网格中执行SLA的并行多群体协作PSO框架

Parallel Multi-swarm Cooperative PSO Framework of Implemented SLA in Grid

针对广域范围内的计算资源加入粒子群优化(PSO)执行从而提高计算效率并降低计算成本,提出一个网格环境下执行SLA的并行多群体协作PSO框架(PMCF)。给出一个适应网格环境的并行多群体协作PSO(PMCPSO)算法,基于WS-Agreement研究了利用网格技术和PMCPSO算法设计PMCF,把一个元任务调度器用于无缝的资源发现、选取及协商服务质量,通过模拟试验对PMCF进行了评估。

For joining the computational resources in large field into the implementation of PSO to enhance the computational efficiency and reduce the computational cost, a Parallel Multi-swarm Cooperative PSO Framework（PMCF） of Service Level Agreement（SLA） implementing is presented in this paper. A Parallel Multi-swarm Cooperative PSO（PMCPSO） algorithm adapted to Grid environment is proposed. Based on WS-Agreement, how to make use of Grid technologies and PMCPSO algorithm to design PMCF becomes the emphasis of research. PMCF has a meta-scheduler for the seamless resource discovery, selection and negotiation of QoS. The performance of PMCF is evaluated by the experiment.