Statistical and Applied Mathematical Sciences Institute
19 T. W. Alexander Drive
P.O. Box 14006
Research Triangle Park, NC 27709-4006
Tel: 919.685.9300 FAX: 919.685.9310
info@samsi.info
2003-04 Program on
Network Modeling for the Internet
Because of the size and complexity of the internet, and the nature of the protocols, Internet traffic has proven to be very challenging to model effectively. Yet modeling is critical to improving Quality of Service and efficiency. The main research goal of this program is to address these issues by bringing together researchers from three communities:
The timing is right for simultaneous interaction among all three communities, because of the current trend away from dealing with Quality of Service issues through over-provisioning of equipment. This trend suggests that heavy traffic models will be ideally situated to play a leading role in future modeling of Internet traffic, and in attaining deeper understanding of the complex drivers behind Quality of Service. This SAMSI program will catalyze this process through building strong bridges among the three communities.
An additional goal is to enhance contact between these research communities and potential industrial partners. The location of SAMSI is ideal for this purpose, because the Research Triangle is becoming a world center for the networking industry.
Workshops. There will be three workshops. The first two, to be held back-to-back on September 17-20, 2003, are deliberately connected, to enable close contacts among heavy traffic researchers, statisticians and network researchers. The third, on October 12-14, 2003, will focus on the rapidly developing area of Internet tomography.
Research. Research will bring the three communities together through collaborative work centered on real data, as well as carefully designed laboratory experiments. Observed Internet behavior will be modeled using ideas from the applied probabilists. The goodness of fit of these models (Do they behave like real traffic?) will be assessed by the statisticians, working together with network researchers. The results will be the basis for new insights about Internet traffic, and for a new generation of mathematics and models from the applied probabilists. SAMSI aims to establish the contacts that will allow this process to continue well beyond the duration of the program.
SAMSI-University Fellows, appointed for a semester or the entire year, will help lead the research, as well as join academic life at a SAMSI partner university; see www.samsi.info/univfellows200210.html for further information. Salary and timing are ideal for senior sabbatical visitors with partial funding from their home institutions.
Long-term Research Visitors will participate in the research; support for expenses is available. Please write to internet@samsi.info if you are interested in exploring this possibility.
Postdoctoral Fellows can be appointed for two or more years; see www.samsi.info/postdoc200210.html for information and application instructions.
Workshop Attendees both inform the course of the research and have early access to the results, and may receive support for expenses. Workshops will be announced individually on the SAMSI Web site.
Members of underrepresented groups are especially encouraged to apply.
Kevin Jeffay (North Carolina), James Landwehr (Avaya Laboratories),
John P. Lehoczky (Carnegie Mellon), J. S. Marron (North Carolina/SAMSI,
Co-Chair), Ruth Williams (California San Diego, Co-Chair), Walter
Willinger (AT\&T Research), Donald Towsley (Massachusetts).
© 2002, Statistical and Applied Mathematical Sciences Institute. All rights reserved.