Abstract: Replication of data is an important method to improve the availability of applications in distributed data-intensive Grid environment. How to choose the replicas is a key factor to affect the performance of Data Grid. Within this paper, an ant algorithm for file replica selection in Data Grid is proposed and discussed. It is implemented and the advantages of the new ant algorithm are investigated in a grid simulator called OptorSim after the simulator has been expanded. The evaluation demonstrates that this new ant algorithm can reduce data access latency, decrease bandwidth consumption and distribute storage site load.

Key words: replica selection, ant algorithm, Data Grid, OptorSim

摘要： 在分布着大量数据和计算能力的数据网格环境中，采用数据副本是提高网格应用可用性的重要方法。如何对数据网格中大量的数据副本进行优化选择是影响数据网格性能的重要因素。因此提出一种基于蚂蚁算法的数据网格副本选择策略，并在网格仿真器OptorSim中对该算法进行实现及性能分析。仿真实验结果表明该算法可以减少数据访问延迟及带宽消耗，并有效做到网格中存储节点间的负载平衡。

关键词: 副本选择, 蚂蚁算法, 数据网格, OptorSim