Network Modeling for the Internet
Formulation of
Suite of Model
| Leader | Zhengyuan Zhu (University of North Carolina, Chapel Hill), zhuz@email.unc.edu |
| Meeting | Every Other Tuesday 2:00 -
3:00 pm, room 104 |
| Members | Jay
Aikat (University of North Carolina, Chapel Hill),
aikat@cs.unc.edu Kevin Jeffay (University of North Carolina, Chapel Hill), jeffay@cs.unc.edu Steve Marron (SAMSI/University of North Carolina, Chapel Hill), marron@email.unc.edu Jonathan Mattingly (Duke University), jonm@math.duke.edu Krishanu Maulik (EURANDOM, The Netherlands), maulik@eurandom.tue.nl Cheolwoo Park (SAMSI), cwpark@email.unc.edu Juhyun Park (University of North Carolina, Chapel Hill), parkj@email.unc.edu Surajit Ray (University of North Carolina, Chapel Hill), sray@bios.unc.edu David Rolls (SAMSI), rollsd@uncw.edu Haipeng Shen (University of North Carolina, Chapel Hill), haipeng@email.unc.edu |
| Outline Objectives | Our goal is to develop useful
statistical models for Internet traffic
flow that are simple to analyze and simulate, and can capture the
characteristics of actual traffic data that are important to
electronics engineers and computer scientists. To achieve that goal,
we will try to combine the top-down approach with the bottom-up
approach.
The class of statistical models we propose is based on the fact that
Internet traffic is an aggregation of individual connections, each with
a
random starting time, random duration (from heavy-tail distribution)
and a
random throughput. We will use a bottom-up approach to model the
starting
time, duration, throughput and their dependence structure by analyzing
the
data derived from the actual traffic flow, and study the statistical
property of the random process of aggregated packet counts thus
obtained. The working group will try to address the following problems: 1. Identify appropriate measures for evaluating how well a model fits the data (i.e., how to determin if model generated data is close to actual traffic data). 2. Find good models under those measures. 3. Develop methodology to estimate the parameters of the model as well as making statistical inference. 4. Simulate traffic flow efficiently under such model. 5. Other things group members want to address. Computer scientists are particularly welcome to join our group! |
| References |
© 2002, Statistical and Applied Mathematical Sciences Institute. All rights reserved.