Modelling of pipelined circuit switching in multicomputer networks

Geyong Min*, Hamid Sarbazi-Azad, Mohamed Ould-Khaoua

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

Several recent studies have revealed that pipelined circuit switching (PCS) can provide superior performance characteristics over wormhole routing. This paper proposes an original analytical model of PCS in k-ary n-cube networks augmented with virtual channel support. The model uses random walk theory to analyze the backtracking actions of the header flit during the path setup phase in PCS, and M/G/I queueing systems to compute the mean waiting time that a message experiences at a source before entering the network. Results from simulation experiments show close agreement to those predicted by the model.

Original languageEnglish
Title of host publicationProceedings - IEEE Computer Society's Annual International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunications Systems, MASCOTS
PublisherIEEE Computer Society
Pages299-306
Number of pages8
Volume2000-January
ISBN (Print)076950728X
DOIs
Publication statusPublished - 2000
Event8th International Symposium on Modeling, Analysis and Simulation of Computer and Telecommunication Systems, MASCOTS 2000 - San Francisco, United States
Duration: Aug 29 2000Sept 1 2000

Other

Other8th International Symposium on Modeling, Analysis and Simulation of Computer and Telecommunication Systems, MASCOTS 2000
Country/TerritoryUnited States
CitySan Francisco
Period8/29/009/1/00

Keywords

  • Adaptive routing
  • K-ary N-cubes
  • Message latency
  • Multicomputer
  • Performance analysis
  • Queueing theory
  • Virtual channels

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Computer Networks and Communications
  • Software
  • Modelling and Simulation

Fingerprint

Dive into the research topics of 'Modelling of pipelined circuit switching in multicomputer networks'. Together they form a unique fingerprint.

Cite this