Queuing Theory

© Macmillan Publishers India Ltd

1
1997,2003,2007,2009
“Good managers have a bias for action.”
Thomas J Peters

© Macmillan Publishers India Ltd

2
1997,2003,2007,2009
 Introduction

Queuing theory can be applied to a variety of operational situations where it is not possible to
predict accurately the arrival rate (or time) of customers and service rate (or time) of service
facility (or facilities). Thus it is needed to determine the level of service (either the service rate or
the number of service facilities) that balances the following two conflicting costs:

 cost of offering the service

 cost incurred due to delay in offering service

The first cost is associated with the service facilities and their operation, and the second
represents the cost of customer’s waiting time.
Since cost of waiting is difficult to estimate, it is usually measured in terms of loss of sales or
goodwill when the customer who is a human being has no sympathy with the service. But, if the
customer is a machine waiting for repair, then cost
© Macmillan of waiting
Publishers India Ltd is measured in terms of the cost of
3
1997,2003,2007,2009
 Queuing Theory

Figure illustrates both types of costs as a function of level of service. The optimum service level
is one that minimizes the sum of the two costs.

© Macmillan Publishers India Ltd

4
Queuing Costs vs Level of Service
1997,2003,2007,2009
 Queuing theory

Examples of Queues

Situation Customers Service Facilities

 Petrol pumps (stations) Automobiles Pumps/Passionel

 Hospital Patients Doctors/Nurses/Rooms

 Airport Aircraft Runways

 Post office  Letters Sorting system

 Job interviews Applicants Interviewers

© Macmillan Publishers India Ltd

5
1997,2003,2007,2009
 The Structure of Queuing System

The Major Components of a Queuing System

 Calling population (or input source)
 Service process (or mechanism)
 Queue discipline
 Queuing process

© Macmillan Publishers India Ltd

6
1997,2003,2007,2009
 The Structure of Queuing System

Calling Population

The calling population need not be homogeneous and may consist of several subpopulations. For
example, patients arriving at OPD of a hospital are usually of three categories: walk-in patients,
patients with appointments and emergency patients. Each patient class places different demands
on service facility, but the waiting expectations of each category differ significantly.

The calling population is characterized by

 its size

 behaviour of arrivals

 pattern of arrivals

 arrival time distribution

© Macmillan Publishers India Ltd
7
1997,2003,2007,2009
 The Structure of Queuing System

Size of Calling Population

whether it is homogeneous or consist of several subpopulations is considered to be either finite

(limited) or infinite (unlimited). If customer’s arrival depends on the number of customers
already in the system (in service plus in queue), the calling population is called limited or finite.

For example, A factory having only four machines which often require repair/service and any
of these in working condition is an example of finite population. Then at any point in time, the
remaining machines could possibly require service.

However, if new customer’s arrival is independent of the number of customers already in the
system, the calling population is called unlimited or infinite. For example, customers arriving at a
bank or super market, students arriving to get admission at the university, cars arriving at a
highway petrol pump, etc. © Macmillan Publishers India Ltd
8
1997,2003,2007,2009
 The Structure of Queuing System

Behaviour of Arrivals

 Balking: Customers do not join the queue either by seeing the number of customers already
in service system or estimating the excessive waiting time for desired service.

 Reneging: Customers after joining the queue, wait for sometime in the queue but leave before
being served on account of certain reasons.

 Jockeying: Customers move from one queue to another hoping to receive service more
quickly (a common scene an railway booking window).

© Macmillan Publishers India Ltd

9
1997,2003,2007,2009
 The Structure of Queuing System

Pattern of Arrivals:

Customers may arrive in batches (such as the arrival of a family at a restaurant) or individually
(such as the arrival of a train at a platform). These customers may arrive at a service facility
either on scheduled time (by prior information) or on unscheduled time (without
information).The arrival process (or pattern) of customers to the service system is classified into
two categories: static and dynamic.

