LECTURE - 3

M/M/1 Queuing Model

Learning objective
To explain the state of the system
To analyze M/M/1 queues
9.6 M|M|1 queuing model
Features of the M|M|1 queuing system are presented in Table 9.3

An infinite population with independent arrivals

Calling Population
and not influenced by the queuing system

Arrival Process Poisson distribution of arrival rate

Queuing configuration Single waiting line with unlimited space

Queue discipline First come, First serve

Service Process Exponential service time distribution

Table 9.3: Features of M|M|1 queuing system

M|M|1 Queuing model following Poisson distribution of arrival () and service rate
() with single server is presented in Figure 9.4. There is no balking and reneging in
this model. The numbers of customers in a queue as well as in the system are
represented by Lq and Ls respectively.

Server


Lq

Ls
Figure 9.4: M|M|1 queuing system
To analyze M|M|1 Queuing model we require information regarding the number of
customers currently present in the system, which is represented by the state of the
system.

9.6.1 State of the system

The state of queuing system is represented by a single number n, the number
of customers currently in the system.
It utilizes memory-less property of exponential distribution. As per this
property the time since the last arrival and the time the current customer has
been in the service process are irrelevant to the future behavior of the
system.
Consider the system to be in steady state, which means that the system has
been running for so long that the current state doesnt depend on the starting
condition.
By computing the long run probabilities of being in each state, we will
determine the performance measures of queuing models as long term steady
state performance measures
Hence, the customers arrive only one customer at a time. The system state
can change only by one unit at a time.

9.6.2 Transition from one state to another state in a queuing

system
If, currently there are n customers in the system, then the following changes can
happen in the system
The state of the system increases from n to n +1, if arrival occurs in to
the service system. The rate of increase is represented by , the arrival
rate.
The state of the system decreases from n to n-1, if departure to the system
occurs. The rate of decrease is represented by , the service rate.
The transition from one state to other state can take place as shown in Figure 9.5.


1 2 3 n-1 n n+1

Figure 9.5: Transition of state in a queuing system

At any given point,

If the system is in the state n,
i. The state of the system moves from n-1 to n at the rate of ( P n1 )( )
ii. The state of the system moves from n to n-1 at the rate of ( P n )( )

Where, ( Pn 1) and ( Pn ) are the probabilities of being in n-1 and n state respectively
as shown in Figure 9.6.
 ( Pn 1 )( )

1 2 3 n-1 n n+1

( Pn )( )
Figure 9.6: Probabilities associated with state of the queuing system

As per the stability condition

( Pn1 )( )=(P n )( )
or

()
( Pn )= (P n1 )

or
( Pn )=( )( Pn1 )

9.6.3 Performance measures of M|M|1 queues

The memory less property is utilized to define the state of the queuing system. To
determine the performance measures, first we will find the probability of having n
number of customers in the queuing system.
Probability of having 1 customer (i.e. n=1) in the service system is:

P1= P0
Similarly,
P2 =P1
P2 =2 P0
.
.
Pn =n P0
We also know that the sum of the probabilities is 1 i.e.,
P0 +P1 +P2 +.. ..=1
or
P0 +P0 + 2 P0 +. .. .= 1
or
(1+ +2 +. . .)P 0= 1
where (1++2 +.. .) is an infinite series,
1
Sum of infinite series can be written as,( )
1-
Hence, P0=1-
i.e. Probability of no customer in the system

To determine performance measuresLs , Lq , W s , Wq in the queuing system we need to

determine the average number of customers in the system.

The average number of customers, Ls in the system can be written as,


L s = nP n
n=0

Where, Pn =n P0
and P0 =(1)

Ls =(1 ) nn
n=0

n1
=(1 ) n
n=0

( n )
We know that =nn1 ,


Ls =(1 )
( )

n=0
n

1
=(1 ) (
1 )
or

Ls =

Using Littles law, according to which the average number of customers in the
service system is the product of arrival rate and average time a customer spends
in the system.

Average time a customer spends in the system, WS can be written using

Littles law as given below

Ls
W s=

We have determined Ls and we know , hence

Ls
W s=

1 1
( ) ()( )
W s=

x

=

 Average time a customer spends in the queue, Wq, can be determined by
subtracting expected service time or average service time from average time a
customer spends in the system, Ws

Wq W s - Expected service time

Ws -
1 1

-
- ( )

( - )


( - )

Using Littles law to determine the average number of customers in the queue,

L q = W q
2

( )
The Performance measures of a M|M|1 Queue can be written as below.
n
Probability of n customers in the service system (1 )

Average number of customers in the system L s=

2

Average number of customers in the queue =
( )

1

 Average time a customer spends in the system =

Average time a customer spends in the queue =
( )
Example 1

The mean time between arrivals of customers in a bank is 3 minutes. Write the
expression for the exponential distribution for average time between arrivals for
any time t (t>=0).

If a customer has already arrived in the bank, what is the probability that the
next customer will come after 10 minutes?

What is the probability that 5 customers will arrive in the one hour interval?

Solution:

The mean time between arrivals = 3 mins

Average arrival rate = (1/3)

= 0.333 arrivals/min

= 20 arrivals/hour

The exponential distribution for average time between arrivals for any time t,
f(t) is
0 .333t
f (t )= 0 .333 e ,t0
The cumulative distribution F(t) is

-0 .333t
F(t )= 1- e ,t0

Solution:
 The probability that the next customer will come after 10 minutes is

F(10)= 1- e -0.333(10 )
0.9643

There is 96.4 percent chance that another customer will arrive in the next 10
minute interval.

Probability that 5 customers will arrive in one-hour interval will follow Poisson
distribution with = 20
(20)5 e20
S0, f (5 )=
5!
5. 49 10-5

Example 2:

In a health clinic, the average rate of arrival of patients is 12 patients per hour.
On an average, a doctor can serve patients at the rate of one patient every four
minutes.

Assume, the arrival of patients follows a Poisson distribution and service to

patients follows an exponential distribution.

Find the average number of patients in the waiting line and in the clinic.

Find the average waiting time in the waiting line or in the queue and also the
average waiting time in the clinic.

Solution:

Poisson arrivals, Exponential service and a single doctor on service, follows a

M|M|1 Queuing model.

The average rate of patient arrival, = 12 patients per hour

The average rate of serving a patient , = 1 in 4 minutes = 15 patients per hour

The utilization factor = (12/15) = 0.8

 Average number of patients in the clinic

L s= = 4 patients

Average number of patients in the waiting line
2
L q = = 3. 2 patients
( )
Average waiting time in the clinic,

W s = 1 = 0.267 hours

Average waiting time in the queue,

W q= = 0 .333 hours
( )