Improved File Search And Sharing Mechanism In Structured P2P Systems

— File querying is important functionality which indicates the performance of p2p system. To improve file query performance, cluster the common interested peers based on physical proximity. Existing methods are dedicated to only unstructured p2p systems and they don’t have strict policy for topology construction which decreases the file location efficiency. In this project, we propose a structured p2p file sharing system which is clustered and aware of proximity. It forms a cluster based on node proximity as well as a sub-cluster with the nodes having common interest. A novel lookup function named as DHT and file replication algorithm which supports efficient file lookup and access is utilized. To reduce the overheads and file searching delays, the file querying may become inefficient due to the sub-interest super node overload/failure. Even though the sub-interest based file querying improves querying efficiency, it is still not sufficiently scalable when there are a very large number of nodes in a sub-interest group. We then propose a distributed intra-sub-cluster file querying method in order to further improve the file querying efficiency.