Comparative Study of Some Nature-Inspired Meta-Heuristics for Task Scheduling in a Computational Grid System

Abstract

Grid computing has emerged as an intelligent distributed computing paradigm due to the huge improvements in performance of wide-area network and powerful yet low-cost computers. Computational grids accumulate and share the power of heterogeneous, geographically dispersed, and multi-domain-administered computational resources to offer high performance or high-throughput computing. To realize the promising potential of computational grids, an effective and efficient task scheduling system is primarily important. Task scheduling in computational grid is one of the most challenging and complex tasks. In other words, the task scheduling in computational Grid is considered as NP-hard problem due to the problem complexity and intractable nature of the problem. Such a problem could be solved using meta-heuristic algorithms. In this chapter, several nature-inspired meta-heuristics are compared with respect to the parameters (i.e., minimization of makespan and flowtime) for scheduling tasks in computational grids. The nature-inspired meta-heuristics involved in this chapter are genetic algorithm (GA), particle swarm optimization (PSO), ant colony optimization (ACO), and cuckoo search (CS) algorithms. Experimental results show that the GA outperforms the other methods in terms of average makespan and the PSO algorithm performs best among all 4 algorithms in terms of average flowtime.