Data grid is an ideal platform for the data-intensive and computing-intensive computing, which coordinates resources sharing in large scale environment. Replica is an important key technology in data grid. It improves the efficiency of data access and reduces the delay of data transfer. Replica selection is to choose the most appropriate replica among nodes. A replica selection strategy based on ant algorithm is proposed within this study. And then implementation and performance analysis of the algorithm are studied in the grid simulator Optorsim. Experimental results show that the ant algorithm can reduce data access latency, reduce bandwidth consumption and distribute the load of storage site, improve data access speed.
展开▼