Mathematical statistical characterization of high-speed computer network traffic data

The studying of high-speed networks is an actual problem in both fields of modeling of communication networks and in mathematical statistics. In this paper the methods of time series analysis are applied to the traffic data of the routers by the FDDI ring of the University of Debrecen. After deseasonalization the five data series examined are found to be long-memory processes. The long-memory parameters are estimated by three different methods.