Website Fingerprinting Defenses at the Application Layer

Abstract

Website Fingerprinting (WF) allows a passive network adversary to learn the websites that a client visits by analyzing traffic patterns that are unique to each website. It has been recently shown that these attacks are particularly effective against .onion sites, anonymous web servers hosted within the Tor network. Given the sensitive nature of the content of these services, the implications of WF on the Tor network are alarming. Prior work has only considered defenses at the client-side arguing that web servers lack of incentives to adopt countermeasures. Furthermore, most of these defenses have been designed to operate on the stream of network packets, making practical deployment difficult. In this paper, we propose two application-level defenses including the first server-side defense against WF, as .onion services have incentives to support it. The other defense is a lightweight client-side defense implemented as a browser add-on, improving ease of deployment over previous approaches. In our evaluations, the server-side defense is able to reduce WF accuracy on Tor .onion sites from 69.6% to 10% and the client-side defense reduces accuracy from 64% to 31.5%.

Website Fingerprinting Through the Cache Occupancy Channel and its Real World Practicality

“…These works have investigated both the selection of features (such as packet sizes, packet timings, direction of communication), as well as the design of classifiers (such as support vector machines, random forests, Naive Bayes) to make accurate predictions [10,29,35,36,37,45,46,53,60,66,67,73,88,89,93]. In response, several defense mechanisms have been proposed in the literature [11,12,15,63,90]. The common idea behind these defenses is to inject random delays and spurious cover traffic to perturb the traffic features and therefore obfuscate users' communications.…”

“…Because the attack observes the internal state of the target PC, rather than the network traffic. It offers the potential of overcoming traffic shaping, often proposed as a defense for website fingerprinting [11,12,15,63,90]. Similarly, the attack may be applicable in scenarios where network-based fingerprinting is known to be less effective, such as when the browser caches the contents of the website [36].…”

How Unique is Your .onion?

“…This is a limitation of our approach to countermeasures, but might be a limitation to website fingerprinting defenses in general, as large sites are easily identified by website fingerprinting attacks. However, we believe that our results can inform the design of application-layer defenses that alter websites in order to perturb site-level features [8]. This would allow to optimize existing application-layer defenses by focusing on the features that our site-level feature analysis has identified as most identifying, thus reducing the performance that these defenses incur in Tor.…”

“…This substantially reduces the number of sites to consider when only targeting onion services, as the universe of onion services is orders of magnitude smaller than the web, which makes website fingerprinting attacks potentially effective in practice. In addition, onion services are used to host sensitive content such as whistleblowing platforms and activist blogs, making website fingerprinting attacks on this sites particularly attractive, and potentially very damaging [8]. For these reasons, we focus our analysis on onion services rather than the whole web.…”

Measuring Information Leakage in Website Fingerprinting Attacks and Defenses

“…Our paper includes these defenses for information leakage evaluation. We leave recently proposed defenses [9,48] in our future work.…”