论文部分内容阅读
在P2P系统内每秒钟或许有成百上千个节点加入或离开系统,这种高动态性必然极大的增加DHT-based系统的维护开销,导致这种开销的主要原因是节点构建其路由信息时产生的开销。因此为了降低节点构建路由信息时带来的开销,提出了Chord系统的一种新的加入算法—JoinBP。JoinBP通过利用查询路径上节点的指针表信息来构建加入节点的指针信息,减少了节点在创建指针信息时需要的查询跳数和查询延迟,从而减少了构建finger时产生的开销。理论和模拟试验结果表明,JoinBP很大程度减少了系统的维护开销,同时提高了系统的查询性能。
In P2P systems, there may be hundreds or thousands of nodes joining or leaving the system every second. This high dynamic necessarily increases the maintenance overhead of DHT-based systems. The main reason for this overhead is that nodes construct their routes Information generated overhead. Therefore, in order to reduce the overhead of constructing routing information, a new join algorithm of Chord system is proposed, JoinBP. JoinBP uses the pointer table information of the nodes on the query path to construct the pointer information of the joined node, reducing the number of query hops and query delay required when creating pointer information, thus reducing the overhead incurred in constructing the finger. Theoretical and simulation results show that JoinBP greatly reduces system maintenance overhead and improves system query performance.