QUEUING THEORY

Part I
MGTE 31243-Optimization Methods in Management Science II

By
Ms. Erandika Gamage
University of Kelaniya
 LEARNING OUTCOMES
o Describe the key elements and underlying mathematical concepts of analytical queuing models
o Explain and compute the operating characteristics associated with the single server queuing
model
o Apply the operating characteristic formulas for multiple server queuing model
o Understand the behavior of serial and cyclic queueing network models
o Explain economic trade offs associated with designing and managing queuing systems
o Explain the use of Excel to compute the operating characteristics associated with queuing
models
o Explain the psychology of waiting for designing and managing queues

By Ms. Erandika Gamage

 SUMMARY
o What is Queueing Theory?
o Queueing Situation
o Goals of Queuing Analysis
o Applications of Queuing Theory
o Components of a Queueing System
o Characteristics in a Queuing System
o Equilibrium/ Steady State Condition
o Kendall’s Notation
o Queueing Model
o Seven Types of Queueing Models
o Use of Queuing Add-in to solve problems
o Suggestions for Managing Queues
By Ms. Erandika Gamage
 WHAT IS QUEUING THEORY ?
o What is a Queue?
o Queue is a linear arrangement of items waiting to be served
o Queuing Theory is the Mathematical Study of Waiting Lines/Queues

Ex:  Waiting in line at a bank for a teller

 Waiting for a customer service representative to answer a call
 Waiting for a train/bus to come
 Waiting for a computer to perform a task or respond
 Waiting for an automated car wash to clean a line of cars

By Ms. Erandika Gamage

 WHAT IS QUEUING THEORY ?
• People - Bus queue, Cinema queue
• Items - Vehicle queue, Queue of applications
• Events - Queue of telephone calls, Queue of births

o Queues/Waiting Lines are formed when people or items come faster than they can be

served (due to limited resources for providing a service and resources and demand
mismatched)

By Ms. Erandika Gamage

By Ms. Erandika Gamage
 QUEUING SITUATION
o At its core, a queuing situation involves two parts.
o Someone or something that receive the service—People, Items, Events
o Someone or something that provides the service—Server

Ex:  At a bank - the customers are people seeking to deposit or withdraw money, and the
servers are the bank tellers.
 When looking at the queuing situation of a printer, the customers are the requests that
have been sent to the printer, and the server is the printer.

By Ms. Erandika Gamage

QUESTIONS ARISE IN QUEUING SITUATION
o Those that receive service would like to know
1. How long the queue will be ?
2. How long will I have to wait in the queue ?

o Those who provides service would like to know

1. How many hours the server idle ?
2. Should we have two servers ?

By Ms. Erandika Gamage

 GOAL OF QUEUING ANALYSIS
o To find the best level of service that a firm should provide

Ex:  Supermarket
- Should decide how many cash register checkout position should be opened

 Petrol station
- How many pumps should be opened and how many attendants should be on duty

 Bank
- Should decide how many teller windows to keep open to serve customers during various
hours of the day

By Ms. Erandika Gamage

APPLICATIONS OF QUEUING THEORY
o Telecommunications
o Traffic Control
o Determining the sequence of computer operations
o Predicting computer performance
o Healthcare (hospital bed assignments)
o Airport traffic, airline ticket sales
o Layout of manufacturing system

By Ms. Erandika Gamage

COMPONENTS OF A QUEUING SYSTEM

Units in Queuing System

Server
Unit currently
Population Arrival Units in Queue/Waiting Line Departure
being served

By Ms. Erandika Gamage

COMPONENTS IN A QUEUING SYSTEM
System Customers Servers
Reception Desk People Receptionist

Hospital Patients Nurses

Airport Airplanes Runway
Road network Cars Traffic light
Grocery Shoppers Checkout
Station Computer Jobs CPU, disk, CD

By Ms. Erandika Gamage

CHARACTERISTICS IN A QUEUING SYSTEM
1) Pattern of arrivals
2) Pattern of service provision
3) Number of servers
4) Queue discipline
5) Size of waiting room
6) Size of population

By Ms. Erandika Gamage

CHARACTERISTICS IN A QUEUING SYSTEM
Units in Queuing System
Server

Unit currently
Population Arrival Units in Queue/Waiting Line Departure
being served

o Arrival Characteristics o Waiting Line Characteristics o Service Characteristics

