基于行为特征的超级节点节流算法研究

针对P2P网络中超级节点失效时带来的资源流失、网络拓扑结构变化和重新选举网络开销增加等问题，提出了一种基于用户行为特征统计的超级节点禅让算法。根据节点的失效统计特征预估失效时间，预先指定继任超级节点。仿真实验对比结果表明，该算法可以有效降低超级节点失效时带来的网络波动，降低网络流量消耗。

Research on algorithm for low bandwidth super-peers network based on user’s behavior characteristic

In a super-peers-based P2P network, if super-peers fail or leave, it may bring many problems, such as resource losing, network topology changing and increasing bandwidth occupancy for re-election. To solve these problem, this paper brings up a super-peers abdicate algorithm based on user’s behavior characteristic. According to statistic characteristic of the node failure, the SPAA algorithm can forecast the node leave time and appoint next super-peer. Simulation and analysis show that the SPAA algorithm can effectively reduce the net churn by super-peer failure and reduce network traffic.