research article a two-stage model queueing with no waiting...
TRANSCRIPT
Hindawi Publishing CorporationMathematical Problems in EngineeringVolume 2013 Article ID 679369 5 pageshttpdxdoiorg1011552013679369
Research ArticleA Two-Stage Model Queueing with No WaitingLine between Channels
Vedat SaLlam1 and Muumljgan Zobu2
1 Department of Statistics Ondokuz Mayıs University 55139 Samsun Turkey2Department of Statistics Amasya University 05100 Amasya Turkey
Correspondence should be addressed to Mujgan Zobu mujganzobuhotmailcom
Received 5 December 2012 Revised 22 April 2013 Accepted 20 May 2013
Academic Editor Suiyang Khoo
Copyright copy 2013 V Saglam and M Zobu This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited
We consider a new queuing model with sequential two stations (stages) single server at each station where no queue is allowed atstation 2 andwith no restriction at station 1There is a FCFS service discipline inwhich the input stream is Poisson having rate 120582Theservice time of any customer at server i (119894 = 1 2) is exponential with parameter 120583
119894The state probabilities and loss probability of this
model are givenThe performance measures are obtained and optimized and additionally the model is simulatedThe simulationresults exact results and optimal results of the performance measures are numerically computed for different parameters
1 Introduction
New models of queuing theory have been needed latelyconcerning the developments in areas such as productionline communication and computer systems One of thesenecessary models is for a tandem queuing system Manyimportant studies have been done in this area The meanwaiting time and the mean customer number at two tandemchannels (servers) which are Poisson arrival and exponentialservice time were given in [1] The mean customer numberdistribution of waiting time and the probabilities of variousnumbers of the tandem queuing system at every stage ofthe Poisson arrival and the exponential service time of thetandem queuing system were found in [2] In [3] it wasproved that if arrivals to the system are Poisson process withthe parameter then the output of this system is also Poissonprocess with a parameter 120582 A more complicated examplewith network analysis was studied in [4] In queuing theoryit is usually assumed that service channels are homogeneousHowever it is seen that in real queuing systems servicechannels sometimes are heterogeneous Understanding suchsystems is important for finding solutions to both theoreticaland technical problems Under the condition that the sum
of service rates is fixed homogeneous systems have beencompared with the heterogeneous systems for performancemeasures in [5ndash9] The measures of effectiveness for tandemqueueswith blockingwere calculated according to an approx-imation method and simulated in [10] The tandem queueswith one server in the first queue and 119899 ge 1 servers in thesecond queue where the arrivals to the system with Poissonprocess having parameter 120582 and there is no waiting roombetween the two stages were analyzed and some probabilitiesfor the number of customers were found in [11] In theliterature usually for some similar models to ours blockedtandemqueueing systems have been studied and probabilitiesof number of customers have been obtained Approximateand simulation results of performance measures have beenobtained In our proposed model there is no restriction forthe first stage there is no waiting room between both stagesand no blocking (ie customers leave the system after hav-ing service in the first stage if the second stage is busy)Probabilities of being-nonbeing of customers in the first andsecond stages performancemeasures and the optimal valuesof these measures are theoretically obtained by analyzing ourproposed method Also we have compared some results byobtaining simulation results
2 Mathematical Problems in Engineering
In real life while there is waiting case in the service sys-tems because of obligations and urgency and unavailabilityof desired features the loss may occur in the beginning ofsecond stage as in our proposed model Therefore we havedecided to construct such a model and analyze itThe follow-ing two examples can be proposed as potential applicationsof our model under some conditions
The following examples can be proposed as the applica-tions of our model on some topics
(a) Fruit-Vegetable Packing Line There is a company whichexports fresh fruits and vegetables This company has aproduct packing linewhich consists of two stages for a specialexport fruit The product comes to first stage for qualitycontrol Later if the product is not in the desirable size orquality then it is taken away from the system (the loss occurs)The product is sent to the second stage to be packed with nowaiting time if it qualifies the desirable size and quality
(b) VoIP (Voice Over Internet Protocol) Internet telephonyrefers to communications servicesmdashvoice fax SMS andorvoice-messaging applicationsmdashthat are transported via an IPnetwork rather than the public switched telephone network(PSTN) The steps involved in originating a VoIP telephonecall are signalling andmedia channel setup digitization of theanalogy voice signal encoding packetization and transmis-sion as Internet Protocol (IP) packets over a packet-switchednetwork On the receiving side similar steps (usually in thereverse order) such as reception of the IP packets decodingof the packets and digital-to-analogy conversion reproducethe original voice stream
More generally the server transfers voice messaging tothe recipient In this system the server is first service therecipient is second service If the recipient is busy and thenthe server destroys voice messaging at that moment
