Research on the Real Time Estimation of Traffic Descriptors for High Speed Telecommunications Networks
Project Award Date: 0000-00-00
Ubiquitous high speed networks (HSN) available on a global scale will enable new classes of telecommunications services, e.g., multimedia, to be offered to the customer. The way customers will use future high speed networks is impossible to predict and thus real time performance monitoring and traffic engineering for HSN's is an active area of research. In the context of this research, 'real time' means that observations are collected and processed in time to be useful for controlling and managing the network. The design of the next generation of switches and systems will benefit from well-developed performance monitoring and traffic engineering technologies. To gain the required understanding of the phenomenon, traffic measurements, traffic descriptors, and the fundamental limitations on the real time estimation of those descriptors are needed.
This research program addresses performance monitoring and traffic characterization issues for HSN. This effort will begin with a selection of target HSN services, e.g. scientific visualization using remote supercomputers. From these services and continuing standards efforts, a set of traffic metrics will also be selected. Having defined what needs to be measured, the network characteristics that are correlated with those metrics will be developed. Techniques and algorithms to measure these characteristics in real time will then be determined as well as the quality of these estimators as a function of measurement sampling and offered data rates. These algorithms will then be implemented in the MAGIC testbed and an initial set of measurements collected and analyzed.
Primary Sponsor(s): Bell Northern Research