• Pattern of arrival • Size of waiting room • Pattern of service provision
• Size of population • Queue discipline • Number of servers

By Ms. Erandika Gamage

 1) PATTERN OF ARRIVALS
Pattern of
Arrivals

I. Poisson II. Uniform III. Earlang

Distribution Distribution Distribution

By Ms. Erandika Gamage

 1) PATTERN OF ARRIVALS
I. Poisson Distribution
Arrivals are considered random when they are independent of one another and their occurrence
can’t be predicted exactly
Probability of “x” occurrences/arrivals within a given unit of time or space “t” is given by

𝜆 – Average rate of arrivals

Ex: 20 customers per hour

10 lorries per day
5 items per minute

By Ms. Erandika Gamage

 1) PATTERN OF ARRIVALS
Ex:
The average number of telephone calls received per hour is 2.What is the probability of
receiving 3 telephone calls within the next hour?
𝜆 = 2 per hour x=3 t = 1 hour
=

By Ms. Erandika Gamage

2) PATTERN OF SERVICE PROVISION

Pattern of
Service Provision

I. Poisson II. Uniform III. Earlang

Distribution Distribution Distribution

By Ms. Erandika Gamage

1) PATTERN OF SERVICE PROVISION
I. Poisson Distribution
In many cases it can be assumed that random service times are described by the negative
exponential distribution ( )
Probability of “x” service completions within a given unit of time or space “t” is given by

µ – Average rate of service provision

Ex: 12 customers served per hour

10 vehicles serviced per day
5 items assembled per minute

By Ms. Erandika Gamage

UNIT OF MEASUREMENT FOR ARRIVALS & SERVICE COMPLETIONS
o In a queuing system, the pattern of service completion can be expressed in 2 ways

I. Service rate (average rate of service provision)

Ex: 2 vehicles serviced per hour

II. Service time ( time taken to serve one unit)

Ex: average time taken to service a vehicle is 30 min

“ The unit of measurement for pattern of arrival and pattern of service completion should be equal ”

Ex: A TV repairman finds that the average time spent on his jobs is 30 min per TV set and is negative
exponential. TV sets arrive in a Poisson fashion at the rate of 10 per eight hour day.
Average rate of arrival = 10 per eight hour day
Average rate of service completion = 30 min per TV set
= ½ hours per TV set * 2 *8
By Ms. Erandika Gamage
= 16 per eight hour day
 3) NUMBER OF SERVERS
o Single Server Queue o Multi Server Queue

• Single Server Single Queue • Multi Server Single Queue

Server 1

Server Server 2

Server 3

• Single Server Multiple Queues • Multi Server Multiple Queues

Server 1

Server Server 2

Server 3
By Ms. Erandika Gamage
 3) NUMBER OF SERVERS
o Single Server Queue o Multi Server Queue

I. Single Server Single Queue I. Multi Server Single Queue

Ex: • Students arriving at a library counter  Ex: • Booking at a service station
• Family doctor’s office

II. Single Server Multiple Queues II. Multi Server Multiple Queues
Ex: Ex: • Different cash counters in an electricity office
• Different boarding pass encounters at an airport

By Ms. Erandika Gamage

 3) NUMBER OF SERVERS
o Single Server Queue o Multi Server Queue

• Single Server Single Queue (1) • Multi Server Single Queue (S)

Server 1

Server Server 2

Server 3

By Ms. Erandika Gamage

 4) QUEUE DISCIPLINE
o This is the rule in which units in the queue are being selected for service

I. First Come First Served (FCFS) / First in First Out (FIFO)

Ex: Payment counter at shops

II. Last Come First Served (LCFS)/ Last in First Out (LIFO)
Ex: Elevator

III. Service in Random Order (SIRO)

Ex: Drawing tickets out of a pool of tickets for service

IV. Priority Service

Ex: Hospital Emergency Room (patients who are critically injured will move ahead in treatment)
By Ms. Erandika Gamage
 5) SIZE OF WAITING ROOM
o Maximum allowable size/number of units in the queue

I. Size of waiting room is infinity (∞)

Waiting Room Server

Ex: Tollbooth serving arriving vehicles

II. Limited size of waiting room (N)

Waiting Room Server

Ex: A small restaurant has 10 tables and can serve no more than 50 customers

By Ms. Erandika Gamage

 6) SIZE OF POPULATION
o This is the size of population eligible to receive the service. (Total number of eligible units
outside the queuing system)

I. Infinite Population (∞)

Ex: All people of a city or state (and others) could be the potential customers at a milk parlor.

