0% found this document useful (0 votes)
37 views

Queuing Theory: RK Jana

The document provides an overview of queuing theory, including the key elements of a queuing system such as arrivals, queues, and service mechanisms. It discusses operating characteristics like steady and transient states. Common distributions for arrivals and service times are covered, along with queue disciplines and models. Notations and applications to areas like telecommunications are also summarized.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
37 views

Queuing Theory: RK Jana

The document provides an overview of queuing theory, including the key elements of a queuing system such as arrivals, queues, and service mechanisms. It discusses operating characteristics like steady and transient states. Common distributions for arrivals and service times are covered, along with queue disciplines and models. Notations and applications to areas like telecommunications are also summarized.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 31

Queuing Theory

RK Jana

1
Outline
• Elements of a queuing system
• Operating characteristics of a queuing system
• Steady and transient states of a queuing system
• Distribution of arrivals and service times
• Queue discipline
• Single channel waiting line model with Poisson arrivals and
Exponential service times
• Multiple channel waiting line model with Poisson Arrivals
and Exponential service times
• Software for analysis of queuing systems

2
Elements of a Queuing System

A flow of customers coming towards the service facility forms a


queue on account of lack of capacity to serve them all at a time.

Some Examples:
 Persons waiting at doctor’s clinic
 Persons waiting at railway booking office
 Machines waiting to be repaired
 Ships waiting in the harbor to be unloaded
 Airplanes take off, landing
Customers may be: persons, machines, vehicles, parts etc 3
Continued…

Customers

Input Process Queue Service


Mechanism

Queuing system: Customers arriving for service,


waiting for service if it is not immediate, leaving the
system after being served.

The theoretical study of waiting lines, expressed


in mathematical terms. 4
Input Process
 Source (population size)
 Finite
 Infinite

 Arrival behavior
 One by one

 Batch
 Fixed size
 Variable size

 Change of arrival pattern


 Stationary
 Non-stationary/Transient (time dependent)
5
Queue Discipline

 Queue size
 Finite

 Infinite

 Queue discipline
 First come, first served (FCFS)

 Last come, first served (LCFS)

 Random selection for service (RSS)

 Priority queue

 Preemptive
 non-preemptive

6
Service Mechanism

 Bulk service (computer with parallel processing, bus


service)
 State dependent (service depends on number of waiting
customers. Example: internet access)
 State independent
 Stationary/non-stationary

7
Number of Service Channels

 Single channel
 Parallel channels (provides identical service)
 Series (customers go through a number of services, public
offices, manufacturing process)

8
Capacity of the System

 Finite source queue (finite capacity of waiting room, restriction on


queue length)
 Infinite (no restriction on queue length)

9
Applications of Queuing Theory

 Telecommunications
 Traffic control
 Determining the sequence of computer operations
 Predicting computer performance
 Health services (eg. control of hospital bed
assignments)
 Airport traffic, airline ticket sales
 Layout of manufacturing systems

10
QT in Performance Measurement

 Analyze waiting time distribution


 To know average waiting time of a customer
 To know queue length distribution
 Calculate current work backlog
 Measurement of the idle time of server
 Measurement of the busy time of server
 System utilization

11
Questions in Queuing Systems

1. What is the number of people in the system?


2. What is the number of people in the queue?
3. What is the waiting time of a customer in the system?
4. What is the waiting time of a customer in the queue?

12
Notations

n: # customers in the system (in queue & in service)


Pn(t): Transient state probability of exactly n customers in the
system at time t (it is assumed that the system starts its
operation at time zero)
Pn: Steady state probability of exactly n customers in the
system
E(s): Expected number of customers in the system
E(q): Expected number of customers in the queue
= E(s) – Number of customers being served

13
Notations: Continued…

λn : Expected number of arrivals per unit time (mean


arrival rate) of customers when n customers are
present in the system
µn : Expected number of customers served per unit time
(mean service rate) when n customers are present in the
system
λ: Mean arrival rate when λn is constant for all n
µ: Mean service rate when µn is constant for all n ≥ 1
E(w1): Expected waiting time per customer in the system
E(w2): Expected waiting time per customer in the queue
14
The Poisson Process

 Axiom 1: The number of arrivals in non-overlapping


intervals are statistically independent.
 Axiom 2: The probability of more than one arrival between
time t and time (t + ∆t) is O(∆t). That is the
probability of more two or more arrivals during the
small time is negligible.
 Axiom 3: The probability that an arrival occurs between
time t and time (t + ∆t) is {λ.∆t + O(∆t)}

15
The Arrival Theorem

If the arrivals are completely random, then the probability


distribution of the number of arrivals in a fixed time interval
follows Poisson distribution.

 t
e ( t )
n
Pn (t )  , n  0,1, 2, L
n!

