A novel passive website fingerprinting attack on tor using fast fourier transform

One of the main applications of low latency anonymity networks, such as Tor, is to protect data and users' privacy from interception over the Internet. This paper presents a novel passive website fingerprinting attack to defeat Tor's anonymity mechanism with a significant improvement in comparison with similar methods. Unlike the state-of-the-art approaches, the proposed method does not need to change the sequence of IP packets exchanged between users and the first relay in the network. We introduce a new method based on the Fast Fourier Transform to calculate the similarity distance of instances from traffic patterns. In this way, the time complexity of feature extraction is reduced by a factor of 400 during the classification process. Considering a closed-world scenario, our method easily spots on target websites with a minimum success rate of 95%. As yet another notable payoff, the accuracy keeps up more robustly while the polarity of target website grows gradually.