Abstract
In the past, different intersection attacks on Chaum Mixes have been proposed and shown to work well in simulation environments. In this work we describe intersection attacks that have been performed on data from anonymized proxy log files. This approach creates all new problems that arise in real systems, where real-world users do not behave like those in the idealized model. E.g. the attack ...
Abstract
In the past, different intersection attacks on Chaum Mixes have been proposed and shown to work well in simulation environments. In this work we describe intersection attacks that have been performed on data from anonymized proxy log files. This approach creates all new problems that arise in real systems, where real-world users do not behave like those in the idealized model. E.g. the attack algorithm has to cope with a fixed number of observations. From the performed first experiments on the “dirty” real world data we get valuable insight into theory and practice of real anonymizers.