Task Scheduling Using the PSO-IWD Hybrid Algorithm in Cloud Computing with Heterogeneous Resources

Optimal Task Scheduling is one of the most important challenges for achieving high performance in distributed environments such as cloud computing. The primary purpose of task scheduling is to allocate tasks to resources so that some of the system performance metrics will be optimized such as runtime or parallelism. Task scheduling is an NP-complete problem, so heuristic or metha-heuristic algorithms are used to solve it. Because cloud providers offer computing resources based on the pay-as-you-go model, the scheduling algorithm affects the users cost of the cloud. In this paper, a new cloud task scheduling algorithm based on particle swarm optimization as a metha-heuristic method is proposed that assigns users tasks to free resources in cloud computing environments. To enhance the convergence rate of the particle swarm optimization method, the intelligent water drops algorithm is applied. The results of this algorithm on random graphs showed a significant improvement in the performance of the proposed method compared to other task scheduling algorithms.