Tik-Tok: The Utility of Packet Timing in Website Fingerprinting Attacks

This paper investigates how packet timing can enhance the effectiveness of website fingerprinting (WF) attacks, which aim to identify websites visited over Tor. Previous research has often discounted timing information due to its volatility; however, this study introduces new burst-level timing features and evaluates their contribution in combination with directional features. The results demonstrate that including timing data improves the accuracy of WF classifiers, particularly in realistic open-world settings. The paper also introduces the Tik-Tok attack, a novel deep learning-based method that utilizes both timing and directional information, achieving significant accuracy improvements, especially against WTF-PAD and onion sites.

Recommended citation: Mohammad Saidur Rahman, Payap Sirinam, Nate Mathews, Kantha Girish Gangadhara, Matthew Wright. (2020). "Tik-Tok: The Utility of Packet Timing in Website Fingerprinting Attacks." Proceedings on Privacy Enhancing Technologies (PoPETS), 2020. DOI: 10.2478/popets-2020-0043.
Download Paper