Peer-to-Peer Innovation Project
Research Project initiated and financed by |
Project Team |
Summary
The latest developments in the Internet suggest that future telecommunication applications will be realized in distributed Peer-to-Peer (P2P) environments. In a distributed P2P network each peer is considered equal. That is, each peer runs the same piece of software and performs exactly the same tasks as any other peer in the network. Consequently, those applications no longer depend on central unities. While distributed applications show the potential to reduce the total cost of ownership, little is known about the robustness, scalability and performance of those systems.Current P2P systems like Gnutella, eDonkey, Overnet, and Kazaa are mainly used for media sharing and other legally questionable applications. The time needed to complete a search in those file-sharing systems is not really critical to the end-user since file download time exceeds the preceding lookup time of the files location by magnitudes. Real-time applications with quality of service demands, like VoIP telephony, chatting, or instant messaging on the other hand are dependent on the time needed to find their communication partner. To be able to build business case services based on P2P systems, the ability of such mechanisms to satisfy real-time requirements has to be studied.
The Resource Management Framework (RMF) provides a unified model of P2P computing which is independent of the underlying protocols. In this project a proof-of-concept is provided for the RMF prototype regarding different aspects. The primary research focus is thereby on the scalability as well as the robustness of the RMF subject to different conditions. The corresponding protocol overhead is also examined in detail, leading to suggestions for architecture improvements.
Conference Articles
| 1. | Andreas Binzenhöfer, Gerald Kunzmann, and Robert Henjes A Scalable Algorithm to Monitor Chord-based P2P Systems at Runtime. Third International Workshop on Hot Topics in Peer-to-Peer Systems (Hot-P2P) in conjunction with the IEEE International Parallel & Distributed Processing Symposium ( IPDPS 2006), Rhodes Island, Greece, April 2006. [PDF (204 kB)] [Bibtex] |
| 2. | Gerald Kunzmann, Andreas Binzenhöfer, and Robert Henjes Analyzing and Modifying Chord's Stabilization Algorithm to Handle High Churn Rates. MICC & ICON 2005, Kuala Lumpur, Malaysia, November 2005. [PDF (121 kB)] [Bibtex] |
| 3. | Andreas Binzenhöfer, Dirk Staehle, and Robert Henjes On the Stability of Chord-based P2P Systems. GLOBECOM 2005, St. Louis, MO, USA, November 2005. [PDF (437 kB)] [Bibtex] |
| 4. | Andreas Binzenhöfer, Dirk Staehle, and Robert Henjes On the Fly Estimation of the Peer Population in a Chord-based P2P System. 19th International Teletraffic Congress (ITC19), Beijing, China, September 2005. [PDF (1.13 MB)] [Bibtex] |
| 5. | Phuoc Tran-Gia and Andreas Binzenhöfer On the Stochastic Scalability of Information Sharing Platforms. 2005 Tyrrhenian International Workshop on Digital Communications, Sorrento, Italy, July 2005. [PDF (501 kB)] [Bibtex] |
| 6. | Andreas Binzenhöfer, Phuoc Tran-Gia Delay Analysis of a Chord-based Peer-to-Peer File-Sharing System. ATNAC 2004, Sydney, Australia, December 2004. [PDF (257 kB)] [Bibtex] |
