Structured P2P Search Method to Support Complex Query

It is difficult for the structured P2P (Peer-to-Peer) systems to support complex query such as multi-keyword query and range query, although they are scalable to lookup data objects with unique keys. In this paper, we propose a new method called multidimensional semantic query (MSQ), which provides efficient multi-keywords query and searching mechanics for structured P2P systems to overcome the shortcomings of the existing multi-keyword query (MKQ). The vital idea of MSQ is to apply the space filtering curves to cluster similar data objects in high dimension semantic space to an adjacent range in one dimensional value space. The multidimensional semantic query model, the construction process of Hilbert space curves, and the process of data object′s publication and query are presented in detail. The simulated experimental results show that MSQ is better than MKQs to improve the query accuracy and the number of communication messages. Furthermore the theory analysis also shows that MSQ has capacity to hold load balance and fault tolerance.