【摘要】 结构化P2P网络虽然具有扩展性良好的数据查找机制,但只支持基于键的准确匹配搜索。为提供更丰富的数据查询能力,本文提出一种基于主题重叠网络的结构化P2P搜索算法——主题重叠网络搜索算法(TONS)。其基本思想是在结构化P2P网络之上,将结点按主题组织成分层的重叠网络,使含有相似主题的结点相互链接在一起;利用主题中继结点所具有的全局导航能力,TONS能够基于内容将查询限定在P2P网络的局部范围内,并且通过在重叠网络中随机添加一些长距离链接,使重叠网络具有Small-World特性,改善TONS的搜索性能。实验结果表明,TONS大大提高了搜索的查全率,减少了P2P网络信息搜索时的平均路径距离和平均消息数目。更多还原

【Abstract】 Structured P2P systems are not suitable for complex queries because they only support keyword-based and exact-match lookup. To provide complex queries in DHT-based P2P system, a topic overlay networks search algorithm (TONS)is proposed in this paper. The basic idea is to construct a hierarchy topic overlay network on top of the existing structured P2P network at first. The novel overlay network can be generated by dynamically clustering peer nodes that include relevant topic document set and linking them together. And then with the global navigation capacity global navigation of the overlay network, TONS can restrict a query in partial nodes of P2P network and achieves a very competitive trade-off between the search latencies/bandwidth and maintenance overheads. Furthermore, to improve the search performance, some long distance linkages are randomly added to the topic overlay networks to organize them into a small-world networks. The experiment results showed that TONS can reduce network bandwidth, improve the search efficiency and precisions.更多还原