2 The Model
Let 1198991and 119899
2be the number of customers in the first and
second stages respectively at any time of 119905 including thosebeing served where 119899
1= 0 1 2 119899
2= 0 1
For stage 119894 let 120585119894(119905) be defined as follows
120585119894 (119905) =
1198991 119894 = 1
1198992 119894 = 2
(1)
We can denote
11987511989911198992(119905) = Prob 120585
1 (119905) = 1198991 1205852 (119905) = 1198992 (2)
The random process
120585119894 (119905) 119894 = 1 2 119905 ge 0 (3)
is a continuous-time two-dimensionalMarkov chain For 119899 ge1 the state space of this chain becomes
119864 = (0 0) (0 1) (119899 0) (119899 0) (4)
We wish to find the steady-state probability 11990111989911198992
11990111989911198992
= lim119905rarrinfin
11987511989911198992(119905) (5)
The usual procedure leads to the steady-state equations forthis Markov chain
0 = minus12058211990100+ 120583211990110 (6)
0 = minus (120582 + 1205832) 11990101+ 120583111990110+ 120583111990111 (7)
0 = minus (120582 + 1205831) 11990111989910+ 1205821199011198991minus10
+ 120583211990111989911 1198991ge 1 (8)
0 = minus (120582 + 1205831+ 1205832) 11990111989911+ 1205821199011198991minus11
+ 1205831(1199011198991+10
+ 1199011198991+10) 119899
1ge 1
(9)
We define 120588119894= 120582120583
119894for 119894 = 1 2 and 120588 = 120582(120583
1+ 1205832)
21 State Probabilities The probability 1199011198991
denotes the prob-ability of finding 119899
1customers in the first stage at an arbitrary
point in time (see [12]) We can write these as
1199011198991
= 11990111989910+ 11990111989911= 1205881198991
1(1 minus 120588
1) 119899
1ge 0 (10)
11990100+ 11990101= 1 minus 120588
1 (11)
Using this equation and (6) we obtain the following
11990101=120582 (1 minus 120588
1)
120582 + 1205832
=(1 minus 120588
1) 1205882
1 + 1205882
11990100=1205832
12058211990101=1 minus 1205881
1 + 1205882
(12)
By substituting the expression 1199011198991+1
in (9) we get
(120582 + 1205831+ 1205832) 11990111989911= 1205821199011198991minus11
+ 12058311199011198991+1 1198991ge 1 (13)
Let us take 119886 = 120588(1 + 120588) and choose 11990111989911a place to put 119910
1198991
in (13) for simplicity In this case the following equation canbe obtained
1199101198991
= 1198861199101198991minus1+ 1198861199101198991
1198991ge 1 (14)
Both sides of (14) are divided into 1198861198991 and later the index 1198991
is changed to 119896 Then we sum this obtained value
1199101198991
1198861198991minus 1199100=
1198991
sum
119896=1
119901119896
119886119896minus1=1205881(1 minus 120588
1) (1198861198991 minus 1205881198991
1)
(1 minus (1205881119886)) 1198861198991
(15)
11990111989911minus 119886119899111990101= 119886
1205881(1 minus 120588
1) (1198861198991 minus 1205881198991
1)
119886 minus 1205881
(16)
11990111989911= 119886
1205881(1 minus 120588
1) (1198861198991 minus 1205881198991
1)
119886 minus 1205881
+(1 minus 120588
1) 1205882
1 + 1205882
1198861198991 (17)
Mathematical Problems in Engineering 3
When the value of 119886 in (17) is substituted the following equa-tion is obtained
11990111989911=(1 minus 120588
1) 1205882
1 + 1205882
1205881198991
1 (18)
We get the probability 11990111989910from (10) and (18)
11990111989910=(1 minus 120588
1)
1 + 1205882
1205881198991
1 (19)
22 Loss Probability The customerrsquos loss probability is givenas
119901119871=
infin
sum
1198991=0
11990111989911=
1205882
1 + 1205882
(20)
In other way the formula (20) can be obtained from ldquoErlangrsquosloss formulardquo or ldquoErlangrsquos B formulardquo for the119872119872119888119888 queueIn [13] it is denoted as 119861(119888 120588
2) and formulated as the follows
119861 (119888 1205882) =
1205882119888
sum119888
119899=01205881198992119899 (21)
where 1205882is the utilization factor Substituting 119888 = 1 in (21)
we have (20)
3 The Measures of Performance andOptimization of Performance Measures
31 TheMean Sojourn Time Let 119879 be a random variable thatdescribes the sojourn time of customers in the system Usingthe law of total expectation we can write is as follows
119864 (119879) = 119864 (119879 | 119860) 119875 (119860) + 119864 (119879 | 119860)119875 (119860) (22)
where 119875(119860) is the probability of the loss of a customer Nowit is clear that
119864 (119879 | 119860) =1
1205831minus 120582
119864 (119879 | 119860) =1
1205831minus 120582
+1
1205832
(23)
Thus
119864 (119879) =1205831+ 1205832
(1205831minus 120582) (120583
2+ 120582)
(24)
Our main results about the problem of minimizing the meansojourn time can be explained by the following theorem
Theorem 1 If sum of two service rates 1205831+ 1205832= 120583 is fixed
then the mean sojourn time of this tandem system attains itsminimum value for 120583
1= 1205832 + 120582 and 120583
2= 1205832 minus 120582
Proof We will prove the theorem by using the followinginequality
(
119898
prod
I=1
119886119894)
1119898
le1
119898
119898
sum
119894=1
119886119894 forall119886
119894gt 0 forall119898 isin Z
+ (25)
From inequality (25) we have
1
(1205831minus 120582) (120583
2+ 120582)
ge4
1205832 (26)
If we replace the expressions 1205831+1205832= 120583 and 41205832 in equality
(24) we obtain the minimum value of 119864(119879) as follows
min119864 (119879) = 4
120583 (27)
where the equality (27) is provided with 1205831= 1205832 + 120582 and
1205832= 1205832 minus 120582
32 The Mean Number of Customers Let 119873 be the randomvariable that describes the number of customers in the sys-tem
119864 (119873) =
infin
sum
1198991=0
1
sum
1198992=0
(1198991+ 1198992) 11990111989911198992
