How nearly do arriving customers see time-average behavior?

Erol A. Peköz, Sheldon M. Ross, Sridhar Seshadri

Research output: Contribution to journalArticlepeer-review

Abstract

Customers arriving at a queue do not usually see its time-average behavior unless arrivals occur according to a Poisson process. In this article we study how nearly customers see time-average behavior. We give total variation error bounds for comparing the distance between the time- and customer-average distributions of a queueing system in terms of properties of the interarrival distribution. Some refinements are given for special cases and numerical computations are used to demonstrate the performance of the inequalities.

Original languageEnglish (US)
Pages (from-to)963-971
Number of pages9
JournalJournal of Applied Probability
Volume45
Issue number4
DOIs
StatePublished - Dec 2008
Externally publishedYes

Keywords

  • Arrival theorem
  • Arrivals that see time averages
  • PASTA

ASJC Scopus subject areas

  • Statistics and Probability
  • Mathematics(all)
  • Statistics, Probability and Uncertainty

Fingerprint Dive into the research topics of 'How nearly do arriving customers see time-average behavior?'. Together they form a unique fingerprint.

Cite this