Search and replication in unstructured peer-to-peer networks

Abstract
Decentralized and unstructured peer-to-peer networks suchas Gnutella are attractive for certain applications becausethey require no centralized directories and no precise controlover network topology or data placement. However, theflooding-based query algorithm used in Gnutella does notscale; each query generates a large amount of tra#c andlarge systems quickly become overwhelmed by the queryinducedload. This paper explores, through simulation, variousalternatives to Gnutella's query...

This publication has 3 references indexed in Scilit: