Signaling and networking in unstructured peer-to-peer networks

This work deals with the efficiency of Peer-to-Peer (P2P) networks, which are distributed and self-organizing overlay networks. We contribute to their understanding and design by using new measurement techniques, simulations and analytical methods. In this context we first present measurement methods and results of P2P networks concerning traffic and topology characteristics as well as concerning user behavior. Based on these results we develop stochastic models to describe the user behavior, the traffic and the topology of P2P networks analytically. Using the results of our measurements and analytical investigations, we develop new P2P architectures to improve the efficiency of P2P networks concerning their topology and their signaling traffic. Finally we verify our results for the new architectures by measurements as well as computer-based simulations on different levels of detail.