Waiting Line and

Queuing Theory
CHARACTERISTICS OF A WAITING-LINE
SYSTEM

– Arrival Characteristics
– Waiting-Line Characteristics Service Facility
Characteristics
– Measuring the Queue’s Performance
– Queuing Costs
 You’ve Been There Before!

‘The other line Thank you for holding.

Hello...are you there?
always moves faster.’
‘If you change lines, the
one you left will start to
move faster than the one
you’re in.’

© 1995 Corel Corp.

 WAITING LINES
• First studied by A. K. Erlang in 1913
– Analyzed telephone facilities

• Body of knowledge called queuing theory

– Queue is another name for waiting line

• Decision problem
– Balance cost of providing good service with
cost of customers waiting
 Deciding on the
Optimum Level of Service
Cost

Total expected cost

Minimum Cost of providing

total cost service

Negative Cost of
waiting time to
company
Low level Optimal High level
of service service level of service
 Waiting Line Terminology

• Queue: Waiting line

• Arrival: 1 person, machine, part, etc. that arrives
and demands service
• Queue discipline: Rules for determining the
order that arrivals receive service
• Channel: Number of waiting lines
• Phase: Number of steps in service
Three Parts of a Queuing System at
Dave’s Car-Wash
 INPUT CHARACTERISTICS

Input Source
(Population)

Size Arrival
Pattern

Infinite Finite Random Non-

Random

Poisson Other
 Poisson Distribution
• Number of events that P(X)  = 0 .5
.6
occur in an interval of
.3
time
.0 X
– Example: Number of
0 1 2 3 4 5
customers that arrive in
15 min.
• Mean =  (e.g., 5/hr.) P(X)  = 6 .0
.6
• Probability:
.3
.0
e   x X
P( x )  0 2 4 6 8 10
x
 Waiting Line Characteristics

Waiting Line

Length Queue
Discipline

FIFO
Unlimited Limited (FCFS) Random Priority
 SERVICE CHARACTERISTICS

Service
Facility

Configuration

Single Multi-
Channel Channel

Single
Phase
 Single-Channel, Single-Phase System

Service system
Served
Arrivals Queue units
Service
facility

Ships at Ship unloading system Empty

sea Waiting ship line ships
Dock
 Single-Channel, Multi-Phase System
Service system
Served
Arrivals Queue units
Service Service
facility facility

Cars McDonald’s drive-through

Cars
in area Waiting cars & food
Pay Pick-up
 Multi-Channel, Single Phase System

Service system
Served
units
Service
Arrivals Queue facility

Service
facility

Example: Bank customers wait in single line for one

of several tellers.
 Multi-Channel, Multi-Phase System

Service system
Served
units
Service Service
Arrivals Queue facility facility

Service Service
facility facility

Example: At a Laundromat, customers use one of several

washers, then one of several dryers. At Passport office too
 Waiting-Line Performance Measures

• Average queue time, Wq

• Average queue length, Lq
• Average time in system, Ws
• Average number in system, Ls
• Probability of idle service facility, P0
• System utilization, 
• Probability of k units in system, Pn > k
Assumptions of the Basic Simple Queuing
Model
• Arrivals are served on a first come, first served basis
• Arrivals are independent of preceding arrivals
• Arrival rates are described by the Poisson probability
distribution, and customers come from a very large
population
• Service times vary from one customer to another, and are
independent of one and other; the average service time is
known
• Service times are described by the negative exponential
probability distribution
• The service rate is greater than the arrival rate
 Types of Queuing Models

• Simple (M/M/1)
– Example: Information booth at mall
• Multi-channel (M/M/S)
– Example: Airline ticket counter
• Constant Service (M/D/1)
– Example: Automated car wash
• Limited Population
– Example: Department with only 7 drills
 Remember:  &  Are Rates

•  = Mean number of
If average service time is 15
arrivals per time period minutes, then μ is 4
– e.g., 3 units/hour customers/hour

•  = Mean number of
people or items served
per time period
– e.g., 4 units/hour
• 1/ = 15 minutes/unit
Derivation of (M/M/1):(GD/∞/∞)
N=Large no. of ‘N’ Systems
Pn=prob. That ‘n’ persons are present in a system
λn = arrival rate (Poisson)
μn = service rate
λ/μ= ρ= traffic intensity taken as <1

For steady state behavior , number ‘n’ will not change with time
but system with ‘n’ present may not be same as system with
‘n’ present at other moment of time
Rate at which ‘n’ system become n+1 = λn *Pn*N
Rate at which ‘n’ system become n-1 = μn *Pn*N
Combined rate which ‘n’ system change to either (n+1) or (n-1)
λn *Pn*N + μn *Pn*N = (λn +μn )Pn*N
Because no. ‘n’ remains constant the rate of change must be
equal to the rate at which other non ‘n’ system become ‘n’
systems
λ n-1 *Pn-1 *N + μn+1 *Pn+1*N = (λn +μn )Pn*N eq.1

For n=0 , λ n-1, Pn-1 & μ0 have no meaning , Hence

μ1*P1 = λ0*P0

P1 = λ0*P0
μ1
OR λ0*P0 - μ1*P1 =0
Re-arranging eq. 1
λn *Pn - μn+1 *P = λn-1 *P
n+1 n-1 - μn *P n

