Abstract

In this paper, we propose a completely distributed topology generation mechanism named HPC5 for Gnutella network. A Gnutella topology will be efficient and scalable if it generates less number of redundant queries. This can be achieved if it consists of a fewer number of short length cycles. Based on this principle, our protocol directs each peer to select neighbors in such a way that any cyclic path present in the overlay network will not generate any redundant query. We show that our approach can be deployed into the existing Gnutella network without disturbing any of its parameters. We also show that the probability of inconsistencies arising during topology generation, using our mechanism, which may lead to the formation of a small number of short length cycles is very low. However, we have also proposed an inconsistency handling protocol that detects such short length cycles and effectively removes them. We implemented a Gnutella prototype to compare and validate the efficiency of our protocol over existing Gnutella. Simulation results indicate that our mechanism outperforms existing Gnutella in terms of network coverage (the number of unique peers explored during query propagation in limited flooding) and message complexity. Structural analysis indicates that the proposed enhancement conserves the robustness of existing Gnutella network. Finally, we draw comparisons of the proposed protocol with a state-of-the-art topology optimization protocol named Distributed Cycle Minimization Protocol (DCMP); the simulation results indicate that HPC5 outperforms DCMP in terms of message overhead and network coverage.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.