28623 articles – 22140 references  [version française]

inria-00070775, version 1

M/G/1/MLPS compared to M/G/1/PS

Samuli Aalto 1, Urtzi Ayesta, Eeva Nyberg-Oksanen

N° RR-5219 (2004)

Abstract: Multilevel Procesor Sharing scheduling disciplines have recently been resurrected in papers that focus on the differentiation between short and long TCP flows in the Internet. We prove that, for M/G/1 queues, such disciplines are better than the Processor Sharing discipline with respect to the mean delay whenever the hazard rate of the service time distribution is decreasing.

  • 1:  MAESTRO (INRIA Sophia Antipolis)
  • INRIA – Université Montpellier II - Sciences et techniques
  • Domain : Computer Science/Other
  • Keywords : SCHEDULING / MULTILEVEL PROCESSOR SHARING / MLPS / MEAN DELAY / UNFINISHED TRUNCATED WORK / M/G/1
  • Internal note : RR-5219
 
  • inria-00070775, version 1
  • http://hal.inria.fr/inria-00070775
  • oai:hal.inria.fr:inria-00070775
  • From: Rapport De Recherche Inria <>
  • Submitted on: Friday, 19 May 2006 21:35:57
  • Updated on: Wednesday, 31 May 2006 14:24:25