|
|
Internet Traffic -- Characteristics and Models
|
Author
|
Joachim Charzinski
Siemens Information and
Communication Networks
|
Abstract
|
A major portion of the traffic in present networks is Internet data
traffic. This kind of traffic has characteristics quite different from
the traffic known in yesterday's telecommunication (telephone) networks.
Terms like "fractal", "self-similar" and "long-range dependent" are
employed to describe the new traffic characteristics, to deter people
from asking further questions and even to argue that the New Traffic
cannot possibly be understood at all within any classical framework.
This tutorial sets the different observations into relation, describing
causes responsible for the effects that are visible on different levels
of observation. The implications of heavy tails are introduced and
discussed in the context of measurement, characterization and
traffic generation. Based on research results from the literature
and our own measurements, we characterize different aspects of user,
application and backbone traffic. Some modeling approaches and the
resulting models for Internet traffic are discussed and practical
advice is given for handling Internet traffic models in simulations.
|
Who should attend?
|
The tutorial is directed towards people interested in IP traffic
phenomena. It will give an overview of user and application
traffic characteristics to both researchers and performance
engineers interested in understanding Internet traffic,
simulating Internet users or applications or dimensioning
IP based networks.
A notion of probability and a background in basic performance
analysis is helpful but not required. Most of the presented
effects and connections between them can be understood without
|
Biography
|
Joachim Charzinski is project leader for IP traffic characterization and
engineering at Siemens Information and Communication Networks in Munich,
Germany. His research interests include IP traffic characterization, user
perceived quality of service, performance analysis and simulation, and
protocol engineering. He received his Dipl.-Ing. and Dr.-Ing. degrees
(approximately equivalent to MS and PhD) from University of Stuttgart,
Germany, and joined Siemens in 1997.
|
|