基于反馈的非结构化P2P网络混合搜索算法

Feedback-based hybrid search algorithm in unstructured peer-to-peer networks

为了提高P2P网络搜索效率和有效地减少冗余消息,提出一种基于反馈的混合搜索算法,将查询消息在某些高度数节点处转发给多个邻居节点而在其他节点处只转发给一个邻居节点,并且在搜索过程中利用先前搜索的反馈信息来指导搜索,以选择能够返回最多最近响应结果的邻居节点转发消息。算法分析和实验结果表明,该算法不仅实现了较高的成功率、响应率和搜索效率,还能有效地减少冗余消息,并且具有良好的自适应能力,能快速适应不断变化的网络行为。

In order to improve search efficiency and reduce redundant messages efficiently in P2P networks, this paper proposed a feedback-based hybrid search algorithm. It forwarded the query messages to a subset of neighbors at some high-degree nodes and to only ones at other nodes, and utilized the feedback from the previous searches to guide current ones in the searching process to select those neighbors that can return the most and nearest results to query. The algorithm analysis and experimental results show that this algorithm not only achieves higher success rate, query hits and search efficiency, but also reduces redundant messages efficiently. Furthermore, it has a better self-adaptive ability to rapidly adapt the dynamical hanging network behaviors.