=120582 (1205831+ 1205832)
(1205831minus 120582) (120583
2+ 120582)
(28a)
or
119864 (119873) = 120582119864 (119879) (28b)
The mean number of customers in the system is opti-mized fromTheorem 1 and the equality (28b) as below
min119864 (119873) = 120582min119864 (119879) = 4120582120583 (29)
The independence of the number of customers can beexpressed by the following theorem
Theorem 2 If the random variables 1198731and 119873
2are taken as
the number of customers in the first and second stages respec-tively then119873
1and119873
2are independent random variables
Proof The joint probability mass functions of 1198731and 119873
2
random variables is
11990111989911198992
= 119875 (1198731= 1198991 1198732= 1198992) 119899
1= 0 1 2 119899
2= 0 1
(30)
If 1198992= 0 and (119899
2= 1) are substituted in (30) the equations
(19) and (18) are obtained respectivelyThe marginal probability mass function of119873
1is given as
(10)
4 Mathematical Problems in Engineering
Table 1 For 120582 = 030 1205831= 080 120583
2= 100
Iterationnumber
Simulation results Exact results Optimal results119864(119879) 119864(119873) 119864(119879) 119864(119873) 119864(119879) 119864(119873)
100 28163 08452 27692 08308 22222 066671000 28111 08433 27692 08308 22222 066675000 28124 08436 27692 08308 22222 06667
Table 2 For 120582 = 01 1205831= 09 120583
2= 09
Iterationnumber
Simulation results Exact results Optimal results119864(119879) 119864(119873) 119864(119879) 119864(119873) 119864(119879) 119864(119873)
100 22661 02263 22500 02250 22220 022221000 22599 02261 22500 02250 22220 022225000 22500 02260 22500 02250 22220 02222
Table 3 For 120582 = 001 1205831= 06 120583
2= 07
Iterationnumber
Simulation results Exact results Optimal results119864(119879) 119864(119873) 119864(119879) 119864(119873) 119864(119879) 119864(119873)
100 31042 00311 31034 00310 30769 003081000 31034 00310 31034 00310 30769 003085000 31043 00310 31034 00310 30769 00308
Themarginal probability mass function of1198732is obtained
from Erlangrsquos B formula or (20)
1199011015840
0= 119875 (119873
2= 0) =
1
1 + 1205882
(31)
1199011015840
1= 119875 (119873
2= 1) =
1205882
1 + 1205882
(32)
11990111989911198992
= 1199011198991
1199011015840
1198992
1198991= 0 1 2 119899
2= 0 1 (33)
If (10) and (31) are substituted in (30) the equation (19) isobtained and if (10) and (32) are substituted in (30) theequation (18) is obtained Thus the independence of 119873
1and
1198732has been demonstrated
4 Numerical Results
The random arrivals and service times were generated fromexponential distribution as seconds by using MATLAB 7100(R2010a) programming for this proposedmodelThe numberof customers taken was 10000 and was performed in threeiterations Performance measures are calculated for differentvalues of 120588
119894(119894 = 1 2) and three different iterations steps that
is 100 1000 and 5000 These results were shown in Tables 12 and 3
5 Conclusions
A new queuing discipline is given for a Markov model whichconsists of two consecutive channels and no waiting line
between channels In this model steady-state equations themean sojourn time the mean number of customers andloss probability are obtained Additionally two theorems aregivenwhich are about optimization of performancemeasuresand the independent of the number of customers respec-tively Performance measures are calculated for different val-ues of 120588
119894(119894 = 1 2) and for three different iterations steps that
is 100 1000 and 5000 Moreover results of these measuresare compared in the tables above It has been seen thatthe simulation results approximated the theoretical resultsAlthough the iteration number is increased the simulationresults of performance measures have not changed Howeveras 120588119894(119894 = 1 2) converges to zero both the simulation results
and exact results approximately are equal to optimal resultsThus it is said that our proposed queueing model operateswell
For further research a model in which a customer whocompleted his service in channel 1 blocks channel 1 withprobability 120587 or leaves the systemwith probability 1minus120587 whilechannel 2 is busy can be studied
References
[1] G G OrsquoBrien ldquoThe solution of some queuing problemsrdquo Jour-nal of the Society For Industrial and Applied Mathematics vol 2pp 133ndash142 1954
[2] R R P Jackson ldquoQueuing system with phase-type servicerdquoOperational Research Quarterly vol 5 pp 109ndash120 1954
[3] P J Burke ldquoThe output of a queuing systemrdquo OperationsResearch vol 4 pp 699ndash704 1956
Mathematical Problems in Engineering 5
[4] J R Jackson ldquoNetwork of waiting linesrdquo Operation Researchvol 5 pp 518ndash524 1957
[5] R B Cooper Introduction to Queuing Theory The MacmillanCompany New York NY USA 1972
[6] D Gross and C M Harris Fundamentals of Queueing TheoryJohn Wiley amp Sons New York NY USA 3rd edition 1998
[7] V Saglam and H Torun ldquoOn optimization of stochastic servicesystemwith two heterogeneous channelsrdquo International Journalof Applied Mathematics vol 17 no 1 pp 1ndash6 2005
[8] V Saglam and A Shahbazov ldquoMinimizing loss probability inqueuing systems with heterogeneous serversrdquo Iranian Journalof Science and Technology vol 31 no 2 pp 199ndash206 2007
[9] A H Taha Operating Research Macmillian Publishing NewYork NY USA 1982
[10] A Brandwajn and Y-L L Jow ldquoAn approximation method fortandem queues with blockingrdquoOperations Research vol 36 no1 pp 73ndash83 1988
[11] A A Akinsete ldquoBlocked network of tandem queues with with-drawalrdquo Kragujevac Journal of Mathematics vol 23 pp 63ndash732001
[12] U N Bhat An Introduction to Queueing Theory Boston MassUSA 2008
