A PARALLEL NETWORK SIMULATION PARTITIONING ALGORITHM IN THE COMPLEX PHYSICAL COMPUTING ENVIRONMENT
Traditional topology partition of parallel network simulation usually uses topology graph partition,that is to induce actual network simulate task to a weighted graph,Then use certain tools to complete the partition of simulation tasks,such as METIS,Chaco; These traditional tools usually perform very good when the simulation environment is of the same structure.However,in the complex physical computing environment,such as a large number of computing nodes whose calculation ability is different,a wide gap on the communication capabilities and so on,the traditional topology partition tools is difficult to achieve efficiency optimal,resulting in computing power a tremendous waste.Therefore,we introduce a Cluster Instructed Partitioning Algorithm (CIPA),suitable for partitioning simulation tasks for heterogeneous platform.In complex network environment,this method organizes the computing nodes by clustering,and distributes the simulated topology to the computing nodes properly combined with the improved topology graph partition.It can get a balanced partitioning result,and greatly improve the efficiency of the entire system.
Parallel network simulation Topology graph partition Partitioning algorithm Cluster analysis
Xiangzhan Yu Zhaoxin Zhang Dan Liu
School of Computer Science and Technology,Harbin Institute of Technology,Harbin 150001,China IBM China Design Center,Beijing,China
国际会议
杭州
英文
842-847
2012-10-30(万方平台首次上网日期,不代表论文的发表时间)