Regenerative Simulation of Response Times in Networks of Queues: Statistical Efficiency.

Abstract

In this report the calculation of variance constants entering into central limit theorems used to obtain confidence intervals from passage time simulations is considered. Using results of Hordijk, Iglehart, and Schassberger for the calculation of moments in discrete time and continuous time Markov chains, we calculate variance constants pertinent to mean passage times. This is done first for the 'marked job method' for passsage time simulation which is based on the tracking of a distinguished job, and then for the 'decomposition method' in which observed passage times for all of the jobs enter into the construction of point and interval estimates. The results of this paper provide a means of comparing the statistical efficiency of the two estimation methods.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 1979
Accession Number
ADA075174

Entities

People

  • Donald Iglehart
  • Gerald S. Shedler

Organizations

  • Stanford University

Tags

DTIC Thesaurus Topics

  • Construction
  • Decomposition
  • Efficiency
  • Equations
  • Identities
  • Intervals
  • Markov Chains
  • Operations Research
  • Probability
  • Random Variables
  • Security
  • Simulations
  • Simulators
  • Statistical Estimation
  • Stochastic Processes
  • Transitions
  • United States

Fields of Study

  • Mathematics

Readers

  • Mathematical Modeling and Probability Theory.
  • Statistical inference.