© Macmillan Publishers India Ltd

10
1997,2003,2007,2009
 The Structure of Queuing System

 In static arrival pattern, the control depends on the nature of arrival rate (random or
constant): In random (or unscheduled) arrivals the times are random variable, and hence
requires to understand the average and frequency distribution of the times. In both the cases,
the arrival process can be described either by the average arrival rate (average number of
arrivals per unit of time) or by the average inter-arrival time (average time between two
consecutive arrivals).
 The dynamic arrival process is controlled by both service facility and customers. The service
facility adjusts its capacity to match changes in the service intensity, by either varying the
staffing levels at different timings of service, varying service charges (such as telephone call
charges at different hours of the day or week) at different timings, or allowing entry with
appointments. The variation in the service intensity also affects the customer’s behaviour.
They either balk or renege from the service system when confronted with a long or slow
moving waiting line.

© Macmillan Publishers India Ltd

11
1997,2003,2007,2009
 The Structure of Queuing System

The Arrival Time Distribution

 Poisson distribution- a discrete probability distribution, helps in the analysis of probabilities

for the number of customers that may arrive in any specific interval to time, whereas
exponential distribution provides probabilities for times gap between two consecutive
arrivals.

Let n customers arrive during a time interval 0 to t. If l is the expected (or average)
number of arrivals per unit time, then expected number of arrivals during a time interval t
becomes lt. Poisson probability distribution function is given by

(lt )n e - , n = 0, 1, 2, …
P(x = n | Pn = lt) =
n!

© Macmillan Publishers India Ltd

12
1997,2003,2007,2009
 The Structure of Queuing System

Let T be the time between successive arrivals. Since a customer can arrive at any time,
T must be a continuous random variable. The probability of no arrival in the time
interval from 0 to t will be equal to the probability that T exceeds t, so
P(T  t )  P( x  0 | Pn  t )  e  t

The cumulative probability that the time T between two successive arrivals is t
or less is given by

P(T  t )  1  P(T  t )  1 e  t ; t0

This probability is also called the cumulative probability distribution function,

also the distribution of the random variable T is called exponential distribution.
Its probability distribution function (pdf) is given by
 e t ; t  0
f (t )©Macmillan
 Publishers India Ltd
13
 0 ;t 0
1997,2003,2007,2009
 The Structure of Queuing System

The inter-arrival time in many situations is approximated by an exponential distribution (also

called negative exponential distribution). The shape of the typical exponential probability
distribution function and cumulative exponential probability distribution function is:

© Macmillan Publishers
Exponential Probability Distribution India Ltd
Cumulative Exponential Probability Distribution 14
1997,2003,2007,2009
 The Structure of Queuing System

The mean of exponential distribution is the expected or average time E(T) between arrivals.
Thus,
with l arrivals per unit of time, E(T) = 1/l.
 Poisson distribution of arrivals with arrival rate l is equal to the negative exponential
distribution of inter-arrival times with average inter-arrival time 1/l. For example, if
l=4 customers per minute, then the inter-arrival time between any two customers is t =
0.40 minute (or 24 seconds) or less is given by

P(T £ 0.4) = 1 – e–4(0.4) = 1 – e–1.6 = 1 – 0.202 = 0.798

© Macmillan Publishers India Ltd

15
1997,2003,2007,2009
 The Structure of Queuing System

If customers arrive at an average rate of l=24 customers per hour, and a customer has already
arrived, then probability of a customer arriving in the next 5 minutes (i.e. t = 1/12 hr) is

P(T < 1/12) = 1 – e–24(1/12) = 1 – e–2 = 1 – 0.1353 = 0.8647

If l = 24 customers per hour, t = 1/12 hr and lt = 24×(1/12) = 2, the probability of n = 2

customers arriving within the next 10 minutes is:
e 24 (1/12 ) (24 / 12) 2
P(x = 2) = 0.27 
2!

© Macmillan Publishers India Ltd

16
1997,2003,2007,2009
 The Structure of Queuing System

Queuing Process

The queuing process refers to the number of queues, and their respective lengths. The number of
queues single, multiple or priority queues depend upon the layout of a service system. The
length (or size) of the queue depends upon the operational situation such as physical space, legal
restrictions, and attitude of the customers.

 If due to any reason(s), a service system is unable to accommodate more than the required
number of customers at a time , then such queues are called finite (or limited) source queue.
Examples of finite source queues are cinema halls, restaurants, etc.

