Independent Tasks Scheduling using Parallel PSO in Multiprocessor Systems

Abstract

Multiprocessor system often aims to minimize the schedule length of the submitted jobs. For this, efficient scheduling of the parallel tasks on multiprocessor system is required. As the scheduling is an NP-Hard problem, meta-heuristics are often applied for this. This work applies parallel particle swarm optimization technique for multiprocessor task scheduling. As the scheduler designed using parallel PSO itself can execute in parallel on the multiprocessor systems the convergence is faster. The proposed algorithm takes care of local as well as global convergence. The performance evaluation of the proposed model is done by simulation and the result is quite encouraging.