Multiple-Objective Particle Swarm Optimization Algorithm for Independent Task Scheduling in Distributed Heterogeneous Computing Systems

Task scheduling is a crucial issue in distributed (disbursed) heterogeneous processing environment and significantly influence the performance of the system. The task scheduling problem has been identified to be NP-complete in its universal frame. In this paper the task scheduling problem is investigated using multiple-objective particle (molecule) swarm optimization algorithm with crowded displacement operator (MOPSO-CD). Particle swarm optimization is a populace based meta-heuristic which mimics the convivial conduct of feathered creatures running. In this algorithm particles move in the problem's search space to achieve near optimal solutions. The performance of this algorithm is compared with non-domination sorting genetic algorithm II (NSGA-II). The proposed scheduling algorithm intends to find the near optimal solution with aim to minimize the make-span and flow time. The exploratory results demonstrate that the proposed multi-objective PSO algorithm is more productive and gives better outcomes when contrasted with those of NSGA-II. KeywordsTask scheduling, Independent tasks, Meta heuristic, Particle swarm intelligence, Nondomination sorting genetic algorithm, Make-span, Flow-time.