[13] W J Stewart Probability Markov Chains Queues and Simula-tion Princeton University Press Princeton NJ USA 2009
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
2 Mathematical Problems in Engineering
In real life while there is waiting case in the service sys-tems because of obligations and urgency and unavailabilityof desired features the loss may occur in the beginning ofsecond stage as in our proposed model Therefore we havedecided to construct such a model and analyze itThe follow-ing two examples can be proposed as potential applicationsof our model under some conditions
The following examples can be proposed as the applica-tions of our model on some topics
(a) Fruit-Vegetable Packing Line There is a company whichexports fresh fruits and vegetables This company has aproduct packing linewhich consists of two stages for a specialexport fruit The product comes to first stage for qualitycontrol Later if the product is not in the desirable size orquality then it is taken away from the system (the loss occurs)The product is sent to the second stage to be packed with nowaiting time if it qualifies the desirable size and quality
(b) VoIP (Voice Over Internet Protocol) Internet telephonyrefers to communications servicesmdashvoice fax SMS andorvoice-messaging applicationsmdashthat are transported via an IPnetwork rather than the public switched telephone network(PSTN) The steps involved in originating a VoIP telephonecall are signalling andmedia channel setup digitization of theanalogy voice signal encoding packetization and transmis-sion as Internet Protocol (IP) packets over a packet-switchednetwork On the receiving side similar steps (usually in thereverse order) such as reception of the IP packets decodingof the packets and digital-to-analogy conversion reproducethe original voice stream
More generally the server transfers voice messaging tothe recipient In this system the server is first service therecipient is second service If the recipient is busy and thenthe server destroys voice messaging at that moment
2 The Model
Let 1198991and 119899
2be the number of customers in the first and
second stages respectively at any time of 119905 including thosebeing served where 119899
1= 0 1 2 119899
2= 0 1
For stage 119894 let 120585119894(119905) be defined as follows
120585119894 (119905) =
1198991 119894 = 1
1198992 119894 = 2
(1)
We can denote
11987511989911198992(119905) = Prob 120585
1 (119905) = 1198991 1205852 (119905) = 1198992 (2)
The random process
120585119894 (119905) 119894 = 1 2 119905 ge 0 (3)
is a continuous-time two-dimensionalMarkov chain For 119899 ge1 the state space of this chain becomes
119864 = (0 0) (0 1) (119899 0) (119899 0) (4)
We wish to find the steady-state probability 11990111989911198992
11990111989911198992
= lim119905rarrinfin
11987511989911198992(119905) (5)
The usual procedure leads to the steady-state equations forthis Markov chain
0 = minus12058211990100+ 120583211990110 (6)
0 = minus (120582 + 1205832) 11990101+ 120583111990110+ 120583111990111 (7)
0 = minus (120582 + 1205831) 11990111989910+ 1205821199011198991minus10
+ 120583211990111989911 1198991ge 1 (8)
0 = minus (120582 + 1205831+ 1205832) 11990111989911+ 1205821199011198991minus11
+ 1205831(1199011198991+10
+ 1199011198991+10) 119899
1ge 1
(9)
We define 120588119894= 120582120583
119894for 119894 = 1 2 and 120588 = 120582(120583
1+ 1205832)
21 State Probabilities The probability 1199011198991
denotes the prob-ability of finding 119899
1customers in the first stage at an arbitrary
point in time (see [12]) We can write these as
1199011198991
= 11990111989910+ 11990111989911= 1205881198991
1(1 minus 120588
1) 119899
1ge 0 (10)
11990100+ 11990101= 1 minus 120588
1 (11)
Using this equation and (6) we obtain the following
11990101=120582 (1 minus 120588
1)
120582 + 1205832
=(1 minus 120588
1) 1205882
1 + 1205882
11990100=1205832
12058211990101=1 minus 1205881
1 + 1205882
(12)
By substituting the expression 1199011198991+1
in (9) we get
(120582 + 1205831+ 1205832) 11990111989911= 1205821199011198991minus11
+ 12058311199011198991+1 1198991ge 1 (13)
Let us take 119886 = 120588(1 + 120588) and choose 11990111989911a place to put 119910
1198991
in (13) for simplicity In this case the following equation canbe obtained
1199101198991
= 1198861199101198991minus1+ 1198861199101198991
1198991ge 1 (14)
Both sides of (14) are divided into 1198861198991 and later the index 1198991
is changed to 119896 Then we sum this obtained value
1199101198991
1198861198991minus 1199100=
1198991
sum
119896=1
119901119896
119886119896minus1=1205881(1 minus 120588
1) (1198861198991 minus 1205881198991
1)
(1 minus (1205881119886)) 1198861198991
(15)
11990111989911minus 119886119899111990101= 119886
1205881(1 minus 120588
1) (1198861198991 minus 1205881198991
1)
119886 minus 1205881
(16)
11990111989911= 119886
1205881(1 minus 120588
1) (1198861198991 minus 1205881198991
1)
119886 minus 1205881
+(1 minus 120588
1) 1205882
1 + 1205882
1198861198991 (17)
Mathematical Problems in Engineering 3
When the value of 119886 in (17) is substituted the following equa-tion is obtained
11990111989911=(1 minus 120588
1) 1205882
1 + 1205882
1205881198991
1 (18)
We get the probability 11990111989910from (10) and (18)
11990111989910=(1 minus 120588
1)
1 + 1205882
1205881198991
1 (19)
22 Loss Probability The customerrsquos loss probability is givenas
119901119871=
infin
sum
1198991=0
11990111989911=
1205882
1 + 1205882
(20)
In other way the formula (20) can be obtained from ldquoErlangrsquosloss formulardquo or ldquoErlangrsquos B formulardquo for the119872119872119888119888 queueIn [13] it is denoted as 119861(119888 120588
