Statistical Analysis of Multiserver Queueing Simulations

Abstract

This paper presents methods for computing point and interval estimates of descriptors of simulated queueing systems. The methods rely on the renewal process representation of certain queueing models. Two suggested procedures combine sample state times and job completion counts data to produce estimates with desirable statistical properties. One procedure uses linear combinations of the data to derive minimum length interval estimates. The other derives the maximum likelihood estimates of three descriptors (transformations of which can be used to give maximum likelihood esti mates of a variety of other system descriptors). The paper also describes how to extend these results to other queueing systems and illustrates the approach for a queueing system with finite waiting room.