Title: A Peer-to-Peer Large-scale Collaboration Network
Author: Thomas Bocek
e-mail: bocek@ifi.uzh.ch
Partner: UniZH
Supervisor: Prof. Dr. Burkhard Stiller
Committee: Prof. Dr. Burkhard Stiller, NN
Year of start: 2005
Year of end: 2008
Funding institution: University of Zurich, IFI
The aim of the PhD thesis is the development and analysis of a fully decentralized large-scale collaboration network. Collaboration networks require new algorithms to provide decentralized control and deal with misbehaving peers. Thus, the network is designed to be robust, scalable, and to provide incentive schemes against malicious peers and resource over usage. Key aspects to take into account in those algorithms are resource trading and human interaction, such as reviewing and voting. An example application of such a network enables storage, search, retrieval, and manipulation of documents in a large network. Control mechanisms, such as reviewing and voting for or against modified documents by other peers identify and exclude misbehaving peers.
The following four key topic are addressed in this thesis:
* Incentive schemes for P2P collaboration networks * Similarity search algorithm for searching documents * Decentralized voting algorithms * Extending existing P2P storage with incentive schemes, similarity searches and decentralized voting.