Performance Modeling of P2P File Sharing Applications

Peer-to-Peer (P2P) based file sharing applications have become highly popular in today's Internet due to the spread of platforms such as Napster, Gnutella, KaZaa, eDonkey, BitTorrent, and others. There are several interesting issues in P2P-based file sharing applications that affect their performance. As their diffusion grew, the scientific community started to analyze these applications using analytical and mathematical techniques, measurements and simulations. In this paper we make a first step towards a comprehensive survey of the existing literature on the performance analysis of P2P-based file sharing applications. In particular, we first provide a classification structure of performance issues in P2P-based file sharing that exploits different viewpoints that could be chosen to analyze the behavior of such applications. Furthermore, we discuss and summarize a subset of the proposals that are based on analytical and mathematical models of P2P-based file sharing applications. The discussion of measurement and simulation based experiments is left for later development.