Scheduling of scientific workflows using Threshold accepting algorithm for Computational Grids

Shajulin Benedict, V. Vasudevan

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

1 Scopus citations

Abstract

Grid computing environment involves all kind of resources namely network software, data, storage and processing units, evolving towards Global computing to solve a single large problem using Grid scheduling architecture that addresses the interaction between the resource management and data management In this paper, two different approaches have been proposed to solve Grid scheduling problem with the objectives of maximizing the Job completion ratio (JCR) and minimizing the lateness. A population based evolutionary algorithm that involves evolution during the search process and a single point local search meta-heuristics that work on a single solution called as hybrid evolutionary algorithm. A Threshold accepting algorithm (TA) proposed is a single point local search meta-heuristic. Proposed algorithms are evaluated and the experimental results are presented for comparison.

Original languageEnglish
Title of host publication2007 IEEE International Conference on Service Operations and Logistics, and Informatics, SOLI
DOIs
StatePublished - 2007
Externally publishedYes
Event2007 IEEE International Conference on Service Operations and Logistics, and Informatics, SOLI - Philadelphia, PA, United States
Duration: 27 Aug 200729 Aug 2007

Publication series

Name2007 IEEE International Conference on Service Operations and Logistics, and Informatics, SOLI

Conference

Conference2007 IEEE International Conference on Service Operations and Logistics, and Informatics, SOLI
Country/TerritoryUnited States
CityPhiladelphia, PA
Period27/08/0729/08/07

Keywords

  • Genetic algorithm
  • Grid computing
  • Scheduling
  • Threshold accepting

Fingerprint

Dive into the research topics of 'Scheduling of scientific workflows using Threshold accepting algorithm for Computational Grids'. Together they form a unique fingerprint.

Cite this