2) and formulated as the follows
119861 (119888 1205882) =
1205882119888
sum119888
119899=01205881198992119899 (21)
where 1205882is the utilization factor Substituting 119888 = 1 in (21)
we have (20)
3 The Measures of Performance andOptimization of Performance Measures
31 TheMean Sojourn Time Let 119879 be a random variable thatdescribes the sojourn time of customers in the system Usingthe law of total expectation we can write is as follows
119864 (119879) = 119864 (119879 | 119860) 119875 (119860) + 119864 (119879 | 119860)119875 (119860) (22)
where 119875(119860) is the probability of the loss of a customer Nowit is clear that
119864 (119879 | 119860) =1
1205831minus 120582
119864 (119879 | 119860) =1
1205831minus 120582
+1
1205832
(23)
Thus
119864 (119879) =1205831+ 1205832
(1205831minus 120582) (120583
2+ 120582)
(24)
Our main results about the problem of minimizing the meansojourn time can be explained by the following theorem
Theorem 1 If sum of two service rates 1205831+ 1205832= 120583 is fixed
then the mean sojourn time of this tandem system attains itsminimum value for 120583
1= 1205832 + 120582 and 120583
2= 1205832 minus 120582
Proof We will prove the theorem by using the followinginequality
(
119898
prod
I=1
119886119894)
1119898
le1
119898
119898
sum
119894=1
119886119894 forall119886
119894gt 0 forall119898 isin Z
+ (25)
From inequality (25) we have
1
(1205831minus 120582) (120583
2+ 120582)
ge4
1205832 (26)
If we replace the expressions 1205831+1205832= 120583 and 41205832 in equality
(24) we obtain the minimum value of 119864(119879) as follows
min119864 (119879) = 4
120583 (27)
where the equality (27) is provided with 1205831= 1205832 + 120582 and
1205832= 1205832 minus 120582
32 The Mean Number of Customers Let 119873 be the randomvariable that describes the number of customers in the sys-tem
119864 (119873) =
infin
sum
1198991=0
1
sum
1198992=0
(1198991+ 1198992) 11990111989911198992
=120582 (1205831+ 1205832)
(1205831minus 120582) (120583
2+ 120582)
(28a)
or
119864 (119873) = 120582119864 (119879) (28b)
The mean number of customers in the system is opti-mized fromTheorem 1 and the equality (28b) as below
min119864 (119873) = 120582min119864 (119879) = 4120582120583 (29)
The independence of the number of customers can beexpressed by the following theorem
Theorem 2 If the random variables 1198731and 119873
2are taken as
the number of customers in the first and second stages respec-tively then119873
1and119873
2are independent random variables
Proof The joint probability mass functions of 1198731and 119873
2
random variables is
11990111989911198992
= 119875 (1198731= 1198991 1198732= 1198992) 119899
1= 0 1 2 119899
2= 0 1
(30)
If 1198992= 0 and (119899
2= 1) are substituted in (30) the equations
(19) and (18) are obtained respectivelyThe marginal probability mass function of119873
1is given as
(10)
4 Mathematical Problems in Engineering
Table 1 For 120582 = 030 1205831= 080 120583
2= 100
Iterationnumber
Simulation results Exact results Optimal results119864(119879) 119864(119873) 119864(119879) 119864(119873) 119864(119879) 119864(119873)
100 28163 08452 27692 08308 22222 066671000 28111 08433 27692 08308 22222 066675000 28124 08436 27692 08308 22222 06667
Table 2 For 120582 = 01 1205831= 09 120583
2= 09
Iterationnumber
Simulation results Exact results Optimal results119864(119879) 119864(119873) 119864(119879) 119864(119873) 119864(119879) 119864(119873)
100 22661 02263 22500 02250 22220 022221000 22599 02261 22500 02250 22220 022225000 22500 02260 22500 02250 22220 02222
Table 3 For 120582 = 001 1205831= 06 120583
2= 07
Iterationnumber
Simulation results Exact results Optimal results119864(119879) 119864(119873) 119864(119879) 119864(119873) 119864(119879) 119864(119873)
100 31042 00311 31034 00310 30769 003081000 31034 00310 31034 00310 30769 003085000 31043 00310 31034 00310 30769 00308
Themarginal probability mass function of1198732is obtained
from Erlangrsquos B formula or (20)
1199011015840
0= 119875 (119873
2= 0) =
1
1 + 1205882
(31)
1199011015840
1= 119875 (119873
2= 1) =
1205882
1 + 1205882
(32)
11990111989911198992
= 1199011198991
1199011015840
1198992
1198991= 0 1 2 119899
2= 0 1 (33)
If (10) and (31) are substituted in (30) the equation (19) isobtained and if (10) and (32) are substituted in (30) theequation (18) is obtained Thus the independence of 119873
1and
1198732has been demonstrated
4 Numerical Results
The random arrivals and service times were generated fromexponential distribution as seconds by using MATLAB 7100(R2010a) programming for this proposedmodelThe numberof customers taken was 10000 and was performed in threeiterations Performance measures are calculated for differentvalues of 120588
119894(119894 = 1 2) and three different iterations steps that
is 100 1000 and 5000 These results were shown in Tables 12 and 3
5 Conclusions
A new queuing discipline is given for a Markov model whichconsists of two consecutive channels and no waiting line
between channels In this model steady-state equations themean sojourn time the mean number of customers andloss probability are obtained Additionally two theorems aregivenwhich are about optimization of performancemeasuresand the independent of the number of customers respec-tively Performance measures are calculated for different val-ues of 120588
119894(119894 = 1 2) and for three different iterations steps that
is 100 1000 and 5000 Moreover results of these measuresare compared in the tables above It has been seen thatthe simulation results approximated the theoretical resultsAlthough the iteration number is increased the simulationresults of performance measures have not changed Howeveras 120588119894(119894 = 1 2) converges to zero both the simulation results
