A Compact Table-Based Search Mechanism for Peer-to-Peer Systems
Jing Zheng · Computer Engineering and Science · 2006
Recently unstructured P2P systems have been deployed widely on the Internet. However, current search techniques used in the existing Gnutella-like peer-to-peer systems are often very inefficient. In this paper we propose the CSTM mechanism to improve the performance and scalability of such systems. In CSTM, each peer maintains a local state table containing the keyword information of data on all neighbors within T hops to guide query, and new peers joining in the system select their neighbors by the degree of similarity. CSTM adopts a query cache to utilize query locality and build shortcut connections to lately accessed peers. Simulations show that CSTM can reduce message cost one order of magnitude remarkably while maintaining short search path lengths compared with flooding or random forwarding algorithms.