© Macmillan Publishers India Ltd

17
1997,2003,2007,2009
 The Structure of Queuing System

 If a service system is able to accommodate any number of customers at a time, then it is called
an infinite (or unlimited) source queue. For example, in a sales department where the
customer orders are received, there is no restriction on the number of orders that can come in
so that a queue of any size can form.
 If arriving customers found long queue(s) in front of a service facility, they often do not enter
the service system even though additional waiting space is available. The queue length in such
cases depends upon the attitude of the customers. For example, when a motorist finds that
there
are many vehicles waiting at the petrol station, in most of the cases he does not stop at this
station and seeks service elsewhere.

© Macmillan Publishers India Ltd

18
1997,2003,2007,2009
 The Structure of Queuing System

Queue Discipline
The queue discipline is the order (or manner) in which customers from the queue are selected for
service.
 A static queue discipline is based on the individual customer’s status in the queue. Few of
such disciplines are:

• If customers are served in the order of their arrival, such discipline is known as: first-
come, first-served (FCFS) service discipline. Prepaid taxi queue at airports where a taxi
is engaged on a ‘first-come, first-served’ basis is an example of this discipline.
• Last-come, first-served (LCFS) is practised in most cargo handling situations where the
last item loaded is removed first because it reduces handling and transportation cost, the
last ones being easier to reach closer.

© Macmillan Publishers India Ltd

19
1997,2003,2007,2009
 The Structure of Queuing System

In the production process, items arrive at a workplace and are stacked one on top of the other.
Item on the top of the stack is taken first for processing which is the last one to have arrived for
service.
 A dynamic queue discipline is based on the individual customer attributes in the queue. Few
of such disciplines are:
• Service in Random Order (SIRO), where customers are selected for service at random
irrespective of their arrivals in the service system.
• Priority Service, where customers are grouped in priority classes on the basis of some
attributes such as service time or urgency, and FCFS rule is used within each class to
provide service. The payment of telephone or electricity bills by cheque or cash are
examples of this discipline.

© Macmillan Publishers India Ltd

20
1997,2003,2007,2009
 The Structure of Queuing System

• Pre-emptive priority (or emergency), where a customer is allowed to enter into the
service immediately after entering into the system even if a customer with lower
priority is already in service. That is, lower priority customer’s service is interrupted
(pre-empted) to start service for such a customer. This interrupted service is resumed
again after the priority customer is served.

• Non-pre-emptive priority where an important customer goes ahead in the queue, but
service is started immediately on completion of the current service.

© Macmillan Publishers India Ltd

21
1997,2003,2007,2009
 The Structure of Queuing System

Service Process (or Mechanism)

The service mechanism is concerned with the manner in which customers are serviced and leave
the service system. It is characterized by

 The arrangement (or capacity) of service facilities

 The distribution of service times

 Server’s behaviour

 Management policies

© Macmillan Publishers India Ltd

22
1997,2003,2007,2009
 The Structure of Queuing System

Arrangement of
Service Facilities
in Series

© Macmillan Publishers India Ltd

23
1997,2003,2007,2009
 The Structure of Queuing System

… service process

Arrangement of Service Facilities in Parallel

© Macmillan Publishers India Ltd

24
1997,2003,2007,2009
Single Queue, Multiple Service Facilities in Parallel and in
 The Structure of Queuing System

Service Time Distribution

 Average service rate is used to measure the service capacity of service facility in terms of
customers served per unit of time. If µ is the average service rate, then the expected number
of customers served during time interval 0 to t will be µt.
If the service time is exponentially distributed, then the service rate can be shown to be
Poisson distributed. Also, if time at the start of service, is considered as zero, then the
probability that service is not completed by time t is given by

P(x = 0 | Pn = µt) = e–µt

If the random variable T represents the service time, then the probability of service
completion within time t is given by

P(T £ t) = 1 – e–µt,  t ³ 0
 Average length of service time: A fluctuating service time may follow negative exponential
probability distribution, and is denoted by E(T) = 1/µ.

