Title: A hybrid particle swarm optimisation-genetic algorithm applied to grid scheduling

Authors: Wilson A. Higashino; Miriam A.M. Capretz; M. Beatriz F. De Toledo; Luiz F. Bittencourt

Addresses: Department of Electrical and Computer Engineering, Western University, London, Ontario, Canada; Institute of Computing, University of Campinas, Campinas, SP, Brazil ' Department of Electrical and Computer Engineering, Western University, London, Ontario, Canada ' Institute of Computing, University of Campinas, Campinas, SP, Brazil ' Institute of Computing, University of Campinas, Campinas, SP, Brazil

Abstract: Scheduling problems have been thoroughly explored by the research community, but they acquire challenging characteristics in grid computing systems. In this context, it is important to have a scheduling strategy that can make efficient use of the available grid resources. This article focuses on the application of the particle swarm optimisation (PSO) meta-heuristic to the scheduling of independent users' jobs on grids. It is shown that the PSO method can achieve satisfactory results in simple problem instances, yet it has a tendency to stagnate around local minima in high-dimensional problems. Therefore, this research also proposes a novel hybrid particle swarm optimisation-genetic algorithm (H_PSO) method that aims to increase swarm diversity when a stagnation condition is detected. This new method is evaluated and compared with other heuristics and PSO formulations; the comparison shows that H_PSO can successfully improve the scheduling solution.

Keywords: PSO; particle swarm optimisation; grid scheduling; genetic algorithms; metaheuristics; grid computing; swarm diversity; stagnation.

DOI: 10.1504/IJGUC.2016.077493

International Journal of Grid and Utility Computing, 2016 Vol.7 No.2, pp.113 - 129

Received: 31 Oct 2014
Accepted: 03 Feb 2015
Published online: 04 Jul 2016 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article