Deep Fingerprinting: Undermining Website Fingerprinting Defenses with Deep Learning

TL;DR

Deep Fingerprinting (DF) uses convolutional neural networks to significantly improve website fingerprinting accuracy against Tor, even defeating recent lightweight defenses like WTF-PAD, highlighting the need for stronger privacy protections.

Contribution

This paper introduces Deep Fingerprinting, a novel CNN-based attack that outperforms previous methods and effectively bypasses recent defenses like WTF-PAD and Walkie-Talkie.

Findings

DF achieves over 98% accuracy on unprotected Tor traffic.

DF maintains over 90% accuracy against WTF-PAD defenses.

Walkie-Talkie remains effective, limiting attack accuracy to 49.7%.

Abstract

Website fingerprinting enables a local eavesdropper to determine which websites a user is visiting over an encrypted connection. State-of-the-art website fingerprinting attacks have been shown to be effective even against Tor. Recently, lightweight website fingerprinting defenses for Tor have been proposed that substantially degrade existing attacks: WTF-PAD and Walkie-Talkie. In this work, we present Deep Fingerprinting (DF), a new website fingerprinting attack against Tor that leverages a type of deep learning called Convolutional Neural Networks (CNN) with a sophisticated architecture design, and we evaluate this attack against WTF-PAD and Walkie-Talkie. The DF attack attains over 98% accuracy on Tor traffic without defenses, better than all prior attacks, and it is also the only attack that is effective against WTF-PAD with over 90% accuracy. Walkie-Talkie remains effective, holding the attack to just 49.7% accuracy. In the more realistic open-world setting, our attack remains effective, with 0.99 precision and 0.94 recall on undefended traffic. Against traffic defended with WTF-PAD in this setting, the attack still can get 0.96 precision and 0.68 recall. These findings highlight the need for effective defenses that protect against this new attack and that could be deployed in Tor.