© Macmillan Publishers India Ltd

25
1997,2003,2007,2009
 Performance Measures of a Queuing System

Time-related measures (for the customers)

 The average (or expected) time an arriving customer has to wait in the queue ( Wq)
before being served.
 The average (or expected) time an arriving customer spends in the system (Ws)
including waiting and service.
Measures related to the number of customers

 Expected number of customers who are in the queue (queue length) for service, Lq.
 Expected number of customers who are in the system either waiting in the queue or
being serviced (Ls).

Measures related to cost (Service or waiting)

 The average cost needed to operate the system per unit of time?

 Servers (service centres) that are needed to achieve cost effectiveness?

© Macmillan Publishers India Ltd
26
1997,2003,2007,2009
 Performance Measures of a Queuing System

Measures related to value of time (both for customers and servers)

 Probability(Pw) that an arriving customer has to wait before being served. It is also
called blocking probability.
 Probability(Pr) that a server is busy at any particular point in time. It is the proportion
of the time that a server actually spends with the customer, i.e. the fraction of the time
a server is busy.
 Probability, Pn (n = 0, 1, . . .. )of n customers being in the queuing system when it is in
steady state condition,
 Probability(Pd )of service denial when an arriving customer cannot enter the system
because the queue is full,

© Macmillan Publishers India Ltd

27
1997,2003,2007,2009
 Performance Measures of a Queuing System

Transient-State and Steady-State

A service system progresses through a number of changes, since it started working and attains
stability after sometime. Before the start of the service operations it gets influenced by the initial
conditions specially by the number of customers in the system and the elapsed time. This period
of transition is termed as transient-state. However, as time passes, the system becomes
independent of the initial conditions and of the elapsed time and enters a steady- state condition.

Let Pn (t) be the probability of n customers in the system at time t. In the case of steady-state,
we have lim Pn (t ) = Pn
t® ¥

lim Pn¢( t ) = 0
t®¥

If arrival rate of customers is more than the service rate, then a steady-state cannot be reached
regardless of the length of the elapsed time.
© Macmillan Publishers India Ltd
28
1997,2003,2007,2009
 Performance Measures of a Queuing System

Relationships among Performance Measures

1. Expected number of customers in the queue and system

 ¥
     Ls   n Pn and Lq = S (n - s) Pn
n0 n= s

2. Expected number of customers in the system

Ls = Lq + Expected number of customers in service

l
                 = 
Lq +
m

3. Expected waiting time of the customer in the system Ws =

Average waiting time in the queue+ expected service time
1
= Wq + 
© Macmillan Publishers India Ltd
29
1997,2003,2007,2009
 Performance Measures of a Queuing System

4. Probability of being in the system (waiting and being served) longer than time t
               P(T > t)= e–(m + l)t  and  P(T £ t) = 1 – P(T > t)
where T = time spent in the system,
t = specified time period, e = 2.718

5. Probability of only waiting for service longer than time t

 –(m+l)t
             P(T > t) = e


6. Probability of exactly n customers in the system

n
    
Pn =  1   
    
© Macmillan Publishers India Ltd
30
1997,2003,2007,2009
 Performance Measures of a Queuing System

7. Probability that the number of customers in the system, n exceeds a

given number, r
r 1

P(n > r) =  


8.
1 1
 Ls =  Ws or Ws  Wq   Ls
 

1 1
 Lq = Ls – = Wq or Wq  Ws 

 Lq


© Macmillan Publishers India Ltd

31
1997,2003,2007,2009
 Classification of Queuing Models

Queuing Models
Different models in queuing theory are classified by using notations suggested by D.G. Kendall
(1953) in the form (a/b/c). A.M. Lee (1966) added the symbols d and c to the Kendall’s notation.
The standard format used to describe the main characteristics of queues is

          {(a/b/c) : (d/e)}

where a = arrivals distribution

b = service time (or departures) distribution
c = number of service channels (servers)
d = maximum number of customers allowed in the system
(in queue plus in service)
e = queue (or service)©discipline.
Macmillan Publishers India Ltd
32
1997,2003,2007,2009
 Classification of Queuing Models

Notations:

Certain descriptive notations used to replace notation a and b for the arrival and service times
distribution are:

M = Markovian (or exponential) inter-arrival time or service-time distribution.