For n=1
λ1*P1 - μ2 *P2 = λ0 *P0 - μ1 *P1 =0

λ1*P1 - μ2 *P2 =0

P2 = λ1*P1 λ 1* λ 0 * P0
μ2 μ1 μ2
 λ0. λ1 …………………………….. λn-1 *Po
Pn = μ1. μ2 ………………….μn

Also P0 +P1 +……………………..=1 eq.4

For single channel, constant arrival and service rate.
λ= λ0 = λ1
μ = μ 1 = μ2
Pn = λn * P 0

μn
P n = ρn * P 0 eq.5
• Substitute(5) into (4)
P0 + ρP0 +ρ2P0 +……………=1
P0 (1 + ρ + ρ2 +…………..)=1

NOW 1 + ρ + ρ2 +…………..= 1/1 – ρ as λ/μ<1

Hence P0 /(1- ρ) =1

P0 = 1- ρ eq. 6
Put (6) into (5)
Pn = ρn (1-ρ) eq. 7
• Average no. of customer in the system

n  n.Pn
n 1

= P1 + 2P2 + 3P3+……………
from (7)
= ρ (1-ρ) + 2.ρ2 (1-ρ) + 3.ρ3 (1-ρ) +
………..
= (1-ρ)( ρ + 2.ρ2 + 3.ρ3 +…………. )
= (1-ρ) ρ (1+ 2.ρ + 3.ρ2 +…………. ) 1/ (1-ρ)2
= (1-ρ) ρ/ (1-ρ)2
= ρ/(1-ρ)
• With μ as service rate , avg. waiting time of the
customer is ρ / (1-ρ) μ

average no. in the queue(not counting one in service)

nq = 1*P2 + 2*P3 + 3*P4 +……………
= ρ2 (1-ρ) + 2.ρ3 (1-ρ) + 3.ρ4 (1-ρ) + ………..

= ρ2 (1-ρ) (1+ 2.ρ + 3.ρ2 +…………. )

= ρ2 (1-ρ) / (1-ρ)2

= ρ2 / (1-ρ)
 Simple (M/M/1) Model Equations

Average number of units Ls =
 -
in queue
1
Average time in system Ws =
 -
2

Average number of units in Lq =
 ( -  )
queue

Average time in queue Wq =
 ( -  )

System utilization  =

Simple (M/M/1) Probability Equations

Probability of 0 units in system, i.e., system idle:


P = 1-  = 1-
0 
Probability of more than k units in system:

()l k+1
P = 
n>k
Where n is the number of units in the system
 Model B (M/M/S) Equations

Average number of

people or units Lq  L s 

waiting for service:

Average time a
1
person or unit Wq Ws 

spends in the
queue
Example:- The arrival rate of customers at a banking counter
follows Poisson distribution with a mean of 45 per hour . The
service rate follows Poisson distribution with a mean of 60
per hour . Calculate

(1)-what is the prob. Of having no customer in the system (P0)

(2) what is the prob. Of having ‘5’ customer in the system (P5)

(3)Calculate avg. no. customers in the system

P0 = (1-ρ) = 1 - 0.75 = 0.25

P5 = ρ5 (1-ρ) = (0.25) *(0.75)5 = 0.0593

• The multichannel case:
consider a queuing process that has ‘m’ channels,
each with service rate μ when the no. customers
present (n) is less than or equal to the no. of
channel, i.e.
When n≤m, there will be ‘n’ busy channel and service
will be at the rate nμ, when n>m, all ‘m’ channels
are busy and service is at a rate mμ, hence
λn = λ = n=0,1,2…………
μn = nμ n=0,1,2, ……..m
mμ n= m+1, m+2……….
ρ= λ/μ <m (λ/μ <1, for single channel
λ/mμ <1 for multi channel
• Substitute value of λn and μn in:

0 .1.2 .. n 1

Pn = * P0
1.2 .n

n n
*P0 * P0
Pn =  *2 *3 .......* n =  * n!
n

n
= n ! * P0 for n= 0,1,2,……………m
 n
Pn = * P0
 * 2 *3 .......* m * (m ) n m

[Till n=m, service rate is mμ, after that for remaining

(n-m) the service rate will also be mμ]

 n n
* P0  n m
* P0
 * m!*(m)
n n m m !* m
Substitute value of Pn from (1) and (2) in the following
eq. P0 +P1 +……………………..= 1

2  m 1 m
P0 +ρ P + 0 * P0............................... + m  1! * P0
+
m!
* P0
2!
 m1  m 2
 *P + 2
* P0  ................. 1  (3)
m ! m 0
m ! m

The first ‘m’ term can’t be examined in closed

form, but the remaining can be simplified.
 m  m 1  m 2
* P0 + * P0  2
* P0 ................
m! m ! m m ! m

m   2 3 
= * P0  1   2  3 ................ 
m!  m m m 

This is a G.P. with ratio ρ/m (a+ar+ar2+…………..= a/1-r)

=1/ (1- ρ/m)
When we solve eq.(3) for P0 we get:

1
P0 
2  m 1 m
1     .......................  
2! m  1!  
m !*  1  
 m