and exact results approximately are equal to optimal resultsThus it is said that our proposed queueing model operateswell
For further research a model in which a customer whocompleted his service in channel 1 blocks channel 1 withprobability 120587 or leaves the systemwith probability 1minus120587 whilechannel 2 is busy can be studied
References
[1] G G OrsquoBrien ldquoThe solution of some queuing problemsrdquo Jour-nal of the Society For Industrial and Applied Mathematics vol 2pp 133ndash142 1954
[2] R R P Jackson ldquoQueuing system with phase-type servicerdquoOperational Research Quarterly vol 5 pp 109ndash120 1954
[3] P J Burke ldquoThe output of a queuing systemrdquo OperationsResearch vol 4 pp 699ndash704 1956
Mathematical Problems in Engineering 5
[4] J R Jackson ldquoNetwork of waiting linesrdquo Operation Researchvol 5 pp 518ndash524 1957
[5] R B Cooper Introduction to Queuing Theory The MacmillanCompany New York NY USA 1972
[6] D Gross and C M Harris Fundamentals of Queueing TheoryJohn Wiley amp Sons New York NY USA 3rd edition 1998
[7] V Saglam and H Torun ldquoOn optimization of stochastic servicesystemwith two heterogeneous channelsrdquo International Journalof Applied Mathematics vol 17 no 1 pp 1ndash6 2005
[8] V Saglam and A Shahbazov ldquoMinimizing loss probability inqueuing systems with heterogeneous serversrdquo Iranian Journalof Science and Technology vol 31 no 2 pp 199ndash206 2007
[9] A H Taha Operating Research Macmillian Publishing NewYork NY USA 1982
[10] A Brandwajn and Y-L L Jow ldquoAn approximation method fortandem queues with blockingrdquoOperations Research vol 36 no1 pp 73ndash83 1988
[11] A A Akinsete ldquoBlocked network of tandem queues with with-drawalrdquo Kragujevac Journal of Mathematics vol 23 pp 63ndash732001
[12] U N Bhat An Introduction to Queueing Theory Boston MassUSA 2008
[13] W J Stewart Probability Markov Chains Queues and Simula-tion Princeton University Press Princeton NJ USA 2009
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
Mathematical Problems in Engineering 3
When the value of 119886 in (17) is substituted the following equa-tion is obtained
11990111989911=(1 minus 120588
1) 1205882
1 + 1205882
1205881198991
1 (18)
We get the probability 11990111989910from (10) and (18)
11990111989910=(1 minus 120588
1)
1 + 1205882
1205881198991
1 (19)
22 Loss Probability The customerrsquos loss probability is givenas
119901119871=
infin
sum
1198991=0
11990111989911=
1205882
1 + 1205882
(20)
In other way the formula (20) can be obtained from ldquoErlangrsquosloss formulardquo or ldquoErlangrsquos B formulardquo for the119872119872119888119888 queueIn [13] it is denoted as 119861(119888 120588
2) and formulated as the follows
119861 (119888 1205882) =
1205882119888
sum119888
119899=01205881198992119899 (21)
where 1205882is the utilization factor Substituting 119888 = 1 in (21)
we have (20)
3 The Measures of Performance andOptimization of Performance Measures
31 TheMean Sojourn Time Let 119879 be a random variable thatdescribes the sojourn time of customers in the system Usingthe law of total expectation we can write is as follows
119864 (119879) = 119864 (119879 | 119860) 119875 (119860) + 119864 (119879 | 119860)119875 (119860) (22)
where 119875(119860) is the probability of the loss of a customer Nowit is clear that
119864 (119879 | 119860) =1
1205831minus 120582
119864 (119879 | 119860) =1
1205831minus 120582
+1
1205832
(23)
Thus
119864 (119879) =1205831+ 1205832
(1205831minus 120582) (120583
2+ 120582)
(24)
Our main results about the problem of minimizing the meansojourn time can be explained by the following theorem
Theorem 1 If sum of two service rates 1205831+ 1205832= 120583 is fixed
then the mean sojourn time of this tandem system attains itsminimum value for 120583
1= 1205832 + 120582 and 120583
2= 1205832 minus 120582
Proof We will prove the theorem by using the followinginequality
(
119898
prod
I=1
119886119894)
1119898
le1
119898
119898
sum
119894=1
119886119894 forall119886
119894gt 0 forall119898 isin Z
+ (25)
From inequality (25) we have
1
(1205831minus 120582) (120583
2+ 120582)
ge4
1205832 (26)
If we replace the expressions 1205831+1205832= 120583 and 41205832 in equality
(24) we obtain the minimum value of 119864(119879) as follows
min119864 (119879) = 4
120583 (27)
where the equality (27) is provided with 1205831= 1205832 + 120582 and
1205832= 1205832 minus 120582
32 The Mean Number of Customers Let 119873 be the randomvariable that describes the number of customers in the sys-tem
119864 (119873) =
infin
sum
1198991=0
1
sum
1198992=0
(1198991+ 1198992) 11990111989911198992
=120582 (1205831+ 1205832)
(1205831minus 120582) (120583
2+ 120582)
(28a)
or
119864 (119873) = 120582119864 (119879) (28b)
The mean number of customers in the system is opti-mized fromTheorem 1 and the equality (28b) as below
min119864 (119873) = 120582min119864 (119879) = 4120582120583 (29)
The independence of the number of customers can beexpressed by the following theorem
Theorem 2 If the random variables 1198731and 119873
2are taken as
the number of customers in the first and second stages respec-tively then119873
1and119873
2are independent random variables
Proof The joint probability mass functions of 1198731and 119873
2
random variables is
11990111989911198992
= 119875 (1198731= 1198991 1198732= 1198992) 119899
1= 0 1 2 119899
2= 0 1
(30)
If 1198992= 0 and (119899
2= 1) are substituted in (30) the equations
(19) and (18) are obtained respectivelyThe marginal probability mass function of119873
1is given as
