A de-anonymize attack method based on traffic analysis
Ming Juan Song, Gang Xiong, Zhenzhen Li, Junrui Peng, Li Guo · 2013
While providing protection for users' privacy, anonymity network has also been exploited by criminals to carry out crime anonymously. We study the problem how to break the unlinkability between the senders and recipients in order to identify the source of anonymous traffic in this paper. Tor, the most widely deployed anonymity network, is selected as our target. We develop a de-anonymize attack method based on traffic analysis and choose the {time, stream size} as features for k-means algorithm to mine the association between the first hop traffic and last hop traffic of Tor. Experiments show that our method is effective for Tor.