Research & Reviews: Journal of Statistics
Special Issue on Recent Statistical Methodologies and Applications (Vol.2)
e-ISSN: 2278-2273, p-ISSN: 2348-7909
RRJS (2014) 94-98 © STM Journals (2014). All Rights Reserved Page 94
Parameter Estimation of Single Server Queue with
Working Vacations
K.Senthamarai Kannan
1
and A. Jabarali
2
*
1
Professor,
2
Research Scholar, Department of Statistics, Manonmaniam Sundaranar University,
Tirunelveli, Tamil Nadu - 627 012, India
Abstract
Queue is a waiting line phenomenon. The server may take the vacation due to breakdown
or other reasons. If the server is on vacation, there is a situation the customers have to
wait for service during that period. The generalization of the classical vacation model is
considered as the server which gives the service at different rate rather than completely
stopping the service during the vacation period. This paper deals with the estimation of
the parameters in M/M/1 queue during working vacations period. The maximum
likelihood estimation procedure is applied to obtain the estimation of the model
parameters and also the performance measure of queueing system had been discussed.
Finally computational results are presented for the queueing model.
Keywords: M/M/1 queue, vacation model, working vacations, maximum likelihood
estimation, performance measures.
*Author for Correspondence E-mail: statjafar@gmail.com
INTRODUCTION
Statistical inference in queueing theory has
drawn the attention of researchers in the past
few decades. The problem of estimation
concerning the parameters of queueing process
such as arrival rate, service rate and traffic
intensity is the most important thing in the
queueing systems [1, 2]. The pioneer
investigators have derived the Maximum
Likelihood Estimates(MLE) for the arrival and
service parameters of an M/M/1 queueing
model [3] and infinite server queueing model
[4]. Examining the MLE and Moment
Estimate(ME) of the parameters of the inter-
arrival and service time distributions of GI/G/1
queue model are discussed in Ref. [5].
Consequently, the inferential procedures
concerning the traffic intensity of M/E
K
/1
queue model are discussed in Ref. [6]. The
stationary solution of MLE of Markovian two
server queueing model parameters have been
obtained in the case of non-identical servers
[7]. Later, stationary solution of MLE of the
generalized form of multi-server queueing
model in the presence of non-identical servers
and the Confidence Intervals(CI) of these
model parameters are obtained [8].
Meanwhile, the MLE and Bayesian estimates
of the M/M/1/1 queueing model parameters
are explained and discussed the large sample
test for the model parameters [9]. Over the last
two decades, the queueing systems with
vacations have been well studied because of
their applications in modeling the computer
networks, communication and manufacturing/
service systems. In the previous studies of
classical vacation queueing models, it is
assumed that the server stops the primary
service completely and utilizes the vacation
time for other secondary jobs. The working
vacations in M/M/1 queueing model was
introduced in Ref. [10], the concept of
working vacation in M/M/1 queue, where the
server works at a lower rate rather than
completely stopping the primary service
during the vacation period. They have
considered multiple vacation policy and
derived the simple explicit formula for the
mean, variance, the probability generating
function of number of customers in the
system, the Laplace-Stieltjes Transform(LST)
of waiting time distribution and the
applications of these queueing model is
connected to performance analysis of gateway
router in communication networks. Recently,
working vacation in M/M/1 queueing model
was considered in Ref. [11], they have
obtained the stationary distribution of the