Statistical disclosure attacks: Traffic confirmation in open environments
George Danezis · 2003
An improvement over the previously known disclosure attack is presented that allows, using statistical methods, to effectively deanonymize users of a mix system. Furthermore the statistical disclosure attack is computationally efficient, and the conditions for it to be possible and accurate are much better understood. The new attack can be generalized easily to a variety of anonymity systems beyond mix networks. © 2003 by Springer Science+Business Media New York.