D = Deterministic (or constant) inter-arrival time or service time.
G = General distribution of service time (departures), i.e. no assumption
is made about the type of distribution with mean and variance.
GI = General probability distribution – normal, uniform or any empirical
distribution, for inter-arrival time
Ek = Erlang-k distribution for inter-arrival or service time with parameter k

© Macmillan Publishers India Ltd

33
1997,2003,2007,2009
 Classification of Queuing Models

Single Server Queuing Model

Model I :  {(M/M/1) : (¥/FCFS)} Exponential Service – Unlimited Queue
Assumptions:

 Arrivals are described by Poisson probability distribution and come from an infinite
population.
 Single waiting line and each arrival waits to be served regardless of the length of the
queue (i.e. infinite capacity) and no balking or reneging.

 Queue discipline is ‘first-come, first-served’.

 Single server and service times follow exponential distribution.
 Customers arrival is independent but the arrival rate (average number of arrivals) does
not change over time.

 The average service rate is more than the average arrival rate.
© Macmillan Publishers India Ltd
34
1997,2003,2007,2009
 Classification of Queuing Models

The following events (possibilities) may occur during a small interval of time, Dt just before
time t.
 The system is in state n (number of customers) and no arrival and no departure, leaving
the total to n customers.

 Ths system is in state n+1 (number of customers) and no arrival and one departure,
reducing the total to n customers.

 The system is in state n–1 (number of customers) and one arrival and no departure,
bringing the total to n customers.

© Macmillan Publishers India Ltd

35
1997,2003,2007,2009
 Classification of Queuing Models

The probability of no customer and n customers in the system at time t are given
by

P0 = 1 – l = 1 – r
m
l n 1 l
Pn = ( )(
m
-
m
)= rn ( 1 - r ); r < 1 , n = 0 , 1 , 2 , . . .

Pw = Probability that an arriving customer has to wait

= 1 – P0 = l/m

© Macmillan Publishers India Ltd

36
1997,2003,2007,2009
 …Single Server Queuing Model

Performance measures for model I

1.  Expected number of customers in the system

 
n
Ls = n 
0
n Pn   n ( 1   ) 0 < r < 1
n0
,
  
 ; 
= 1   

2. Expected number of customers waiting in the queue (i.e. queue length)

  
Lq =  (n  1 ) Pn   nPn   Pn
n 1 n1 n 1
  
= L s  ( 1  P;0 )   1  P0 
  
2
= (  ) © Macmillan Publishers India Ltd
37
1997,2003,2007,2009
 …Single Server Queuing Model

3. Expected waiting time for a customer in the queue

¥ ¥
Wq = t × d f w ( t ) dt
= t × l(1 - r)e
- m (1 r) t
- dt
0 dt 0
¥
- m (1- r) t -r)t
- m (1
Integrating by parts, =Wq l (1 – r) te -
e
-m ( 1- r) m2 (1 - r )2 0
l 1  Lq
l 1-
= m = =
(m - l)2  (   ) 

4. Expected waiting time for a in the system (waiting and service)

1  1
Ws = W q   
  (   ) 

1 L
=  s © Macmillan Publishers India Ltd
38
  1997,2003,2007,2009
 …Single Server Queuing Model

5. The variance (fluctuation) of queue length

¥ 2 ¥ 2
Var (n) = å n P
n
- å nP
n ( )
n =1 n =1
 
= 2

(1   ) (  )2

6. Probability that the queue is non-empty

P(n > 1) = 1 – P0 – p1
2
(
=1 - 1 -
l
) -(1 - l )
)( = ( )
l
l
m m m m

© Macmillan Publishers India Ltd

39
1997,2003,2007,2009
 …Single Server Queuing Model

7. Probability that the number of customers, n in the system exceeds

a given number k.
¥ ¥ 
P (n ³ k) = =å Pk å = rk
(1(1- –r)r) rk  n  k
n = k nk
n =k

= ( l)
k
; P (n > k) = ( )
l
m
k +1

8.  Expected length of non-empty queue

Lq l2 /m ( m - l) m
L= = =
>
P ( n 1) l m 2 m-l
( / )

