A simulation study of scalable broadcast in high-performance regular networks

A. Y. Al-Dubai, M. Ould-Khaoua, M. S. Obaidat

Research output: Contribution to journalArticle

Abstract

Broadcast is an important communication operation required by many real-world applications encountered in parallel, cluster, and grid computing environments. Broadcasting on regular networks has been widely investigated in the past. However, most of the existing algorithms handle broadcast in a sequential manner and do not scale well; as a consequence, many applications cannot be efficiently supported using existing algorithms. In an effort to avoid this limitation, this article presents a new broadcast algorithm based on coded path routing. In addition to its simplicity, the proposed algorithm has shown to be capable of performing the broadcast operation in a fixed number of message-passing steps, irrespective of the network size. An extensive simulation study has been conducted to evaluate the performance of the proposed algorithm under different traffic working conditions. The analysis reveals that the new algorithm exhibits superior performance characteristics over those of the well-known recursive-doubling and extended-dominating node algorithms.

Original languageEnglish
Pages (from-to)207-220
Number of pages14
JournalSimulation
Volume80
Issue number4-5
DOIs
Publication statusPublished - Apr 2004

Fingerprint

Broadcast
High Performance
Simulation Study
Cluster computing
Cluster Computing
Grid computing
Message passing
Grid Computing
Parallel processing systems
Doubling
Broadcasting
Message Passing
Parallel Computing
Real-world Applications
Simplicity
Routing
Traffic
Path
Evaluate
Communication

Keywords

  • Collective communication
  • Grid and cluster computing
  • Multicast latency
  • Performance analysis
  • Regular networks
  • Simulation

ASJC Scopus subject areas

  • Computer Science Applications
  • Computational Theory and Mathematics
  • Computer Graphics and Computer-Aided Design
  • Software
  • Safety, Risk, Reliability and Quality

Cite this

A simulation study of scalable broadcast in high-performance regular networks. / Al-Dubai, A. Y.; Ould-Khaoua, M.; Obaidat, M. S.

In: Simulation, Vol. 80, No. 4-5, 04.2004, p. 207-220.

Research output: Contribution to journalArticle

Al-Dubai, A. Y. ; Ould-Khaoua, M. ; Obaidat, M. S. / A simulation study of scalable broadcast in high-performance regular networks. In: Simulation. 2004 ; Vol. 80, No. 4-5. pp. 207-220.
@article{d3e2ee1ded754da9b1849ca7539b76dc,
title = "A simulation study of scalable broadcast in high-performance regular networks",
abstract = "Broadcast is an important communication operation required by many real-world applications encountered in parallel, cluster, and grid computing environments. Broadcasting on regular networks has been widely investigated in the past. However, most of the existing algorithms handle broadcast in a sequential manner and do not scale well; as a consequence, many applications cannot be efficiently supported using existing algorithms. In an effort to avoid this limitation, this article presents a new broadcast algorithm based on coded path routing. In addition to its simplicity, the proposed algorithm has shown to be capable of performing the broadcast operation in a fixed number of message-passing steps, irrespective of the network size. An extensive simulation study has been conducted to evaluate the performance of the proposed algorithm under different traffic working conditions. The analysis reveals that the new algorithm exhibits superior performance characteristics over those of the well-known recursive-doubling and extended-dominating node algorithms.",
keywords = "Collective communication, Grid and cluster computing, Multicast latency, Performance analysis, Regular networks, Simulation",
author = "Al-Dubai, {A. Y.} and M. Ould-Khaoua and Obaidat, {M. S.}",
year = "2004",
month = "4",
doi = "10.1177/0037549704044325",
language = "English",
volume = "80",
pages = "207--220",
journal = "Simulation",
issn = "0037-5497",
publisher = "SAGE Publications Ltd",
number = "4-5",

}

TY - JOUR

T1 - A simulation study of scalable broadcast in high-performance regular networks

AU - Al-Dubai, A. Y.

AU - Ould-Khaoua, M.

AU - Obaidat, M. S.

PY - 2004/4

Y1 - 2004/4

N2 - Broadcast is an important communication operation required by many real-world applications encountered in parallel, cluster, and grid computing environments. Broadcasting on regular networks has been widely investigated in the past. However, most of the existing algorithms handle broadcast in a sequential manner and do not scale well; as a consequence, many applications cannot be efficiently supported using existing algorithms. In an effort to avoid this limitation, this article presents a new broadcast algorithm based on coded path routing. In addition to its simplicity, the proposed algorithm has shown to be capable of performing the broadcast operation in a fixed number of message-passing steps, irrespective of the network size. An extensive simulation study has been conducted to evaluate the performance of the proposed algorithm under different traffic working conditions. The analysis reveals that the new algorithm exhibits superior performance characteristics over those of the well-known recursive-doubling and extended-dominating node algorithms.

AB - Broadcast is an important communication operation required by many real-world applications encountered in parallel, cluster, and grid computing environments. Broadcasting on regular networks has been widely investigated in the past. However, most of the existing algorithms handle broadcast in a sequential manner and do not scale well; as a consequence, many applications cannot be efficiently supported using existing algorithms. In an effort to avoid this limitation, this article presents a new broadcast algorithm based on coded path routing. In addition to its simplicity, the proposed algorithm has shown to be capable of performing the broadcast operation in a fixed number of message-passing steps, irrespective of the network size. An extensive simulation study has been conducted to evaluate the performance of the proposed algorithm under different traffic working conditions. The analysis reveals that the new algorithm exhibits superior performance characteristics over those of the well-known recursive-doubling and extended-dominating node algorithms.

KW - Collective communication

KW - Grid and cluster computing

KW - Multicast latency

KW - Performance analysis

KW - Regular networks

KW - Simulation

UR - http://www.scopus.com/inward/record.url?scp=3342962614&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=3342962614&partnerID=8YFLogxK

U2 - 10.1177/0037549704044325

DO - 10.1177/0037549704044325

M3 - Article

AN - SCOPUS:3342962614

VL - 80

SP - 207

EP - 220

JO - Simulation

JF - Simulation

SN - 0037-5497

IS - 4-5

ER -