II. Finite Population (N)

Ex: Customers at University Base Canteen
• In this event the rate of arrival is considered to be proportional to the size of population

By Ms. Erandika Gamage

 KENDALL’S NOTATION
o Standard system of notation used to describe and classify the queueing model that a
queueing system corresponds to.
o The notation was first suggested by D. G. Kendall in 1953

Pattern Pattern of Number Size of

Queue Size of
of Service of Waiting
Discipline Population
Arrivals Provision Servers Room

By Ms. Erandika Gamage

 QUEUING MODEL
Ex: Think of an ATM
It can serve one customer at a time; in a first-in-first-out order; with a randomly-distributed
arrival process and service distribution time; unlimited queue capacity; and unlimited number of
possible customers.

(M / M / 1) : ( FIFO / ∞ / ∞ )

M –Markovian, Poisson distribution for arrival

M – Markovian, Negative Exponential distribution for service provision

By Ms. Erandika Gamage

WHAT HAPPENS WITHIN A QUEUING SYSTEM?

The system starts with empty and idle condition in the beginning (a supermarket just open early
in morning and no customers yet)
Then it gradually go in to one or more peak time where the number of customers in the system
reach the highest level and gradually reduces.
At the end of the service hour (in night before closing the supermarket) either the arrival of the
customers are cutoff or there is really no more customers.
The assumption of steady state in the queuing theory doesn’t represent the reality
But in queuing theory we assume that a queue will acquire steady state condition(system doesn’t
change anymore) after few time, which means queue has reached equilibrium.
Queue has reached equilibrium means the probability that the system is in a given state is not
time dependent

By Ms. Erandika Gamage

 EQUILIBRIUM/STEADY STATE

By Ms. Erandika Gamage

 TYPES OF QUEUING MODELS

1. Model 1- (M / M / 1) : ( FIFO / ∞ / ∞ ) - Single server infinite waiting room

2. Model 2 - (M / M / 1) : ( FIFO / L / ∞ ) - Single server limited size of waiting room
3. Model 3 - (M / M / S) : ( FIFO / ∞ / ∞ ) - Multi server infinite waiting room
4. Model 4 - (M / M / S) : ( FIFO / L / ∞ ) - Multi server limited size waiting room
5. Model 5 - (M / M / 1) : ( FIFO / ∞ / N ) - Single server finite population
6. Model 6 - (M / M / S) : ( FIFO / ∞ / N ) - Multi server finite population
7. Model 7 - Queuing Networks

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Single Server Infinite Waiting Room

Avg number of units in queuing system ( Ls )

Avg number of units in queue ( Lq ) Rate of service

provision (µ)
Rate of arrival ( 𝜆 )

Waiting Room Server

Average time spent by unit in queue ( Wq )

Average time spent by unit in queuing system ( Ws )

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Probability of “n” units in queuing system
P(n) = P(n - 1)
P(n) = P(0)
P(n) =

Variables :
λ = Rate of arrival of units
μ = Rate of service provision
θ=λ/μ

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
P(n) = θn (1 – θ) (*)
P (n) is a measure of probability and cannot be negative
P (n) = θn (1 – θ) cannot be negative
(1- θ) > 0 (since if θ >1 then θn (1 – θ) will be negative)
θ<1
λ/μ<1
λ<μ
Condition for equilibrium is that,
λ<μ

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Probability that the queuing system is empty
 P(n) = θn (1 – θ)
If queuing system is empty then n = 0
Hence,
P(0) = θ0 (1 – θ)
P(0) = (1 – θ)
Probability that the queuing system is empty = (1 – θ)

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Probability that the server is idle
Server will idle only if the queuing system is empty.
Hence,
Probability that the server is idle = (1 – θ)

