The idea to infer the available bandwidth of a network path from end-host measurements dates back to packet pair probing and TCP congestion control. It has played an important role throughout the history of the Internet, with applications in monitoring, traffic engineering, and quality of service. In spite of numerous probing schemes, the foundations of bandwidth estimation are still not well understood. Most estimation methods are based on a simple network model, i.e. a single link with fluid constant rate cross-traffic. In case of multiple links and bursty traffic noticeable estimation bias occurs. These errors cannot be explained in current theories. Within the project the computer servers are used to perform simulations for bandwidth estimation with the network simulator NS-2.
Further information:
Understanding current and new technologies is a key issue to future trends in packet switched networks such as the Internet. For a long time classic queueing theory delivered important insights in this field. However, since the mid 90's it is well known that the assumption of Poisson-processes in classical queuing theory does not model Internet data traffic accurately.
In this project we develop a probabilistic network calculus that effectively takes the statistical multiplexing gain of independent data streams into account. Thus, it can be used as a framework for an efficient end-to-end analysis of communication networks. We provide the basics of a comprehensive, powerful yet intuitive theory that allows for new applications in the field of packet switched networks beyond the state-of-the-art theories, for example the established worst-case analysis. The RRZN cluster was helpful in the numerical analysis and simulation of formulae derived in the Properbounds project.
Publications and further information:
Regionales Rechenzentrum für Niedersachsen - URL: www.rrzn.uni-hannover.de/cluster_ikt.html?&L=1
Dr. Paul Cochrane, Last Change: 24.02.2012
Copyright Gottfried Wilhelm Leibniz Universität Hannover