Modeling of priority-based request scheduling mechanism for finite buffer SIP servers

Demir Y. Yavas, Ibrahim Hokelek, Bilge Gunsel

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper, we present a fluid-flow model to characterize the behavior of the overloaded Session Initiation Protocol (SIP) server with a finite buffer. Our model includes a priority-based request scheduling mechanism (PRSM) which diminishes redundant retransmissions to overcome the overloading of SIP servers. The numerical results closely match with the simulation results, indicating that the proposed models can accurately capture the dynamic behavior of a SIP server with the PRSM. Using the Markov Modulated Poisson Process (MMPP) traffic model, the failure ratio of the PRSM is reduced to zero when the buffer size is higher than 1K while it linearly increases for the conventional SIP. The mean response delay of the PRSM is kept constant around 175 ms while it reaches as high as 179 seconds for the conventional SIP when the buffer size is around 200K.

Original languageEnglish
Title of host publicationProceedings of the 11th International Conference on Queueing Theory and Network Applications, QTNA 2016
PublisherAssociation for Computing Machinery
ISBN (Electronic)9781450348423
DOIs
Publication statusPublished - 13 Dec 2016
Event11th International Conference on Queueing Theory and Network Applications, QTNA 2016 - Wellington, New Zealand
Duration: 13 Dec 201615 Dec 2016

Publication series

NameACM International Conference Proceeding Series

Conference

Conference11th International Conference on Queueing Theory and Network Applications, QTNA 2016
Country/TerritoryNew Zealand
CityWellington
Period13/12/1615/12/16

Bibliographical note

Publisher Copyright:
© 2016 ACM.

Keywords

  • Fluidflow
  • Priority queuing
  • Server-overloading
  • SIP retransmission

Fingerprint

Dive into the research topics of 'Modeling of priority-based request scheduling mechanism for finite buffer SIP servers'. Together they form a unique fingerprint.

Cite this