p1 -FP: Extraction, Classification, and Prediction of Website Fingerprints with Deep Learning

RegulaTor: A Straightforward Website Fingerprinting Defense

Website Fingerprinting (WF) attacks are used by local passive attackers to determine the destination of encrypted internet traffic by comparing the sequences of packets sent to and received by the user to a previously recorded data set. As a result, WF attacks are of particular concern to privacy-enhancing technologies such as Tor. In response, a variety of WF defenses have been developed, though they tend to incur high bandwidth and latency overhead or require additional infrastructure, thus making them difficult to implement in practice. Some lighter-weight defenses have been presented as well; still, they attain only moderate effectiveness against recently published WF attacks. In this paper, we aim to present a realistic and novel defense, RegulaTor, which takes advantage of common patterns in web browsing traffic to reduce both defense overhead and the accuracy of current WF attacks. In the closed-world setting, RegulaTor reduces the accuracy of the state-of-the-art attack, Tik-Tok, against comparable defenses from 66% to 25.4%. To achieve this performance, it requires 6.6% latency overhead and a bandwidth overhead 39.3% less than the leading moderate-overhead defense. In the open-world setting, RegulaTor limits a precision-tuned Tik-Tok attack to an F 1-score of. 135, compared to .625 for the best comparable defense.

Roza: a new and comprehensive metric for evaluating classification systems

Many metrics such as accuracy rate (ACC), area under curve (AUC), Jaccard index (JI), and Cohen's kappa coefficient are available to measure the success of the system in pattern recognition and machine/deep learning systems. However, the superiority of one system to one other cannot be determined based on the mentioned metrics. This is because such a system can be successful using one metric, but not the other ones. Moreover, such metrics are insufficient when the number of samples in the classes is unequal (imbalanced data). In this case, naturally, by using these metrics, a sensible comparison cannot be made between two given systems. In the present study, the comprehensive, fair, and accurate Roza (Roza means rose in Persian. When different permutations of the metrics used are superimposed in a polygon format, it looks like a flower, so we named it Roza.) metric is introduced for evaluating classification systems. This metric, which facilitates the comparison of systems, expresses the summary of many metrics with a single value. To verify the stability and validity of the metric and to conduct a comprehensive, fair, and accurate comparison between the systems, the Roza metric of the systems tested under the same conditions are calculated and comparisons are made. For this, systems tested with three different strategies on three different datasets are considered. The results show that the performance of the system can be summarized by a single value and the Roza metric can be used in all systems that include classification processes, as a powerful metric.

GANDaLF: GAN for Data-Limited Fingerprinting

We introduce Generative Adversarial Networks for Data-Limited Fingerprinting (GANDaLF), a new deep-learning-based technique to perform Website Fingerprinting (WF) on Tor traffic. In contrast to most earlier work on deep-learning for WF, GANDaLF is intended to work with few training samples, and achieves this goal through the use of a Generative Adversarial Network to generate a large set of “fake” data that helps to train a deep neural network in distinguishing between classes of actual training data. We evaluate GANDaLF in low-data scenarios including as few as 10 training instances per site, and in multiple settings, including fingerprinting of website index pages and fingerprinting of non-index pages within a site. GANDaLF achieves closed-world accuracy of 87% with just 20 instances per site (and 100 sites) in standard WF settings. In particular, GANDaLF can outperform Var-CNN and Triplet Fingerprinting (TF) across all settings in subpage fingerprinting. For example, GANDaLF outperforms TF by a 29% margin and Var-CNN by 38% for training sets using 20 instances per site.

Website Fingerprinting in the Age of QUIC

With the meteoric rise of the QUIC protocol, the supremacy of TCP as the de facto transport protocol underlying web traffic will soon cease. HTTP/3, the next version of the HTTP protocol, will not support TCP. Current website-fingerprinting literature has ignored the introduction of this new protocol to all modern browsers. In this work, we investigate whether classifiers trained in the TCP setting generalise to QUIC traces, whether QUIC is inherently more difficult to fingerprint than TCP, how feature importance changes between these protocols, and how to jointly classify QUIC and TCP traces. Experiments using four state-of-theart website-fingerprinting classifiers and our combined QUIC-TCP dataset of ~117,000 traces show that while QUIC is not inherently more difficult to fingerprint than TCP, TCP-trained classifiers may fail to detect up to 96% of QUIC visits to monitored URLs. Furthermore, classifiers that take advantage of the common information between QUIC and TCP traces for the same URL may outperform ensembles of protocol-specific classifiers in limited data settings.

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

A passive local eavesdropper can leverage Website Fingerprinting (WF) to deanonymize the web browsing activity of Tor users. The value of timing information to WF has often been discounted in recent works due to the volatility of low-level timing information. In this paper, we more carefully examine the extent to which packet timing can be used to facilitate WF attacks. We first propose a new set of timing-related features based on burst-level characteristics to further identify more ways that timing patterns could be used by classifiers to identify sites. Then we evaluate the effectiveness of both raw timing and directional timing which is a combination of raw timing and direction in a deep-learning-based WF attack. Our closed-world evaluation shows that directional timing performs best in most of the settings we explored, achieving: (i) 98.4% in undefended Tor traffic; (ii) 93.5% on WTF-PAD traffic, several points higher than when only directional information is used; and (iii) 64.7% against onion sites, 12% higher than using only direction. Further evaluations in the open-world setting show small increases in both precision (+2%) and recall (+6%) with directional-timing on WTF-PAD traffic. To further investigate the value of timing information, we perform an information leakage analysis on our proposed handcrafted features. Our results show that while timing features leak less information than directional features, the information contained in each feature is mutually exclusive to one another and can thus improve the robustness of a classifier.

Protecting against Website Fingerprinting with Multihoming

Anonymous communication tools, such as Tor, are extensively employed by users who want to keep their web activity private. But recent works have shown that when a local, passive adversary observes nothing more than the timestamp, size and direction (incoming or outgoing) of the packets, it can still identify with high accuracy the website accessed by a user. Several defenses against these website fingerprinting attacks have been proposed but they come at the cost of a significant overhead in traffic and/or website loading time. We propose a defense against website fingerprinting which exploits multihoming, where a user can access the Internet by sending the traffic through multiple networks. With multihoming, it is possible to protect against website fingerprinting by splitting traffic among the networks, i.e., by removing packets from one network and sending them through another, whereas current defenses can only add packets. This enables us to design a defense with no traffic overhead that, as we show through extensive experimentation against state-of-the-art attacks, reaches the same level of privacy as the best existing practical defenses. We describe and evaluate a proof-ofconcept implementation of our defense and show that is does not add significant loading-time overhead. Our solution is compatible with other state-of-the-art defenses, and we show that combining it with another defense further improves privacy.