Number of hours server idle per day

Suppose a working day has H hours,
Number of hours server idle per day = H (1 – θ)

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Average number of units in queuing system ( Ls )
Ls =

Average number of units in queue ( Lq )

Lq =

By Ms. Erandika Gamage

 LITTLE’S LAW
o First proven by mathematician John Little in 1961

o Long-term average number of units (L) in a queuing system is equal to the long-term average
rate of arrival (λ) multiplied by the average time that a unit spends in the system (W)
L = λW

Note:
• Little’s law assumes that the system is in “equilibrium”

• Valid for any queuing model

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Average time spent by unit in queuing system ( Ws )
According to Little’s Law, L = λW
Hence,
Ls = λ Ws
Ws =

Average time spent by unit in queue ( Wq )

According to Little’s Law, L = λW
Hence,
Lq = λ W q
Ws =

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Example :
A TV repairman finds that the average time spent on his jobs is 30 min per TV set and is
negative exponential. TV sets arrive in a Poisson fashion at the rate of 10 per eight hour day.
a) What is the repairman idle time each day?
b) How many jobs are ahead of the set just taken for repairs?
c) How long on the average must a TV set be kept with the repairman?

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Example : Answers
Rate of arrival of TV sets (𝜆) = 10 (per eight hour day)
Average time take to repair one set = 30 min = 1 / 2 hours
Number of sets repaired per day = 8 / (1/2) = 16 (per eight hour day)
Rate of service provision (µ) = 16 (per eight hour day) 
λ= 10 μ= 16 θ= λ / μ = 10/16 = 5/8
Number of working hours per day (H) = 8

a) Repairman idle time per day = H (1-θ) = 8 (1-5 /8) = 3 hrs

b) Number of sets ahead of the set just taken for repair L Q,

LQ = θ2/ (1-θ) = (5/8)2/ (1-5/8) = 1
By Ms. Erandika Gamage
MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Example : Answers
c) How long must a TV set be kept with the repairman relates to W S,
WS = LS /λ
But LS = θ / (1-θ)
LS = (5/8) / [1- (5/8)] = 5 /3
WS = ( 5 /3 ) * ( 1 /10 ) = 1 /6 (eight hour day)
WS = 1/6 * 8 = 1 1/3 hrs

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Exercise 1:
At a photocopying shop there is only one photocopying machine and customer arrive in a
Poisson fashion at a rate of 14 per hour. The average service time is 4 min and is negative
exponential.
a) What is the average length of queue?
b) What is the average number in the queuing system?
c) Why the difference between the above two estimates not equal to one?
d) How long will a customer have to wait at the photocopying shop?

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Exercise 1: Answers
Rate of arrival of customers (λ) = 14 per hour
Time taken to serve one customer = 4 min
Number of customers served in one hour = 60/4 = 15
That is rate of service provision (µ) = 15 per hour
λ = 14 μ = 15 θ = λ / μ = 14/15

a) Average length of queue LQ,

LQ = θ2 / (1-θ)
LQ = (14/15)2 / [ 1- (14/15)] = (14/15) * (14/15)* 15 = 196/15 =13.06

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Exercise 1: Answers
b) Average number in the queuing system LS ,
LS = θ/(1-θ)
LS =(14/15) / [1- (14/15)] = (14/15) * 15 = 14

c) The difference between LS and LQ need not be equal to 1 because both LS and LQ are
estimates or averages.

d) How long a customer has to wait at the shop is WS,

WS = LS /λ
WS = 14/14 = 1 hour

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Exercise 2:
At a health centre patients arrive in a poisson fashion at the rate of 9 per hour. There is only one
doctor whose service time is negative exponential. The mean service time of doctor is 6 min.
a) What is the probability that the doctor is idle?
b) What is the average number of patients at the health centre?
c) How long will a patient have to wait at the health centre?
d) The doctor does not like to see patients standing and has instructed that probability that a
patient is standing be kept below 5%. How many chairs are required to keep to the doctor’s
request?

