Peer-to-Peer systems, in contrast to the traditional client-server model do not distinguish between clients and servers: all participating computers are peers. Thus, all participating computers have the capability to act both as servers and as clients at the same time. Within this framework we conduct research and development on how to scale peer-to-peer networks to several millions of nodes while at the same time keeping the network management overhead to tolerable levels.
Our current efforts focus on studying the locality patterns of p2p systems in order to identify mechanisms that will exploit this locality and improve the overall scalability of such systems. Furthermore, our research contributions revolve around peer-to-peer network probing (tracing) and scalability improvement techniques, such as caching, network reorganization, and semantic proximity. In addition, we study issues concerning p2p network tomography and security of these networks.
Recently, an initiative for a Peer-To-Peer reading group has started.
Funding is provided by project SecSPeer: Secure and Scalable peer-to-peer computing and communication systems. Funded by the General Secretariat for Research and Technology.