Diogo Barradas will be joining the Cheriton School of Computer Science as a tenure-track Assistant Professor in January 2022.
Education
- PhD, Information Systems and Computer Engineering, Instituto Superior Técnico, Universidade de Lisboa, Portugal, 2021
- MSc, Information Systems and Computer Engineering, Instituto Superior Técnico, Portugal, 2016
- BSc, Information Systems and Computer Engineering, Instituto Superior Técnico, Portugal, 2014
Research interests
Diogo Barradas’s research is in the area of censorship-resistant communication.
People worldwide who are under the pressure of their governments are restricted from free communication — not only the contents of such communication, which can be protected using standard cryptography, but also the very fact that communication is taking place needs to be hidden from powerful adversaries. These adversaries control the entire communication perimeter — e.g., the Great Firewall of China — and they can deeply inspect, modify, and block communication attempts. However, usually they cannot block all communication attempts, since most of them serve an interest of the government such as economic development.
Diogo has significantly advanced the state-of-the-art on the use of personal video communication, such as Skype, as a carrier for covert TCP/IP communication. An advantage is that personal video communication requires high bandwidth to be usable, such that sufficient bandwidth for the covert message might be available. In his work he has shown that all existing methods are easy to detect using simple statistical tests and he has developed the first methods that are undetectable even by advanced machine learning algorithms. Furthermore, he has developed scalable systems at a very low level in the network stack that enable the wide deployment of these methods on the Internet.
Diogo publishes regularly in the top venues for computer and communication security and privacy (ACM CCS, USENIX Security, NDSS), as well as at WWW and in other top venues specialized on computer privacy, such as the PET Symposium, the world’s premier forum exclusively dedicated to computer privacy.
As of March 2021, Diogo had nine peer-reviewed publications, including four in the above top-tier venues. According to Google Scholar, his 2018 USENIX Security paper titled “Effective Detection of Multimedia Protocol Tunneling using Machine Learning,” which developed the most advanced detection techniques, had been cited 27 times as of March 2021.
