On the one hand, the Internet makes worldwide communication possible at an unprecedented scale and speed. On the other hand, it also makes it possible to censor entire countries, in ways that can be difficult to detect. For example, a censoring regime may not simply block Wikipedia traffic, but rather return its own version of a Wikipedia page. Some countries have long been known to censor their own citizens, but only recently has it become apparent that some are censoring *any* traffic crossing their borders—even if neither the source nor destination are in the censoring country.
In this talk, I will present a system that avoids such censors by letting users specify not only where they want their data to go, but also which countries to avoid on the way there and back. The challenge in developing such a system is finding a way to *prove* that it did not go through parts of the network that the user wanted to avoid. How does one prove that something did *not* happen? Our insight is to obtain an "alibi"—proof that the data took a path through the network that would have made it impossible to *also* go through a censor. I will present how we find these alibis, and some results showing which countries can avoid one another.
I worked with several UMD undergraduates (as well as graduate students and faculty members) to design, implement, and evaluate alibi routing. In addition to presenting some the details of our work, I will also share some of my experiences with undergraduate research (and why I believe you should do it).
Please RSVP Here!