Improved PSO for Task Scheduling in Cloud Computing

Abstract

An improved Particle Swarm Optimization (PSO) is used for performing task scheduling in cloud computing with the aim of distributing uniform load on each Virtual Machine (VM) in a datacenter. It is achieved using an objective function which tries to enhance the candidate solutions iteratively and thus finds an optimal mapping of task set to VM set. Experimental results have shown that the improved PSO performs better than the original PSO by maintaining the consistency in scheduling.