Inspired by the observation that most researches on cluster, grid, and cloud computing are intended to reduce the replication cost of files between client and server instead of minimizing the transmission time, this paper presents an efficient data replication mechanism for parallel download called DRMPD to minimize the average transmission time, by replicating the so-called popular files to multiple nodes that are more efficient for download while at the same time taking into account parallelism in download, especially when many applications nowadays can support parallel download. Simulation results indicate that the proposed scheme outperforms the previous schemes in terms of the average transmission time in both the best case and the average case.
關聯:
International Journal of Innovative Computing Information And Control, 8(10B), 7031-7053