© Macmillan Publishers India Ltd

40
1997,2003,2007,2009
 …Single Server Queuing Model

9. Probability of an arrival during the service time when system

contains r customers
¥ (l t ) r e -l t
. me - t dt
m

r!
P (n = r) = 0

l m
= 
( l+m ) ( l +m )

© Macmillan Publishers India Ltd

41
1997,2003,2007,2009
 … Queuing Model

Example Arrivals at telephone booth are considered to be Poisson with an average time of 10
minutes between one arrival and the next. The length of phone call is assumed to be distributed
exponentially, with mean 3 minutes.
 What is the probability that a person arriving at the booth will have to
wait?

 The telephone department will install a second booth when convinced

that an arrival would expect waiting for at least 3 minutes for a phone
call. By how much should the flow of arrivals increase in order to justify
a second booth?

 What is the average length of the queue that forms from time to time?

 What is the probability that it will take him more than 10 minutes
© Macmillan Publishers India Ltd
altogether to wait for the phone and complete his call?
1997,2003,2007,2009
42
 … Queuing Model

Solution: l = 1/10 = 0.10 person per minute and m=1/3=0.33 person per minute

a) Probability that a person has to wait at the booth.

P (n  0)  1  P=
0   / 010
. / 0.33  0.3

b) The installation of second booth will be justified only if the arrival rate is
more than the waiting time. Then expected waiting time in the queue will be

Wq =  (    )

   016
.
 3 = 0or  
.33 (0.33   )

Hence, the increase in the arrival rate is 0.16 – 0.10 = 0.06 arrivals per minute.
© Macmillan Publishers India Ltd
43
1997,2003,2007,2009
 … Queuing Model

c) Average length of non-empty queue

 0.33
L   2 customers (approx.)
 0.23
d) Probability of waiting for 10 minutes or more is given by

¥ l (m - l )t
P ( t ³ 10) = (m - l ) e - dt
10 m
¥ - 0 . 23t ¥
= (0.3) (0.23) e
- 0.23t
dt = 0.069
e = 0.03
10
- 0.23
10

This shows that 3 per cent of the arrivals on an average will have to wait for
10 minutes or more before they can use the phone.
© Macmillan Publishers India Ltd
44
1997,2003,2007,2009
 … Queuing Model

Example A road transport company has one reservation clerk on duty at a time.

He handles information of bus schedules and makes reservations. Customers

arrive at a rate of 8 per hour and the clerk can service 12 customers on an average

per hour. After stating your assumptions, answer the following:

 What is the average number of customers waiting for the service of the clerk?
 What is the average time a customer has to wait before getting service?
 The management is contemplating to install a computer system to handle the
information and reservations. This is expected to reduce the service time from 5 to 3
minutes. The additional cost of having the new system works out to Rs 50 per day. If
the cost of goodwill of having to wait is estimated to be 12 paise per minute spent
waiting before being served, should the company install the computer system? Assume
© Macmillan Publishers India Ltd
8 hours working day. 1997,2003,2007,2009
45
 … Queuing Model

Solution:  Given that l = 8 per hour; m = 12 per hour

a) The average number of customers waiting for the service in the system
 8
Ls    2 customers
 12  8
b) The average time spent by a customer in the system
1 1 1
Ws    hour  15 minutes.
 12  8 4
The average waiting time in the queue for a customer
 8 1
Wq    hour  10 minutes.
 (   ) 12 (12  8) 6

© Macmillan Publishers India Ltd

46
1997,2003,2007,2009
 … Queuing Model

c) Comparing additional cost of Rs 50 per day with the goodwill cost of

customers with existing system before installing computer system.
The average cost for waiting time in the system,
Ws = 15 minutes is 0.12 × 15 = Rs 1.80.

