Publication Details

An Evolutionary Design Technique for Collective Communications on Optimal Diameter-Degree Networks

JAROŠ, J.; DVOŘÁK, V. An Evolutionary Design Technique for Collective Communications on Optimal Diameter-Degree Networks. In 2008 Genetic and Evolutionary Computational Conference GECCO. New York: Association for Computing Machinery, 2008. p. 1539-1546. ISBN: 978-1-60558-131-6.
Czech title
Evoluční technika návrhu kolektivních komunikačních plánů na propojovacích sítích s optimálním poměrem průměr-stupeň uzlu
Type
conference paper
Language
English
Authors
Jaroš Jiří, doc. Ing., Ph.D. (DCSY)
Dvořák Václav, prof. Ing., DrSc.
Keywords

Collective communications, communication scheduling, evolutio-nary design, interconnection networks.

Abstract

Scheduling collective communications (CC) in networks based on optimal graphs and digraphs has been done with the use of the evolutionary techniques. Inter-node communication patterns scheduled in the minimum number of time slots have been obtained. Numerical values of communication times derived for illustration can be used to estimate speedup of typical applications that use CC frequently. The results show that evolutionary techniques often lead to ultimate scheduling of CC that reaches theoretical bounds on the number of steps. Analysis of fault tolerance by the same techniques revealed graceful CC performance degradation for a single link fault. Once the faulty link is located, CC can be re-scheduled during a recovery period.

Published
2008
Pages
1539–1546
Proceedings
2008 Genetic and Evolutionary Computational Conference GECCO
ISBN
978-1-60558-131-6
Publisher
Association for Computing Machinery
Place
New York
UT WoS
000268226900373
EID Scopus
BibTeX
@inproceedings{BUT30492,
  author="Jiří {Jaroš} and Václav {Dvořák}",
  title="An Evolutionary Design Technique for Collective Communications on Optimal Diameter-Degree Networks",
  booktitle="2008 Genetic and Evolutionary Computational Conference GECCO",
  year="2008",
  pages="1539--1546",
  publisher="Association for Computing Machinery",
  address="New York",
  isbn="978-1-60558-131-6",
  url="https://www.fit.vut.cz/research/publication/8658/"
}
Files
Back to top