Error analysis for regenerative queueing estimators with special reference to gradient estimators via likelihood ratio

DOI10.1007/BF02094338MaRDI QIDQ1197804zbMATH OpenFDO

Authors Michael Shalmon, Reuven Y. Rubinstein

Publication date 16 January 1993

Published in Annals of Operations Research (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1007/BF02094338

zbMATH Keywords

analytical formulas for the variancequeueing modelsregenerative estimatorstraffic intensity

Mathematics Subject Classification ID

Queueing theory (aspects of probability theory) (60K25) Queues and service in operations research (90B22)

Review text

Excerpt from introduction: The main contribution of this paper is in obtaining analytical formulas for the variance of the regenerative estimators for the \(M/G/1\) queue, including those for the mean and (arbitrarily) higher order moments of the FCFS waiting time, and also those for the likelihood ratio estimator of the gradient of the moments of the FCFS waiting time with respect to either the service rate or the arrival rate. At arbitrary traffic intensity the formulas are recursive, while the heavy traffic approximations are explicit and of very simple form. While the formulas of this paper are strictly true only for the \(M/G/1\) queue, they can also be used as guide-lines for the design of simulations and of stochastic optimizations of more complex queueing models, particularly those with Poisson arrivals.

Recommendations

• Comparison of gradient estimation techniques for queues with non- identical servers
• On the consistency of second derivative perturbation analysis estimators for the M/G/1 queue
• Stochastic Optimization by Simulation: Convergence Proofs for the GI/G/1 Queue in Steady-State
• Conditioning for variance reduction in estimating the sensitivity of simulations
• Simulation methods of queues: An overview

Cites work

• Analysis of the GI/GI/1 Queue and its Variations via the LCFS Preemptive Resume Discipline and Its Random Walk Interpretation
• Perturbation analysis and optimization of queueing networks
• Perturbation Analysis Gives Strongly Consistent Sensitivity Estimates for the M/G/1 Queue
• Planning Queueing Simulations
• Queueing Simulation in Heavy Traffic
• Sensitivity Analysis for Simulations via Likelihood Ratios
• The total waiting time in a busy period of a stable single-server queue, II

This page was built for publication: Error analysis for regenerative queueing estimators with special reference to gradient estimators via likelihood ratio

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1197804)