P2P网络中节点加入和退出的随意性,影响着文件的可用性；为达到满意的性能,P2P社区必须对文件进行复制或替换.由此,提出引入遗传算法解决P2P文件可用性优化问题,它适合高效的、自适应的文件复制.为在组内成员节点之间复制源文件,将文件划分成大小相等的小片,依据链按时选择该节点的概率,将小片传送到成员节点；然后,每个节点向组内剩下的其它节点复制传播.该方法减少了文件复制的时间,提高了文件可用性.%The availability of a P2P file sharing system is heavily affected by the high churn rate of users. Since nodes are intermittently connected, the P2P community must replicate and replace files to achieve satisfactory per-formance. The Genetic Algorithms is applied in optimizing file availability in P2P files delivery. This algorithm is suitable for an efficient and adaptive replication of files. In order to replicate an original file among group member nodes, the file is divided into small chips of equal size, and each chip is transferred to a member node in proportion to the probability of selecting a link. Then, each node propagates it to the remaining nodes in the group. This method achieves reduced file replication time, and improves the files availability.
展开▼
