In the Data Grid environment, the primary goal of data replication is to shorten the data access time experienced by the job. Nowadays, scientific research and commercial application generate large amount of data that are required by users around the world. Data access time depends on bandwidth and storage space in each site. The major bottleneck to supporting fast data access in Grids is the high latencies of Wide Area Networks and Internet. We proved that, on the one hand high bandwidth and storage space make the enhancement of scheduling efficiency in data grids and on the other hand scheduling and data replication strategies with the high bandwidth and high storage space equally treat and have the same results. In this paper, we studied LFU, LRU, BHR and HRS algorithms and evaluated them through simulation by Gridsim.