Dissertation talk: Threat modeling and circumvention of Internet censorship

Presentation | September 28 | 1:30-2:30 p.m. | 373 Soda Hall

 David Fifield

 Electrical Engineering and Computer Sciences (EECS)

Research on the circumvention of Internet censorship requires modeling the behavior of censors. Circumvention systems are designed according to assumptions arising from models: the more accurate the models, the more effective the circumvention. A circumvention system designed under inadequate assumptions risks irrelevance, through being easily blockable or impractical to deploy.

I will present observations on the behavior of real-world censors, and on circumvention systems (already deployed or under development) that take advantage of their weaknesses. Censorship is not only a technical phenomenon but also a social and economic one: censors should be understood not only as a set of capabilities, but also as entities subject to various influences and limitations. This work aims at a richer understanding of the problem of censorship, leading to better circumvention.