By Ms. Erandika Gamage

MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Exercise 2: Answers
Rate of arrival of patients (λ) = 9 per hour
Time taken for one patient = 6 min
Number of patients treated in one hour = 60/6 = 10
λ=9 μ = 10 θ = λ / μ = 9/10

a) Probability that doctor is idle P(0),

P(0) = (1-θ) = (1- 9/10) = 1/10

b) Average number at health center LS,

LS = θ/(1-θ) = (9/10) / [1- (9/10)] = 9

c) How long a patient has to wait WS,

By Ms. Erandika Gamage
WS = LS / λ = 9 /9 = 1hour
MODEL 1 - (M / M / 1) : ( FIFO / ∞ / ∞ )
Exercise 2: Answers
d) Let number of chairs at health center = x
Then a patient will stand only if the number of patients at health center is more than x
Probability that patient is standing = P(x+1) + P(x+2) + ….+ ∞
= θ(x+1) (1-θ) + θ(x+2) (1-θ) + ……..+ ∞
= θ(x+1) (1-θ) [ 1+ θ + θ2 + …… + ∞]
= θ(x+1) (1-θ) * 1/ (1-θ)
= θ(x+1)
Therefore θ(x+1) = 5%
(9/10)(x+1) = 0.05
(x+1) log (9/10) = log (0.05)
(x+1) ≈ log (0.05) / log (9/10) = 29
x ≈ 28
Hence the number of chairs required is 28
By Ms. Erandika Gamage
MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Single Server Limited Size of Waiting Room

Avg number of units in queuing system ( Ls )

Avg number of units in queue ( Lq ) Rate of service

provision (µ)
Rate of arrival ( 𝜆 )

Waiting Room Server

Average time spent by unit in queue ( Wq )

Average time spent by unit in queuing system ( Ws )

By Ms. Erandika Gamage

MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Probability of “n” units in queuing system
Probability of “n” units in queuing system

n <= L n>L
P(n) = P(n - 1) P(n) = 0
P(n) = P(0)
P (n) =

Probability that queuing system is empty

P (0) = Variables :
λ = Rate of arrival of units
μ = Rate of service provision
θ=λ/μ
By Ms. Erandika Gamage
L = Size of waiting room
MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Probability that the server is idle
Server will idle only if the queuing system is empty.
Hence,
Probability that the server is idle =

Number of hours server idle per day

Suppose a working day has H hours,
Number of hours server idle per day = H*

By Ms. Erandika Gamage

MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Average number of units in queuing system ( Ls )
LS =

Average number of units in queue ( Lq )

Lq = LS - [1- P(0)]

By Ms. Erandika Gamage

MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Average time spent by unit in queuing system ( Ws )
Ws =

Average time spent by unit in queue ( W q )

Wq =

Variables :
λ = Rate of arrival of units
μ = Rate of service provision
θ=λ/μ
L = Size of waiting room
H = Number of working hours per day

By Ms. Erandika Gamage

MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Example 1:
Sid Das Brick Distributors currently employs 1 worker whose job is to load bricks on outgoing
company trucks. An average of 24 trucks per eight hour working day arrive at the loading
platform, according to a Poisson distribution. The worker loads them at a rate of 4 trucks per
hour, following approximately the exponential distribution in his service times. But the loading
area cannot accommodate more than three trucks either being loaded or waiting.
a) What is the probability that there will be at least two trucks either being loaded or waiting?
b) How many hours does the loader has to provide service within the day?
c) How many hours a truck has to be waited until loading being completed?
d) A truck driver working for Sid Das earn an average of $10 per hour. Brick loader receive
about $6 per hour. Truck drivers waiting in the queue or at the loading platform are drawing
a salary but are productively idle and unable to generate revenue during that time. What
would be the total hourly cost to the firm?

By Ms. Erandika Gamage

MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Example 1: Answers
Number of servers = 1
Number of trucks arrive per day = 24
Rate of arrival of trucks (𝜆) = 24/8 = 3 per hour
Rate of service provision (µ) = 4 per hour
Size of waiting room (L) = 3
θ = λ/μ = ¾

a) The summation of the total probabilities = P(0) + P(1) + P(2) + P(3) = 1

Probability that there will be at least two trucks in the system = P(2) + P(3)
= 1 – [P(0) + P(1)]

By Ms. Erandika Gamage

MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Example 1: Answers
P (n) =
P (0) = = = 0.3657
 
