Thursday, November 29, 2012 at 3:00pm
Traffic measurement is a vital tool in the management of communications networks. Sampling reduces resource usage in the measurement infrastructure while supporting both routine and exploratory queries. The demands of network management to dive deeply into usage patterns, coupled with the decreasing relative cost of measurement transmission bandwidth, and the need to join traffic measurements with other operational datasets, together favor an evolution towards time-limited retention of unsampled measurements. In this setting, sampling provides summaries that can be rapidly queried, triggering deeper dives as required, and retained for baselining and retrospective analysis. The analytic principle underpinning this talk is that sampling must optimally trade-off measurement volume against measurement query accuracy. I show how this principle has been used to design sampling algorithms currently deployed in measurement infrastructures, and can flexibly provide enhancements to these algorithms that support various accuracy goals..
Bio: Nick Duffield is a Distinguished Member of Technical Staff and an AT&T Fellow at AT&T Labs-Research, Florham Park, NJ, where he has worked since 1995. He previously held post-doctoral and faculty positions in Dublin, Ireland, and Heidelberg, Germany. He received a BA in Natural Sciences in 1982 and an MMath (Part III Maths) in 1983, both from the University of Cambridge, UK, and a PhD in Mathematical Physics from the University of London, U.K., in 1987. His current research focuses on measurement and inference of networks and traffic. He is a co-inventor of the Smart Sampling technologies that lie at the heart of AT&T's scalable Traffic Analysis Service. He was Charter Chair of the IETF working group on Packet Sampling. He was an Associate Editor for the IEEE/ACM Transactions on Networking from 2007-2011. He is an IEEE Fellow, and was a co-recipient the Sigmetrics 2012 Test of Time Award, for work in Network Tomography. He will serve as TPC Co-Chair for Performance 2013. Further details and publications can be found at www.research.att.com/people/Duffield_Nicholas_G.