Improving Computation Power by Reducing Query Response Time in Peer-to-Peer Environment

Nandagopal Nandagopal

doi:10.3844/jcssp.2011.434.439

Abstract

Problem statement: The Peer-to-Peer (P2P) was an emerging model and was being widely adopted in today’s internet computing.P2P traffic contributes the largest portion of the internet traffic which was unnecessary and led to delay. In existing Scalable Bipartite Overlay (SBO) network, Out of “N” peers, one group of peers were probing the messages (or) queries whereas the other group of peers were computing. This resulted in high query response time .Since the system was bipartite all peers were not capable of computing and probing, resulting in increased computation time. Approach: To overcome the above issue we propose a system RQR (Reduction on query response time) which is decentralized and unstructured where each and every peer can perform both probing and computing with no restrictions. Results: Optimal path finding algorithm is designed to find all possible path existing in the P2P network along with the optimal path. Conclusion: Our comparison with earlier approaches show that less than 60% of reduction on query response time which increases the system performance.

Highlights

A P2P network is composed of an unbounded set of peers
Decentralized architecture further can be divided into structured, unstructured and hybrid p2p networks .In structured p2p overlays a distributed hash table data structure is used in which every data item can be located within a small number of hops at the expense of keeping some state information locally at the nodes (Demetrios et al, 2007)
In ICRQR if a new peer is joining in the network, the joining peer does both probing and computing based on the request from other peers and the waiting time and processing time will be comparatively less than the existing Scalable Bipartite Overlay (SBO)

Summary

INTRODUCTION

A P2P network is composed of an unbounded set of peers. Each peer has its own role. In SBO (Scalable Bipartite P2P Overlay Network) design, the nodes are categorized into two in which one category of nodes only probe the messages (or) queries where as the other category of peers only computes (Liu et al, 2007). In ICRQR if a new peer is joining in the network, the joining peer does both probing and computing based on the request from other peers and the waiting time and processing time will be comparatively less than the existing SBO. This results in Reduction in query response time and increased throughput.

Processing time

MATERIALS AND METHOD

AND DISCUSSION

CONCLUSION