P (1) = = = 0.2743
3
5
8

Probability that there will be at least two trucks in the system = 1 – (0.3657 + 0.2743)
= 0.36
b) Working hours per day (H) = 8 hours
Probability that the server is idle P(0),
P (0) = 0.3657
Number of hours server idle per day = H*P(0) = 8*0.3657 = 2.9256 hours
Number of hours server doesn’t provide service per day = 2.9256 hours
Number of hours the loader has to provide service within the day = 8 - 2.9256
= 5.0744 hours

By Ms. Erandika Gamage

MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Example 1: Answers
c) hours a truck has to be waited until loading being completed WS,
Ws =
LS =
LS = = 1.1486 
Ws = = = 0.3829 hours

d) Idle time of a truck driver means average waiting time of a truck driver within system W S,
Ws = 0.3829 hours
Number of trucks arrive per day = 24
Total idle hours of all the trucks per day = 0.3829 *24 = 9.1896 hours
Truck driver idle time cost = 9.1896 * 10 = $ 91.896
Number of hours brick loader idle per day = 2.9256 hours
Brick loader idle time cost = 6 * 2.9256 = $ 17.5536
Total cost to the firm per day = $ 17.5536 + $ 91.896 = $ 109.45

By Ms. Erandika Gamage

MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Exercise 1:
At a barber shop customers arrive in a Poisson fashion at the rate of 14 per hour. There is only one
barber who takes 4mins per hair cut. There are five chairs for waiting customers. When a customer
arrives if he finds that all the waiting chairs are occupied he proceeds to another barber shop.

a) What is the probability that the barber is idle?

b) What is the probability that there are three customers at the barber shop?
c) What is the probability that a customer that arrives turns back and proceeds to another barber
shop?
d) How many customers on the average will he lose on a eight hour working day on account of
having only five chairs.
e) If the cost of a hair cut is Rs.50 then on the average how much more would he earn per day if he
has five more chairs.

By Ms. Erandika Gamage

MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Exercise 1: Answers
Rate of arrival of customers (λ) = 14 per hour
Time taken for one hair cut = 4 mins
Number of haircuts completed in one hour = 60 / 4 = 15
Rate of service provision (μ) =15 per hour
Size of waiting room (L)= (5+1) = 6
θ = λ / μ =14 /15
(a) Probability that barber is idle =
=
= (0.06669)/ (0.38304)
= 0.174

By Ms. Erandika Gamage

MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Exercise 1: Answers
(b) P (n) =
If there are three customers then n=3
P (3) =
=
= (14/15)3 * 0.174
=0.14146

By Ms. Erandika Gamage

MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Exercise 1: Answers
(c) A customer arrive will leave and proceed to another barber shop only if the barber shop is full. That is
only if there are six customers in the shop.
So probability that a customer who arrives will leave for another barber shop is P (6)
P (6) =
=
= (14/15)6 * 0.174
= 0.115
(d) Rate of arrival of customers (λ) = 14 per hour
Number of customers that arrive on an 8 hour working day = 8*14 = 112
Probability of losing a customer = 0.115
Therefore number of customers lost per day = 112*0.115 = 12.88
By Ms. Erandika Gamage
MODEL 2 - (M / M / 1) : ( FIFO / L / ∞ )
Exercise 1: Answers
(e) If there are five more chairs then all together there will be 10 waiting chairs and it will be a limited waiting
room size queue with size of waiting room as 11.
A customer will leave for another barber shop if the number of customers at the shop is 11.

Prob. that a customer arriving will leave for another shop =

=
= 0.4681705 * (0.0666666) / (0.5630409)
= 0.0554334
Number of customers lost per day when (L =11) = 0.0554334 * 112
= 6.21
Number of customers lost per day when (L =6) = 112*0.115
= 12.88
Number of customers saved per day = 12.88 – 6.21 = 6.67
The increasing profit = Rs.6.67*50 = Rs. 333.00

By Ms. Erandika Gamage

 Click icon to add picture

Next week :
3. Model 3 - (M / M / S) : ( FIFO / ∞ / ∞ ) -
Multi server infinite waiting room
END OF
4. Model 4 - (M / M / S) : ( FIFO / L / ∞ ) -
PART I
Multi server limited size waiting room

ANY
QUESTIONS ?