Peer-to-peer networks are highly distributed and unreliable networks. Peers log in and off the network at their own needs without any overall plan. In the real peer-to-peer case there are no central nodes planning the resources of the network or having an overview about the state of the network. The paper on hand describes and mathematically analyzes a storage algorithm allowing information to be stored within the network without the originator of the information needs to stay online. Information is optimally “blurred” within the network meaning that the information is reconstructable with a high probability and a long time interval, but stored as least redundant as possible. The main focus is to analyze the mathematical and statistical properties of the presented peer-to-peer storage algorithm. Technical procedures are described at a high level and need further improvement. Thus, the paper on hand is primarily purely statistically peer-to-peer theory at this stage of research.
Gehrke, Nick; Seidenfaden, Lutz; and Baule, Rainer, "Statistical Basics of a Reliable World Wide Web Peer to Peer Storage System" (2005). AMCIS 2005 Proceedings. 129.