(10)
4 Mathematical Problems in Engineering
Table 1 For 120582 = 030 1205831= 080 120583
2= 100
Iterationnumber
Simulation results Exact results Optimal results119864(119879) 119864(119873) 119864(119879) 119864(119873) 119864(119879) 119864(119873)
100 28163 08452 27692 08308 22222 066671000 28111 08433 27692 08308 22222 066675000 28124 08436 27692 08308 22222 06667
Table 2 For 120582 = 01 1205831= 09 120583
2= 09
Iterationnumber
Simulation results Exact results Optimal results119864(119879) 119864(119873) 119864(119879) 119864(119873) 119864(119879) 119864(119873)
100 22661 02263 22500 02250 22220 022221000 22599 02261 22500 02250 22220 022225000 22500 02260 22500 02250 22220 02222
Table 3 For 120582 = 001 1205831= 06 120583
2= 07
Iterationnumber
Simulation results Exact results Optimal results119864(119879) 119864(119873) 119864(119879) 119864(119873) 119864(119879) 119864(119873)
100 31042 00311 31034 00310 30769 003081000 31034 00310 31034 00310 30769 003085000 31043 00310 31034 00310 30769 00308
Themarginal probability mass function of1198732is obtained
from Erlangrsquos B formula or (20)
1199011015840
0= 119875 (119873
2= 0) =
1
1 + 1205882
(31)
1199011015840
1= 119875 (119873
2= 1) =
1205882
1 + 1205882
(32)
11990111989911198992
= 1199011198991
1199011015840
1198992
1198991= 0 1 2 119899
2= 0 1 (33)
If (10) and (31) are substituted in (30) the equation (19) isobtained and if (10) and (32) are substituted in (30) theequation (18) is obtained Thus the independence of 119873
1and
1198732has been demonstrated
4 Numerical Results
The random arrivals and service times were generated fromexponential distribution as seconds by using MATLAB 7100(R2010a) programming for this proposedmodelThe numberof customers taken was 10000 and was performed in threeiterations Performance measures are calculated for differentvalues of 120588
119894(119894 = 1 2) and three different iterations steps that
is 100 1000 and 5000 These results were shown in Tables 12 and 3
5 Conclusions
A new queuing discipline is given for a Markov model whichconsists of two consecutive channels and no waiting line
between channels In this model steady-state equations themean sojourn time the mean number of customers andloss probability are obtained Additionally two theorems aregivenwhich are about optimization of performancemeasuresand the independent of the number of customers respec-tively Performance measures are calculated for different val-ues of 120588
119894(119894 = 1 2) and for three different iterations steps that
is 100 1000 and 5000 Moreover results of these measuresare compared in the tables above It has been seen thatthe simulation results approximated the theoretical resultsAlthough the iteration number is increased the simulationresults of performance measures have not changed Howeveras 120588119894(119894 = 1 2) converges to zero both the simulation results
and exact results approximately are equal to optimal resultsThus it is said that our proposed queueing model operateswell
For further research a model in which a customer whocompleted his service in channel 1 blocks channel 1 withprobability 120587 or leaves the systemwith probability 1minus120587 whilechannel 2 is busy can be studied
References
[1] G G OrsquoBrien ldquoThe solution of some queuing problemsrdquo Jour-nal of the Society For Industrial and Applied Mathematics vol 2pp 133ndash142 1954
[2] R R P Jackson ldquoQueuing system with phase-type servicerdquoOperational Research Quarterly vol 5 pp 109ndash120 1954
[3] P J Burke ldquoThe output of a queuing systemrdquo OperationsResearch vol 4 pp 699ndash704 1956
Mathematical Problems in Engineering 5
[4] J R Jackson ldquoNetwork of waiting linesrdquo Operation Researchvol 5 pp 518ndash524 1957
[5] R B Cooper Introduction to Queuing Theory The MacmillanCompany New York NY USA 1972
[6] D Gross and C M Harris Fundamentals of Queueing TheoryJohn Wiley amp Sons New York NY USA 3rd edition 1998
[7] V Saglam and H Torun ldquoOn optimization of stochastic servicesystemwith two heterogeneous channelsrdquo International Journalof Applied Mathematics vol 17 no 1 pp 1ndash6 2005
[8] V Saglam and A Shahbazov ldquoMinimizing loss probability inqueuing systems with heterogeneous serversrdquo Iranian Journalof Science and Technology vol 31 no 2 pp 199ndash206 2007
[9] A H Taha Operating Research Macmillian Publishing NewYork NY USA 1982
[10] A Brandwajn and Y-L L Jow ldquoAn approximation method fortandem queues with blockingrdquoOperations Research vol 36 no1 pp 73ndash83 1988
[11] A A Akinsete ldquoBlocked network of tandem queues with with-drawalrdquo Kragujevac Journal of Mathematics vol 23 pp 63ndash732001
[12] U N Bhat An Introduction to Queueing Theory Boston MassUSA 2008
[13] W J Stewart Probability Markov Chains Queues and Simula-tion Princeton University Press Princeton NJ USA 2009
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
4 Mathematical Problems in Engineering
Table 1 For 120582 = 030 1205831= 080 120583
2= 100
Iterationnumber
Simulation results Exact results Optimal results119864(119879) 119864(119873) 119864(119879) 119864(119873) 119864(119879) 119864(119873)
100 28163 08452 27692 08308 22222 066671000 28111 08433 27692 08308 22222 066675000 28124 08436 27692 08308 22222 06667
Table 2 For 120582 = 01 1205831= 09 120583
2= 09
Iterationnumber
Simulation results Exact results Optimal results119864(119879) 119864(119873) 119864(119879) 119864(119873) 119864(119879) 119864(119873)
100 22661 02263 22500 02250 22220 022221000 22599 02261 22500 02250 22220 022225000 22500 02260 22500 02250 22220 02222
Table 3 For 120582 = 001 1205831= 06 120583
2= 07
Iterationnumber
