针对P2P网络的聚类性以及网络的逻辑拓扑与物理拓扑不匹配所造成的通信延迟增加、网络开销增大等问题,根据节点的物理位置及节点所包含信息的语义相似度,提出了一种3层结构的语义对等网模型。该模型基于节点物理位置与属性特征动态调整网络结构,自适应地改善自身的资源搜索性能。仿真结果显示,该P2P覆盖网能够以较低的查找时延和代价获得较高的查全率。

In view of the clustering of P2P networks and the issues of increased communication latency and network costs caused by mismatch between the logical topology and the physical topology, this paper proposes a mixed three-layer P2P network model. The model adjusts the network structure dynamically based on the position and the property features of nodes, so that the performance of searching resources can be improved adaptively. The simulation result shows that the P2P networks can achieve a high recall ratio with low search delay and cost.