An efficient multi-keyword search mechanism in structured P2P networks

Structured P2P networks are efficient for searching by a single keyword. However, it poses a challenge to support multiple-keyword search. Resolving of a multiple-keyword search in structured P2P networks needs a large number of hops and generates a large number of messages when using traditional method-distributing inverted indices among peers. Current researches on multiple-keyword search focus on the decrease of number of hops and number of messages, but no good solution is put forward by far. In this paper we propose a multiple-keyword search mechanism in structured P2P networks. This mechanism groups peers with similar searches together. The formation and maintenance of groups are in a self-organizing fashion. Peers in the same group share their cached results. Simulation results show this mechanism outperforms current multiple-keyword search mechanisms in both the number of hops and the number of messages.