Author : Anju S Kumar 1
Date of Publication :7th February 2016
Abstract: The overall performance of peer-to-peer (P2P) file sharing lies on the efficiency of file query. To enhance the efficiency of file query in a structured peer-to-peer system, clustering technique can be used. Clustering peers by their common interests and by their physical proximity can improve file query performance. In the clustering technique the physically-close nodes are formed into a cluster and further physically-close and common-interest nodes are grouped into a sub-cluster based on a hierarchical topology. In the search mechanism, the file query will move to the nearest proximity cluster and to the corresponding interest cluster within that proximity cluster. If all nodes within this cluster can respond to the query, there is a need for a method to choose appropriate node to share the file. In this paper, we propose a method, called the statistical feature matrix form (SMF), which improves the searching mechanism in the structured Peer–to-Peer system by selecting neighbors according to their capabilities. SMF measures the number of shared files, the content quality, the query service and the transmission distance between neighbor nodes. Based on these measurements, appropriate nodes can be selected by finding the rank of each nodes in the cluster, thereby reducing the traffic overhead significantly and enhance the file sharing efficiency in the structured P2P system.
Reference :