高级检索

    P2P网络环境下的一种高效搜索算法:Multilayer Light-Gossip

    Multilayer Light-Gossip—An Efficient Search Algorithm in Peer-to-Peer Networks

    • 摘要: 由于现有非结构化P2P网络路由协议均在应用层实现,缺乏缓存机制和对Internet底层通信子网路由资源的利用,存在可扩展性差和效率不高的问题.在基于层域结构的RLP2P网络环境下,将路由空间分为域间和域内两层,结合泛洪和生成树搜索方式的优点,提出并实现了一种Multilayer Light-Gossip分级搜索算法和域间基于正六边形的蜂窝路由探测策略,把网络中的搜索消息分为域间和域内扩散两类分级扩散,以一定的消息冗余保持网络的稳健性和搜索的有效性,使定位某种服务的工作量和查询范围从网络中的所有结点数降到域内的节点数.运用预测评估方法对级间路由消息进行预分组,使消息能够自适应地沿着一条在时间度量上距离尽量短的路径前进.实验结果表明,Multilayer Light-Gossip算法大幅提高搜索效率和减少冗余消息,在广域环境下具有良好的搜索性能和扩展性.

       

      Abstract: In existing unstructured P2P systems, the application layer protocol simply uses flooding algorithm to route peer's querying and the lack of cache scheme, which is just implemented on application layer and doesn't use down-layer's information routing of Internet. So it has poor scalability and low efficiency. In the RLP2P network based on the architecture of layer-region, the routing space is divided into two layers, the on region layer and the in region layer, advantages of both the flooding distributed forward search and spanning tree algorithm are combined, and a new efficient search algorithm of multilayer light-gossip and routing strategy of beehive based on regular hexagon in region is proposed. First, it classifies the message of the network into diffusions based on region and in region, keeps the robust and validity searched for the networks with certain redundant messages, and makes the workload of locating service and all the network host-count of range query bring down to regions. Second, it adopts the forecast evaluation means realize to route messages that are divided into groups in advance, enable the route messages to forward automatically along a shortest path on time measurement. The simulations results show that multilayer light-gossip algorithm improve search efficiently greatly and lowers redundant messages so that it enables the whole comprehensive performance of the networks maintain a fine state and scalability under the environment of wide area.

       

    /

    返回文章
    返回