Improved website fingerprinting on Tor

By Tao Wang, Ian Goldberg. 2013

344 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Networks and Communications, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Networks and Communications

In Computer Vision and Pattern Recognition

Citations in the corpus

Website Fingerprinting at Internet Scale (2016, methodology): “[5] and Wang et Goldberg [26] showed that these defenses are not as effective as assumed and, in the case of randomized pipelining, might even lead to increased recognition rates on small datasets.”
Deep Fingerprinting: Undermining Website Fingerprinting Defenses with Deep Learning (2018, methodology): “Since then, the success rate of WF attacks against Tor has been incrementally improved, reaching 90% accuracy by two classifiers using edit-distances [8, 39].”
A Systematic Approach to Developing and Evaluating Website Fingerprinting Defenses (2014, methodology): “We use the ﬁrst 200 elements of classes C and C (cid:48) for training and the last 200 elements for testing our feature classiﬁers described below.”
Automated Website Fingerprinting through Deep Learning (2017, positive): “It should be noted that, in contrast to prior work [31], the Tor entry guard node was not pinned over the course of our experiments.”
k-fingerprinting: A Robust Scalable Website Fingerprinting Technique (2015, methodology): “Wang and Goldberg [36] then investigated the practical deployment of website fingerprinting attacks on Tor.”
Effective Attacks and Provable Defenses for Website Fingerprinting (2014, methodology): “Figure 2: True Positive Rate and False Positive Rate changes in OSAD [23] as the open set size increases.”
A Critical Evaluation of Website Fingerprinting Attacks (2014, methodology): “Some authors have also evaluated their classifiers in an open-world scenario [3, 23, 32], where there is a set of k target pages being monitored but the user is allowed to visit pages that are not in that set.”
Circuit Fingerprinting Attacks: Passive Deanonymization of Tor Hidden Services (2015, background): “Previous WF attacks assumed that the adversary is able to eliminate noisy background traffic [10, 35, 36].”