Study On Resource Allocation Algorithms In Cloud Computing Environment

With the innovation and development of computer technology, cloud computinghas appeared. It is a new model of business computing, that brings a lot of infrastructureresources together to realize on-demand services of the data central resources with themature virtualization technology. In cloud computing, as the resources are dynamic,heterogeneous, and large-scale, it becomes an urgent issue to develop an appropriateresource allocation policy based the actual characteristics of cloud computing. Atpresent, intelligent optimization algorithm is the main method to solve this problem.This article has studied on the resource distribution under the cloud computing, andanalyzed the problems of existing resource allocation algorithms. The detailed work ofthis paper includes the following aspects:Firstly, Propose a cloud computing resource-allocation algorithm combinedParticle Swarm Optimization and Genetic Algorithm (PSO-GA). Both traditionalParticle Swarm Optimization algorithm and Genetic Algorithm are likely to becomeembroiled in premature convergence in resource allocation processes. In order to solvethis problem, a PSO-GA resource allocation algorithm has been proposed. Thealgorithm introduces Population Division and Population Cover based on the GeneticAlgorithm, and applies the mutation operator of Particle Swarm Optimization into themutation process of PSO-GA algorithm. Results show that, the PSO-GA algorithm cansolve the problem of premature convergence effectively, and it has improved theconvergence rate of optimal solution and execution efficiency.Secondly, Propose a cloud computing resource-allocation algorithm based onImproved Artificial Fish_swarm Algorithm (IAFA). In the resource allocation process ofcloud computing, if the population size of the PSO-GA algorithm is larger, then itsconvergence rate will be slow and can’t get global optimal solution quickly. In order tosolve this problem, an Improved Artificial Fish_swarm Algorithm (IAFA) has beenproposed in this paper. The algorithm has eliminated the stochastic behavior on the basisof original behaviors and introduced the jump behavior which can urge an artificial fishjump out of the partial extreme value to search for global optimization. The algorithmhas introduced the concepts of life cycle and survival index to save storage space andimprove the efficiency of algorithm. Results show that, IAFA algorithm can get theglobal optimal solution with rapid convergence in the case of large population size. Thirdly, Expand the cloud computing simulation platform CloudSim, and carry onthe simulation with the algorithm proposed in preceding text. The paper has studiedCloudSim resource allocation mechanism, recompiled CloudSim platform, realizedresource allocation simulation program of PSO-GA algorithm、IAFA algorithm andintelligent optimization algorithm on the CloudSim platform, experiments prove thevalidity of these two improved algorithms.