16
The Role of Exponential
Distribution

Most analytic results for queuing situations involve the


exponential distribution either as the distribution of inter-arrival
times or service times or both.

The following three properties help to identify the set of


circumstances in which it is reasonable to assume that an
exponential distribution will occur.
- Lack of memory
- Small service times
- Relation to the Poisson distribution
17
Lack of Memory

In an arrival process, this property implies that the


probability that an arrival will occur in the next few minutes
is not influenced by when the last arrival occurred.
(a) There are many individuals who could potentially arrive at
the system.
(b) Each person decides to arrive independently of the other
individuals.
(c) Each individual selects his or her time of arrival completely
at random.

18
Small Service Time

Prob S < t

1.0

0.632

t
10 20 30 40

This graph shows the probability that the service time S is less
than or equal to t if the mean service time is 10.
19
Continued…

The graph showed that more than 63% of the service times
were smaller than the average service time (10).

Compare this to the normal distribution where only 50% of the


service times are smaller than the average.
The practical implication is that an exponential distribution can
best be used to model the distribution of service times in a
system in which a large proportion of “jobs” take a very short
time and only a few “jobs” run for a long time.

20
Relation with Poisson Distribution

If the time between arrivals has an exponential distribution with


parameter, then in a specified period of time the number of
arrivals will have a Poisson distribution.

21
Distribution of Inter-arrival Time

Let T be the time between two consecutive arrivals. If the


arrivals on n-customers in time t follows Poisson
distribution then T follows exponential distribution.

  e  t , 0  t  
f (t )  
0 , elsewhere

22
The Mean Arrival Time


E (T )   tf (t ) dt
t 0

 
t 0
te  t dt

1


23
Distribution of Service Time

If T be the inter-arrival time, then the probability of n-


complete service in time T is given by:

n (t )  P(n-service in time T )
 t
e (t )
n
 , n  0,1, 2,L
n!

24
The Traffic Intensity

For Poisson arrival and departure with one server, the traffic
intensity (ρ) is given by:

mean arrival rate 


 
mean service rate 
The unit of ρ is Erlang.

25
Queuing Models

The general model can be completely represented by


Kendall’s notation as follows:
(a / b / c) : (d / e / f)
a ≡ arrival distribution d ≡ capacity of the system
b ≡ service distribution e ≡ service discipline
c ≡ # service channels f ≡ size of calling source
Standard Notations:
M ≡ Poisson arrival or departure distribution
Ek ≡ Earlangian or Gamma arrival or departure distribution
GI ≡ General Independent arrival distribution
26
(M / M / 1) : (∞ / FCFS / ∞)

Single channel infinite population model

In a steady state condition ρ < 1, it can be shown that


P0  1   1   ,   1

n
  
Pn     1     (1   ),   1, n  0
n

  
27
Characteristics of the Model

1. Expected number of customers in the system E(n):

  
E (n)   nPn   n(1   )  (1   )   n
n n 1

n 0 n0 n 0
 
d d
 (1   )   (  )  (1   ) 
n
  n

n 0 d  d  n 0
d  1  (1   )   
 (1   )     
d   (1   )  (1   ) 2
1    
28
Continued…

2. Expected queue length E(q): Since there is one server, one


customer is in service & (n-1) customers are in the queue.

  
E (q)   (n  1) Pn   nPn   Pn
n 1 n 1 n 1

   
  nPn    Pn  P0    {1  (1   )}
n0  n 0  1 
2 2  
  , since  Pn   (1   ) n  1
1    (   ) n0 n 0
29
Continued…

3. Probability of queue size greater than some finite number N:

 
P(Queue size  N )   Pn   (1   )  n
n N n N
 
 (1   )  N

n N
n N
 (1   )  N
 , where r  n  N
 r

r 0
N
1 
 (1   )  N
   
N

(1   ) 
30
Continued…

4. Expected waiting time per customer in the system E(w1):


Expected # customers in the system E (n) 1 1
E ( w1 )    
Arrival rate      (1   )

5. Expected waiting time per customer in the queue E(w2):

1 1 
E ( w2 )  E ( w1 )  service time of one customer   
     (1   )

31

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy