Best Information Search using Probabilistic Flooding in Unstructured Peer-to-Peer Networks

Abstract

The "peer-to-peer" (P2P) networks refer to a class of systems and applications that employ distributed resources to perform a function in a decentralized manner.Unstructured peer-to-peer networks do not impose any structure on the overlay networks.Peers in these networks connect in an ad-hoc fashion.Ideally, unstructured P2P systems would have absolutely no centralized system, but in practice there are several types of unstructured systems with various degrees of centralization.In heterogeneous decentralized unstructured P2P networks peers join and leave the application at their own will in an uncoordinated fashion and a central index for resource location is absent.Each peer is only responsible for maintaining a local index of the resources it owns and it is willing to provide to others.There are two main approaches for locating a resource in unstructured decentralized P2P networks; they are flooding and random walk.In random walk based strategies, peers forward a query message (termed as walker) to one randomly chosen neighbour whereas; in flooding based strategies peers forward a query message to all of its neighbours.The proposed work deals with the probabilistic approach which maintains shared table.The shared table consists of "terms" (query keywords) and the file name.When the peer receives the query, the shared table calculates the term frequency which is achieved by counting the number of occurrence of the term which the user has queried in each file.This process increases accuracy and also fetches the data very effectively.