Software improves p2p privacy by hiding in the crowd
Wednesday, April 8, 2009 - 13:28
in Mathematics & Economics
Researchers at the McCormick School of Engineering and Applied Science at Northwestern University have identified a new "guilt-by-association" threat to privacy in peer-to-peer (P2P) systems that would enable an eavesdropper to accurately classify groups of users with similar download behavior. To thwart this threat, they have released publicly available, open source software that restores privacy by masking a user's real download activity in such a manner as to disrupt classification.