中文摘要：

传统资源搜索方法在非结构化对等网络中查找资源时存在的主要问题是检索效率低和通信开销大,文中提出新的基于访问兴趣相似性的P2P网络模型。通过节点的行为特征所表现的访问频谱相似性寻找节点并形成集合,然后从中选择部分适合的对象建立少量的远程链接,可以改善传统的非结构化对等网络资源搜索性能。在选择访问频谱相似节点建立远程链接时,利用了高频访问节点的聚集特性而使工作更加容易。为实现该模型的资源检索,设计了相应的资源搜索算法。仿真结果表明,该模型在一定程度上提高了非结构化P2P资源搜索的效率,同时减少了网络中的通信冗余信息量。

英文摘要：

The low efficiency and large communication overhead are the main problems for traditional resource searching method looking for the resource in unstructured P2P network. To solve this problem,a new P2P network model based on interest similarity is given in this paper. The access spectrum similar nodes are found by node behavior and constitute a set of similar nodes. Through establishing the few remote connections among these nodes will improve the traditional unstructured P2P network resources searching performance. Using the high-frequency access node aggregation characteristic makes the work more easily when selecting and accessing the suitable spectrum similar nodes to establish the remote connection. Meanwhile, a new resource searching algorithm is designed for the model of resource re- trieval. The simulation results show that the model can improve the resources searching efficiency of unstructured P2P network, while re- ducing the information redundancy in the network.