Limiting search cost distribution for the move-to-front rule with random request probabilities

Javiera Barrera, Thierry Huillet, Christian Paroissin

Research output: Contribution to journalArticlepeer-review

4 Scopus citations

Abstract

Consider a list of n files whose popularities are random. The list is updated according to the move-to-front rule. When the induced Markov chain is at equilibrium, we explicitly compute the limiting distribution of the search-cost per item as n tends to infinity. The uniform distribution results in the largest search cost.

Original languageEnglish
Pages (from-to)557-563
Number of pages7
JournalOperations Research Letters
Volume34
Issue number5
DOIs
StatePublished - Sep 2006

Keywords

  • Limiting distribution
  • Move-to-front
  • Random discrete distribution
  • Search cost
  • Size-biased permutation

Fingerprint

Dive into the research topics of 'Limiting search cost distribution for the move-to-front rule with random request probabilities'. Together they form a unique fingerprint.

Cite this