Mathematics of Operations Research · 2013 · 42 citations · 16 references
BusinessMathematicsProbability TheoryQueuing TheoryStochastic ProcessesQueue Length DistributionsQueueing AnalysisQueueing NetworksQueueing ProcessesQueueing SystemsQueueing TheoryAsymptotic Analysis Of QueuesAsymptotic Queueing TheoryFluid ModelsInterarrival Time DistributionsMany-server Limit TheoryMany-server Queue AsymptoticsPriority Queueing With RenegingQueue Length Asymptotics
A multiclass many-server system is considered, in which customers are served according to a nonpreemptive priority policy and may renege while waiting to enter service. The service and reneging time distributions satisfy mild conditions. Building on an approach developed by Kaspi and Ramanan, the law-of-large-numbers many-server asymptotics are characterized as the unique solution to a set of differential equations in a measure space, regarded as fluid model equations. In stationarity, convergence to the explicitly solved invariant state of the fluid-model equations is established. An immediate consequence of the results in the case of exponential reneging is the asymptotic optimality of an index policy, called the cμ/θ rule, for the problem of minimizing linear queue-length and reneging costs. A certain Skorohod map plays an important role in obtaining both uniqueness of solutions to the fluid-model equations and convergence.
16
A Proof for the Queuing Formula: <i>L</i> = λ<i>W</i>
John D. C. Little · Operations Research · 1961 · 2.7K citations
Heavy-Traffic Limits for Queues with Many Exponential Servers
S. Halfin, Ward Whitt · Operations Research · 1981 · 771 citations
A proof of the queueing formula: L=λW
Shang Zhi · Journal of Hebei Institute of Technology · 2001 · 356 citations