Sojourn Times in Markov Queueing Networks: Little's Formula Revisited.

Abstract

It is commonly supposed that L = lambda W applies to 'almost any' queueing system, with lambda the mean customer entrance rate, L the asymptotic expectation of the number of customers in the system, and W the asymptotic sojourn time expectation. We study the formula for irreducible positive recurrent Markov queueing systems whose state vector Z consists of entries representing queue lengths at the respective service stations; such a model permits blocking, finite capacities, jockeying, state-dependent or random routing, bulk and/or Erlang service, and variable arrival and service rates. To define waiting times under various queueing disciplines, Z is augmented by a customer location process to yield the new Markov process Y = (Z,U). It is shown that the standard regenerative process proof of Little's equality fails in the absence of further hypotheses; however, additional assumptions assure the validity of L = lambda W for a broad variety of queueing disciplines.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jun 01, 1980
Accession Number
ADA089658

Entities

People

  • Frederick J. Beutler

Organizations

  • University of Michigan

Tags

Communities of Interest

  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Air Force
  • Communication Systems
  • Computer Communications
  • Computers
  • Construction
  • Engineering
  • Equations
  • Intervals
  • Markov Processes
  • Probabilistic Models
  • Probability
  • Probability Distributions
  • Random Variables
  • Scientific Research
  • Sequences
  • Standards
  • Transitions

Fields of Study

  • Mathematics

Readers

  • Mathematical Modeling and Probability Theory.