An analysis of GNUnet and the implications for anonymous, censorship-resistant networks

Abstract

Peer-to-peer networks are a popular platform for file sharing, but only few of them offer strong anonymity to their users. GNUnet is a new peer-to-peer network that claims to provide practical anonymous and censorship-resistant file sharing. In this paper we show that GNUnet's performance-enhancing features can be exploited to determine the initiator of a download. We also present an efficient filter mechanism for GNUnet. Assuming that content filtering is legally enforced, GNUnet can be censored at a large scale. © Springer-Verlag Berlin Heidelberg 2003.