Enhancement the Efficiency of Work with P2P System

Abstract

The P2P system should be used Proximity information to minimize the load of file request and improve the efficiency of the work .Clustering peers for their physical proximity can also rise the performance of the request file. However, very few currently work in a peer group based on demands as peers on physical proximity. Although structured P2P provides more efficient files requests than unstructured P2P, it is difficult to apply because of their strictly defined topology. In this work, we intending to introduce a system for exchange a P2P file for proximity and level of interest based on structured P2P nodes that form physically block in the cluster and other groups physically close and nodes of public interest in sub-cluster based on the hierarchical topology. Querying an effective file is important for the overall P2P file exchange performance. Clustering peers from their common interests can significantly enhance the efficiency of the request file PAIS use an intelligent file replication algorithm to further rise the efficiency of the request file .Create a copy file that is often requested by a group of physically close nodes in their position. In addition, PAIS improves the search for files within the intra-system sub-cluster through various approaches. First, it further classifies interest in the sub-cluster to a number of subsections of interests and groups with common interest-free sub nodes in the group for file sharing. Secondly PAIS creates an over for each group that connects nodes of less node capacity to a higher throughput for the distributed node overload prevention request file. Third, in order to reduce the search for late files, PAIS uses a set of proactive information so that applicant can file knowledge if its requested file is in the neighboring nodes. Fourth, reduce the overhead of collecting information about files using the PAIS, collection of file information based on the Bloom Filter and the corresponding search for files distributed. Fifth, in order to improve the efficiency of file sharing, PAIS ranks the results with a blob of filters in order. Sixth, while the newly visited file is usually re-visited approach, based on the Bloom filter is improved only through the management of new information flowering filter is added to reduce the delay of file search. The experimental result of the Real-world Planet Lab Experiment shows that PAIS significantly reduces overhead and improves the efficiency of scrolling and without sharing files. In addition, the experimental results show high efficiency within the sub-research cluster of file approaches to improve file search efficiency.