Please note: This seminar will be given online.
Diogo Barradas, Information Systems and Computer Engineering
Instituto Superior Técnico, Universidade de Lisboa
Totalitarian states are known to deploy large-scale surveillance and censorship mechanisms in order to deter citizens from accessing or publishing information on the Internet. However, there is evidence that most oppressive regimes cannot afford to block all channels with the outside world, allowing the operation of widely used services such as video-conferencing solutions. This has sparked the development of censorship-resistant communication tools that rely on the establishment of covert channels on the Internet by encoding covert data within popular encrypted multimedia protocols, e.g., Skype.
Our research efforts on the efficacy of the above tools led to multiple significant findings. First, we showed that the covert channels generated by these tools are trivially prone to detection. In particular, we developed a new machine learning (ML)-based traffic analysis framework which has broken the security assumptions of multiple state-of-the-art tools for embedding covert information within media streams. Second, we enabled sophisticated ML-based network flow classification tasks to be performed at line-speed. To this end, we worked towards the efficient deployment of multiple ML-based traffic analysis frameworks (including our own) in programmable switches. Third, we devised a new technique for creating traffic analysis resistant covert channels over multimedia streams. Our approach, based on the careful modification of the WebRTC media pipeline, allows for the creation of high-speed covert WebRTC flows whose traffic patterns closely resemble those of legitimate WebRTC media flows.
In this talk, I will describe my ongoing research efforts towards the improvement of: (i) circumvention tools that rely on the embedding of covert data within multimedia applications; (ii) methodologies to assess traffic analysis resistance; (iii) the understanding of how censors can deploy covert channel detection capabilities at scale, in response to increasingly sophisticated circumvention tools.
Bio: Diogo Barradas is a Ph.D. candidate in Information Systems and Computer Engineering at Instituto Superior Técnico, Universidade de Lisboa. He received his BSc. (2014) and MSc. (2016) from the same institution. His main research interests include network security and privacy, with particular emphasis on statistical traffic analysis and Internet censorship circumvention. He conducts his research at the Distributed Systems Group at INESC-ID Lisboa.
To join this seminar on Zoom, please go to https://zoom.us/j/98844608035?pwd=bm1wK0xTY0o1dnpjVnpTRVBwRTUydz09.
200 University Avenue West
Waterloo, ON N2L 3G1