Title: Network—Specific Computer Traffic Modeling and Prediction

Speaker: Joel Vaughan Department of Statistics, University of Michigan

Date/Time: Wednesday, August 18, 2010 at 9:00am – 10:00am        

Location: CSRI Building/Room 90 (Sandia NM)

Brief Abstract: Computer network use is becoming increasingly widespread, both in terms of number of users and variety of applications. In order to provide consistently high quality service, network engineers and other professionals must monitor several aspects of the network, including the traffic intensity on the links that comprise the network. As networks grow, this type of monitoring has potential to become burdensome in terms of resources required. Motivated by the prospect of directly measuring only a small subset of links, I will address  the problem of using observed traffic measurements on a small number of links to predict the traffic on other, unobserved links. I will begin the talk with a physically interpretable probabilistic model of the traffic over the network, and then describe how the important characteristics of the traffic on links may be learned through auxiliary data. Although expensive to obtain, this extra data set provides the necessary information to represent important structure in the network, and can significantly improve the results of prediction as compared with more naive approaches. I will conclude with an appropriately  adjusted control chart methodology for network traffic monitoring.

CSRI POC: Bill Hart, 505- 845-2217