Since 8 customers arrived in 1 hour or in 8 hours 64 (=8×8) customers

request service at a total goodwill cost of 1.80 × 64 = Rs 115.20.
The installation of computer system, will increase service rate up to l = 20
customers per hour (3 customers per minute).The average time spent by a
customer in the system would be
1 1 1
Ws    hour  5 minutes
   20  8 12

and the average daily customer queuing (or goodwill) cost is reduced to: 64 ×
© Macmillan Publishers India Ltd
(0.12 × 5) = Rs 38.40. 1997,2003,2007,2009
47
 … Queuing Model

Operational cost of computer is Rs 50 per day. So, the average total daily cost

would be

TC = Computer cost + Goodwill cost = 50 + 38.40 = Rs 88.40

This cost is less than the existing goodwill loss cost and net saving is =
(115.20 – 88.40) = Rs 26.80. Hence, company may install a computer.

© Macmillan Publishers India Ltd

48
1997,2003,2007,2009
 … Queuing Model

Model III:{(M/M/1):(N/FCFS)}Exponential Service–Finite (or Limited) Queue

Since no more than N customers can be accommodated at any time in the system due to certain
reasons. Then any additional customer arriving can not enter into the system and is lost.
Emergency room in a hospital; one-man barber shop with certain number of chairs for waiting
customers, etc. are the examples of finite queue.
The steady-state probability distribution of n customers in the system is:

1- r
P0 = ; r ¹ 1 and r < 1; r = l / m
+
1- rN 1

1-r
( +
1 - rN 1
)
r n ; n £ N ; r ¹ 1( l ¹ m )
Pn =
1
; rPublishers
© Macmillan = 1( l =India
m )Ltd
+
N 1 1997,2003,2007,2009
49
 … Queuing Model

In such a case the steady-state solution exists even for r > 1. Because limited
capacity of the system controls the arrivals by the queue length (=N–1) not
by the relative rates (l ) of arrival and departure (µ). If l < µ and N  ,
then Pn = ( 1 – l/µ) (l/m)n , which is same as in Model I.

© Macmillan Publishers India Ltd

50
1997,2003,2007,2009
 … Queuing Model

Performance measures for model III

1. Expected number of customers in the system

N N
å = å 1- r rn
Ls = nP n
n N +1
n = 1 n = 1 1- r
+
r ( N + 1) r N 1
- ; r ¹ 1(l ¹ m)
1- r 1 - r N +1
= N
; r = 1(l = m)
2

2. Expected queue length or expected number of customers waiting in the

system
  ( 1  PN )
Ls  = Ls 
Lq =  
© Macmillan Publishers India Ltd
51
1997,2003,2007,2009
 … Queuing Model

3.  Expected waiting time of a customer in the system (waiting + service)

Ls
Ws =  (1 P ) N

4.  Expected waiting time of a customer in the queue

1 Lq
Wq = Ws  or
  ( 1  PN )

5. Fraction of potential customers lost ( = fraction of time system is full)

PN = P0 rN

Effective arrival rate, leff = l(1 – PN)

Effective traffic intensity, reff = le / µ.

© Macmillan Publishers India Ltd
52
1997,2003,2007,2009
 … Queuing Model

Example Consider a single server queuing system with Poisson input, exponential service
times. Suppose the mean arrival rate is 3 calling units per hour, the expected service time is
0.25 hour and the maximum permissible calling units in the system is two. Derive the steady-
state probability distribution of the number of calling units in the system, and then calculate
the expected number in the system.

Solution  Given l = 3 units per hour; µ = 4 units per hour, and N = 2

Then traffic intensity,

   /  3/4  0 . 75

© Macmillan Publishers India Ltd

53
1997,2003,2007,2009
 … Queuing Model

The steady-state probability distribution of n customers in the system is:

(1   )  n
Pn = ; 1
N 1
1 
( 1  0.75) ( 0.75) n
=  ( 0.43) ( 0.75) n
1  ( 0.75 ) 2  1
(1   ) 1  0.75 0.25
and P0 =    0.431
1   N 1 1  ( 0.75) 2  1 1  (0.75) 3

The expected number of calling units in the system is given by

N 2
 nPn   n ( 0.43) ( 0.75) n
Ls = n1 n1
2
2
0.43  n ( 0.75) n  0.43 {( 0.75)  2 ( 0 .75) }  0.81
= n 1

© Macmillan Publishers India Ltd

54
1997,2003,2007,2009
© Macmillan Publishers India Ltd
55
1997,2003,2007,2009