Behavior Analysis of Internet Traffic Via Bipartite Graphs and One-Mode Projections

Speaker:        Dr. Kuai Xu
                Division of Mathematical and Natural Sciences
                Arizona State University

Title:          "Behavior Analysis of Internet Traffic Via Bipartite
                Graphs and One-Mode Projections"

Date:           Wednesday, 4 July 2012

Time:           3:00pm - 4:00pm

Venue:          Room 3416 (via lifts 17/18), HKUST

Abstract:

As Internet traffic continues to grow in size and complexity, it has
become an increasingly challenging task to understand behavior patterns of
end-hosts and network applications. This talk will present a
behavioral-oriented approach to cluster Internet end-hosts and network
applications. In this talk, I will first describe our approach of applying
bipartite graphs to model host communications from network traffic, and
building one-mode projection graphs to discover social-behavior similarity
in Internet traffic. I will also present simple yet efficient clustering
algorithms to group end-hosts in the same network prefixes and network
applications into distinctive behavior clusters based on the similarity
matrices and clustering coefficient of one-mode projection graphs .
Finally, I will demonstrate the applications of behavior clusters in
profiling Internet traffic, discovering emerging applications, and
detecting anomalous traffic patterns.


*****************
Biography:

Kuai Xu is an Assistant Professor at the Division of Mathematical and
Natural Sciences, Arizona State University. He received his Ph.D. degree
in computer science from the University of Minnesota in 2006, and his B.S.
and M.S. degrees from Peking University, China, in 1998 and 2001. His
research interests span the areas of Internet measurement, network
security, cloud computing and online social networks. His research has
resulted in over 30 papers in major conference and journals including ACM
SIGCOMM and IEEE/ACM Transactions on Networking, and two awarded United
State patents.