A Two-Fitness Resource Scheduling Strategy Based on Improved Particle Swarm Optimization

Abstract

The performance optimization of cloud platform for big data processing is a research hotspot, among which resource scheduling is the most important. Through the analysis of the internal resource scheduling mechanism of CloudStack, the two-level scheduling of resources plays an important role in task optimal span, load balance and other aspects. In this paper, aiming at optimizing IaaS service performance and taking CloudStack platform as the research object, a dual fitness resource scheduling strategy based on improved particle swarm optimization is proposed. First of all, PSO algorithm with high precision and fast convergence speed is used to optimize the two-level resource scheduling, which can shorten the scheduling time when the scheduling requirements are met. Secondly, aiming at the problem of “prematurity” of particle swarm optimization (PSO), this paper USES simulated annealing algorithm to optimize the traditional PSO. Finally, aiming at the two pole resource scheduling, this paper proposes the virtual machine deployment algorithm based on improved particle swarm and the dual fitness task scheduling algorithm based on Improved Particle Swarm respectively, and carries out simulation in CloudSim simulation tool. The simulation results show that the algorithm proposed in this paper can effectively improve the optimal span and optimize the load balance.