Simulation results Exact results Optimal results119864(119879) 119864(119873) 119864(119879) 119864(119873) 119864(119879) 119864(119873)
100 31042 00311 31034 00310 30769 003081000 31034 00310 31034 00310 30769 003085000 31043 00310 31034 00310 30769 00308
Themarginal probability mass function of1198732is obtained
from Erlangrsquos B formula or (20)
1199011015840
0= 119875 (119873
2= 0) =
1
1 + 1205882
(31)
1199011015840
1= 119875 (119873
2= 1) =
1205882
1 + 1205882
(32)
11990111989911198992
= 1199011198991
1199011015840
1198992
1198991= 0 1 2 119899
2= 0 1 (33)
If (10) and (31) are substituted in (30) the equation (19) isobtained and if (10) and (32) are substituted in (30) theequation (18) is obtained Thus the independence of 119873
1and
1198732has been demonstrated
4 Numerical Results
The random arrivals and service times were generated fromexponential distribution as seconds by using MATLAB 7100(R2010a) programming for this proposedmodelThe numberof customers taken was 10000 and was performed in threeiterations Performance measures are calculated for differentvalues of 120588
119894(119894 = 1 2) and three different iterations steps that
is 100 1000 and 5000 These results were shown in Tables 12 and 3
5 Conclusions
A new queuing discipline is given for a Markov model whichconsists of two consecutive channels and no waiting line
between channels In this model steady-state equations themean sojourn time the mean number of customers andloss probability are obtained Additionally two theorems aregivenwhich are about optimization of performancemeasuresand the independent of the number of customers respec-tively Performance measures are calculated for different val-ues of 120588
119894(119894 = 1 2) and for three different iterations steps that
is 100 1000 and 5000 Moreover results of these measuresare compared in the tables above It has been seen thatthe simulation results approximated the theoretical resultsAlthough the iteration number is increased the simulationresults of performance measures have not changed Howeveras 120588119894(119894 = 1 2) converges to zero both the simulation results
and exact results approximately are equal to optimal resultsThus it is said that our proposed queueing model operateswell
For further research a model in which a customer whocompleted his service in channel 1 blocks channel 1 withprobability 120587 or leaves the systemwith probability 1minus120587 whilechannel 2 is busy can be studied
References
[1] G G OrsquoBrien ldquoThe solution of some queuing problemsrdquo Jour-nal of the Society For Industrial and Applied Mathematics vol 2pp 133ndash142 1954
[2] R R P Jackson ldquoQueuing system with phase-type servicerdquoOperational Research Quarterly vol 5 pp 109ndash120 1954
[3] P J Burke ldquoThe output of a queuing systemrdquo OperationsResearch vol 4 pp 699ndash704 1956
Mathematical Problems in Engineering 5
[4] J R Jackson ldquoNetwork of waiting linesrdquo Operation Researchvol 5 pp 518ndash524 1957
[5] R B Cooper Introduction to Queuing Theory The MacmillanCompany New York NY USA 1972
[6] D Gross and C M Harris Fundamentals of Queueing TheoryJohn Wiley amp Sons New York NY USA 3rd edition 1998
[7] V Saglam and H Torun ldquoOn optimization of stochastic servicesystemwith two heterogeneous channelsrdquo International Journalof Applied Mathematics vol 17 no 1 pp 1ndash6 2005
[8] V Saglam and A Shahbazov ldquoMinimizing loss probability inqueuing systems with heterogeneous serversrdquo Iranian Journalof Science and Technology vol 31 no 2 pp 199ndash206 2007
[9] A H Taha Operating Research Macmillian Publishing NewYork NY USA 1982
[10] A Brandwajn and Y-L L Jow ldquoAn approximation method fortandem queues with blockingrdquoOperations Research vol 36 no1 pp 73ndash83 1988
[11] A A Akinsete ldquoBlocked network of tandem queues with with-drawalrdquo Kragujevac Journal of Mathematics vol 23 pp 63ndash732001
[12] U N Bhat An Introduction to Queueing Theory Boston MassUSA 2008
[13] W J Stewart Probability Markov Chains Queues and Simula-tion Princeton University Press Princeton NJ USA 2009
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
Mathematical Problems in Engineering 5
[4] J R Jackson ldquoNetwork of waiting linesrdquo Operation Researchvol 5 pp 518ndash524 1957
[5] R B Cooper Introduction to Queuing Theory The MacmillanCompany New York NY USA 1972
[6] D Gross and C M Harris Fundamentals of Queueing TheoryJohn Wiley amp Sons New York NY USA 3rd edition 1998
[7] V Saglam and H Torun ldquoOn optimization of stochastic servicesystemwith two heterogeneous channelsrdquo International Journalof Applied Mathematics vol 17 no 1 pp 1ndash6 2005
[8] V Saglam and A Shahbazov ldquoMinimizing loss probability inqueuing systems with heterogeneous serversrdquo Iranian Journalof Science and Technology vol 31 no 2 pp 199ndash206 2007
[9] A H Taha Operating Research Macmillian Publishing NewYork NY USA 1982
[10] A Brandwajn and Y-L L Jow ldquoAn approximation method fortandem queues with blockingrdquoOperations Research vol 36 no1 pp 73ndash83 1988
[11] A A Akinsete ldquoBlocked network of tandem queues with with-drawalrdquo Kragujevac Journal of Mathematics vol 23 pp 63ndash732001
[12] U N Bhat An Introduction to Queueing Theory Boston MassUSA 2008
[13] W J Stewart Probability Markov Chains Queues and Simula-tion Princeton University Press Princeton NJ USA 2009
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of
Submit your manuscripts athttpwwwhindawicom
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttpwwwhindawicom
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Stochastic AnalysisInternational Journal of