Research On Task Scheduling Strategy Of Cloud Computing Based On Improved Particle Swarm Optimization

As a new business computing model, cloud computing is the product of integration and development of grid computing, parallel computing and utility computing. With virtualization technology cloud computing in the system makes use of resources to form a virtual shared pool. Then the users elicit various resources of services and application services from the resource pools through the Internet. Cloud computing in the new need of society and the demand of scale economies is intensively driven, which has the high commercial value and application value. Since cloud computing put forward, it has become the focus research object and attention of the commercial and scientific research institutions at home and abroad.In cloud computing environment, the users obtain the required application service from the resource pools based on "demand requests, volume pays" standards. Due to widespread user groups, different service of quality and dynamics and heterogeneity resources in the system, tasks scheduling of cloud computing becomes a NP hard problem. To the users, task execution time and execution cost are two mainly considered service of quality. Therefore, this paper put forward a task scheduling strategy of cloud computing on improved particle swarm optimization. In this paper, the main research content and the work done by including:First of all, it roundly and detailedly introduces the research status of cloud computing platform and cloud computing task scheduling at home and abroad. Then the related concepts of cloud computing and task scheduling is illustrated, and the solution of task scheduling with all kinds of algorithm is put forward to make certain summary and the elaboration. The algorithm includes traditional task scheduling algorithm and heuristic task scheduling algorithm.According to the commercial characteristics of cloud computing, cloud computing is a service mode based on the user as the center. So from the user’s point of view, the task execution time and execution cost of quality of service targets is considered based on these two goals to establish the mathematical model of task scheduling in cloud computing environment.Then considering the advantages and disadvantages of the particle swarm algorithm and genetic algorithm, an improved discrete particle swarm optimization algorithm is put forward. The realization of the algorithm is implemented through the join of the crossover operation and mutation operation of genetic algorithm based on the realization of the particle swarm optimization algorithm, which makes the particle swarm optimization algorithm solve the problem of task scheduling of such discretization.Finally, the improved discrete particle swarm optimization algorithm is used in cloud computing simulation platform CloudSim to simulate experiment simulator. Experiment will compare improved particle swarm algorithm comparing with the standard particle swarm optimization algorithm and genetic algorithm through the different size case overall task execution time and execution cost from two aspects of the experimental results, which proves that the improved discrete particle swarm algorithm has better performance, faster convergence speed.