From October 2011 until May 2013, I was a systems engineer in Bryan Ford's Decentralized and Distributed Systems Group. Bryan was my research project advisor during my senior year of college, and during that time, we published a research paper about a new protocol, called "Dissent," for anonymous communication. I returned to Yale after college to be a staff member in Bryan's research group. In that capacity, I worked with post-docs, PhD students, and undergrads to identify good research problems, to write research papers, and to convert ideas into code.
The motivating problem of the Dissent project is that, on today's Internet, there is no practical way to send a message in such a way that the recipient of the message does not know who sent it. For me, the lack of online anonymity is troubling because it makes people less likely to look for information that they might need online and thus severely limits the everyday usefulness of the Internet.
During my time at Yale, we published a number of papers on anonymity systems that provide security guarantees against an adversary who can observe all network links and also control some number of participants in the group. Although such systems are not quite practical yet for everyday use, the existence of near-global adversaries in the real world makes these systems interesting (to me, at least!).