Jon Crowcroft has been the Marconi Professor of Communications Systems in the Computer Laboratory at the University of Cambridge since October 2001. He has worked in the area of Internet support for multimedia communications for over 30 years. He graduated in Physics from Trinity College, University of Cambridge in 1979, gained an MSc in Computing in 1981 and PhD in 1993, both from UCL. He is a Fellow the Royal Society, a Fellow of the ACM, a Fellow of the British Computer Society, a Fellow of the IET and the Royal Academy of Engineering and a Fellow of the IEEE.
In this talk, I will discuss the challenges placed on us when the goal is to find anomalies in the Internet. The challenges are statistical, methodological and ethical. Firstly, today's Internet is vastly more complex (not just larger) than its original designers expected. The number of applications numbers in the millions. Systems increasingly involve machine-to-machine communication as well as human-to-human, or simple information retrieval. Thus the dimensionality of the system is massive, and the definition of an "anomaly" is itself non-trivial. Secondly, the scale requires us to limit measurement to sampling, and yet as we've just observed, we have dynamic, high dimensionality data. What to sample, where and when, is increasingly non-obvious. This leads to the third strand of the talk, which is that the most successful measurement methods we have found in the last decade also represent a fairly extensive invasion of privacy, and are subject both to ethical and to legal constraints. And yet the Internet is a critical resource upon which society, for social, economic and practical reasons depends. Just as other utilities (water, electricity) require careful protection, the Internet needs care and attention. The old Jon Postel principle "be conservative in what you send, and liberal in what you accept" is a fine guiding design saying, but unfortunately, there are many intentional and unintentional behaviours in today's net which are a long way from that ideal, and understanding (and possibly repairing them) seems to be ever more important. The talk is based on work in the EU Marie-Curie METRICS project, which is an international training network for PhDs that has been looking at requirements for network measurement and sharing of data, and held a summer school in Estonia in 2015 on Cybersecurity.
One particularly popular moment associated to the colloquium is the “Master Class” where students have the opportunity to give a short (but well-prepared) presentation of his/her work. Each presentation (10 minutes) is followed by an open discussion with the guest speaker (15 minutes) who gives a detailed feedback. The complete program is provided here.