Toronto Networking Seminar

Multiscale network monitoring and anomaly detection

Mark Coates
Department of Electrical and Computer Engineering
McGill University

Date:  Friday, November  9, 2:10pm
Location: BA1220 (Bahen Center)


Network operators need to monitor the end-to-end performance experienced by all users in order to identify and correct for overloaded links and to verify that quality of service agreements are being satisfied.  The key challenge is to obtain end-to-end performance levels without the costly overhead involved in explicitly gathering data for the paths connecting every source-destination pair.  Recently we proposed a method for characterizing end-to-end performance across a network given limited measurements.  Our method uses diffusion wavelets, a methodology for defining multi-scale representations of functions defined on a graph, to obtain an efficient representation for network performance parameters. With this representation, we can employ nonlinear sparse estimation techniques, yielding a state-of-the-art framework for network inference. This talk will describe this monitoring framework, and discuss its application to network anomaly detection and traffic matrix estimation.


Mark Coates received a B.E. (first class honors) in computer systems engineering from the University of Adelaide, Australia in 1995 and a Ph.D. in information engineering from the University of Cambridge, U.K. in 1999. Currently, he is an Assistant Professor at McGill University, Montreal, Canada. He was awarded the Texas Instruments Postdoctoral Fellowship in 1999 and was a research associate and lecturer at Rice University, Texas, from 1999-2001. His research interests include communication and sensor/actuator networks, statistical signal processing, causal analysis, and Bayesian and Monte
Carlo inference.