Research On Improved Particle Swarm Optimization Algorithm Based On Cloudsim In Cloud Computing

Cloud computing is distributed computing, parallel computing, grid computing, virtualization(Virtualization), load balancing, network storage and other traditional computer and network technology integration products, it is based on new technology networks. Cloud computing resource pool performance by a number of different resource nodes, with the growing user demand, how to assign tasks to the limited scale of resource nodes, enhance customer satisfaction, achieve load balancing, cloud computing is an important need to study problem.Efficient task scheduling is an effective step to play an important cloud computing potential. In cloud computing, some of the tasks that need to be assigned to different virtual machines to improve system utilization and minimize completion time. Task scheduling problem is NP-complete problem, so to find a definitive solution is tricky, especially for large-scale task.So this paper proposes a dynamic update inertia weight particle swarm optimization algorithm based on fitness, and particle swarm particle re-coded to discrete, the improved algorithm is applied to cloud computing task scheduling problem among.This article also details the subsequent cloud simulation tool CloudSim, and configure the test environment. On the simulation platform for discrete particle swarm optimization(DPSO) and improved DPSO simulation experiments. Through DPSO algorithm, greedy strategy, order scheduling method comparing simulation results show that the cloud computing task scheduling problem, a small amount of task scheduling, improved our algorithm is not better than the other allocation algorithms, gradually increase the amount of the task, we found the size of the Great, DPSO convergence faster, the advantages compared to other scheduling algorithm is more obvious. Its simulation show that the algorithm for task scheduling problem is feasible and effective, especially for the massive task.