Task scheduling in Cloud Computing Based on Improved Discrete Particle Swarm Optimization

Cloud computing is a technology that allows users to pay as the demand and offers various kind of services just like using water, and how to efficiently schedule the tasks is an important aspect in cloud. In this paper, an improved discrete particle swarm optimization algorithm abbreviated as IDPSO is implemented in the cloud simulation platform Cloudsim. IDPSO uses a dynamic inertia weight optimization method based on sinusoidal strategy to make particles adaptive to different stages in the process of searching for global optimal solution. The simulation results show that the proposed algorithm outperforms Discrete Particle Swarm Optimization(DPSO) and the basic scheduling algorithm First Come First Serve(FCFS) in terms of completion time and converges better than DPSO.