Each call lasts in average 120 s
Traffic intensity =

120 sec / 3600 sec = 2 min / 60 min = 1/30

Probability that a user is busy

User busy 2 min out of 60 = 1/30

adimensional
4

Traffic Generated By More

Than One Users
U1

Traffic intensity
(adimensional, measured in Erlangs):
4

U2

A Ai 4 Ai
i 1

U3
U4

4
4k
Pk active calls Aik 1 Ai
k
E active calls 4 Ai A

TOT

Example

5 users
Each user makes an average of 3 calls per hour
Each call, in average, lasts for 4 minutes

1
calls 4
hours erl
Ai 3

5
hour 60
1
A 5 erl 1erl
5
number of active users
Meaning: in average, there is 1 active call;
but the actual number of active calls varies
from 0 (no active user) to 5 (all users active),
with given probability

0
1
2
3
4
5

probability
0,327680
0,409600
0,204800
0,051200
0,006400
6
0,000320

Second Example

30 users
Each user makes an average
of 1 calls per hour
Each call, in average, lasts
for 4 minutes

4
A 30 1 2 Erlangs
60
SOME NOTES:
-In average, 2 active calls (intensity A);
-Frequently, we find up to 4 or 5 calls;
-Prob(n.calls>8) = 0.01%
-More than 11 calls only once over 1M
TRAFFIC ENGINEERING: how many
channels to reserve for these users!

n. active users
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

binom
1
30
435
4060
27405
142506
593775
2035800
5852925
14307150
30045015
54627300
86493225
119759850
145422675
155117520
145422675
119759850
86493225
54627300
30045015
14307150
5852925
2035800
593775
142506
27405
4060
435
30
1

probab
1,3E-01
2,7E-01
2,8E-01
1,9E-01
9,0E-02
3,3E-02
1,0E-02
2,4E-03
5,0E-04
8,7E-05
1,3E-05
1,7E-06
1,9E-07
1,9E-08
1,7E-09
1,3E-10
8,4E-12
5,0E-13
2,6E-14
1,2E-15
4,5E-17
1,5E-18
4,5E-20
1,1E-21
2,3E-23
4,0E-25
5,5E-27
5,8E-29
4,4E-31
2,2E-33
5,2E-36

cumulat
0,126213
0,396669
0,676784
0,863527
0,953564
0,987006
0,996960
0,999397
0,999898
0,999985
0,999998
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000
1,000000

A Note On Binomial Coefficient

Computation
60
60!

1.39936e 12
12
12
!
48
!

but 60! 8.32099e 81 (overflow problems! ! )
60
60
exp log explog 60! log 12! log 48!
12
12
12
48

60
exp log i log i log i (no overflow! ! before exp...)
i 1
i 1

i 1

60 12
Ai 1 Ai 48
12
12
48
60

exp log i log i log i 12 log Ai 48 log 1 Ai

i 1
i 1
i 1

(no overflow! ! never! )

Infinite Users
Assume M users, generating an overall traffic intensity A
(i.e. each user generates traffic at intensity Ai =A/M).
M
A
We have just found that

k 1
M

Pk active calls, M users

k

k
M! A M
M k
Ai 1 Ai

M
k
k

!
!
M
A

1
M

Let Minfinity, while maintaining the same overall traffic intensity A

M
k
M!
1 Ak
A
A
Pk active calls, users lim

1 1
M M k ! k! M k
M M
M

k
A
M M 1 M k 1
A A
1

lim

M
k! M
Mk

1
M

Ak
k!

A
9

Poisson Distribution
30%
poisson
binomial (M=30)

A=2 erl

25%
20%

A=10 erl

15%
10%
5%
0%
0

Pk A e

A
k!

10

12

14

16

18

20

22

Very good matching with Binomial

(when M large with respect to A)
Much simpler to use than Binomial
(no annoying queueing theory complications)

10

Limited number of channels

THE most important problem
in circuit switching

U1

The number of channels C is

less than the number of users U2
M (eventually infinite)
Some offered calls will be
U3
blocked
What is the blocking
probability?
U4
We have an expression for

X
X

P[k offered calls]

We must find an
expression for

P[k accepted calls]

TOT

As:

No. carried calls versus11 t

No. offered calls versus t

P[block ] PC accepted calls

Channel Utilization Probability

offered traffic: 2 erl - C=3

C channels available
Assumptions:
Poisson distribution (infin.
users)
Blocked calls cleared
It can be proven (from
Queueing theory) that:

35%
30%

offered calls
accepted calls

25%
20%
15%
10%

P[k calls in the system, k (0, C)] 5%

Pk offered calls
C

0%
0

Pi offered calls
i 0

(very simple result!)

P[system full] P[C accepted calls ]

Hence:

PC offered calls
C

Pi offered calls

12

i 0

Blocking Probability: Erlang-B

Fundamental formula for
telephone networks planning
Ao=offered traffic in Erlangs

AoC
block C C! j E1,C Ao
Ao

j 0 j!

Efficient recursive computation

available

E1,C Ao

Ao E1,C 1 Ao
C Ao E1,C 1 Ao

100,00%

blocking probability

10,00%
1,00%
C=1,2,3,4,5,6,7

0,10%
0,01%
0

2
3
offered load (erlangs)

13

NOTE: Finite Users

Erlang-B obtained for the infinite

users case
It is easy (from queueing theory)
to obtain an explicit blocking
formula for the finite users case:

ENGSET FORMULA:

M 1

A
C

C
M 1
k

Ai

k 0
i
C
i

block

Ao
Ai
M

Erlang-B can be re-obtained

as limit case
Minfinity
Ai0
MAiAo
Erlang-B is a very good
approximation as long as:
A/M small (e.g. <0.2)
In any case, Erlang-B is a
conservative formula
yields higher blocking
probability
Good feature for planning
14

Capacity Planning

Target: support users with a given Grade Of Service

(GOS)
GOS expressed in terms of upper-bound for the
blocking probability

GOS example: subscribers should find a line available in

the 99% of the cases, i.e. they should be blocked in no
more than 1% of the attempts

Given:

C channels
Offered load Ao
Target GOS Btarget

C obtained from numerical inversion of

Btarget E1,C Ao
15

Channel Usage Efficiency

Carried load (erl)

Offered load (erl)

Ao

C channels

Ac Ao 1 B

Ao B
Blocked traffic

efficiency : h

Ac Ao 1 E1,C Ao

C
C

Ao
if small blocking
C

Fundamental property: for same GOS, efficiency

increases as C grows!! (trunking gain)

16

Example

blocking probability

100,0%
A = 40 erl
A = 60 erl
A = 80 erl
A = 100 erl

10,0%

1,0%

0,1%
0

20

40

GOS = 1% maximum blocking.

Resulting system dimensioning
and efficiency:

60
capacity C

80

100

40 erl C >= 53 h = 74.9%

60 erl C >= 75 h = 79.3%
80 erl C >= 96 h = 82.6%
100 erlC >= 117 h = 84.6%

120

17

Erlang B Calculation - Tables

18