On Performance Comparisons of GA, PSO and proposed Improved PSO for Job Scheduling in Multiprocessor Architecture

Job Scheduling in a Multiprocessor architecture is an extremely difficult NP hard problem, because it requires a large combinatorial search space and also precedence constraints between the processes. For the effective utilization of multiprocessor system, efficient assignment and scheduling of jobs is more important. This paper proposes a new improved Particle Swarm Optimization (ImPSO) algorithm for the job scheduling in multiprocessor architecture in order to reduce the waiting time and finishing time of the process under consideration. In the Improved PSO, the movement of a particle is governed by three behaviors, namely, inertia, cognitive, and social. The cognitive behavior helps the particle to remember its previous visited best position. This paper proposes to split the cognitive behavior into two sections .This modification helps the particle to search the target very effectively. The proposed ImPSO algorithm is discussed in detail and results are shown considering different number of processes and also the performance results are compared with the other heuristic optimization techniques Genetic Algorithm and Particle Swarm Optimization.