Mathematical statistical characterization of high-speed computer network traffic data
The authors study the traffic of high speed FDDI networks on the basis of data measured in a real life environment. The measurements took place at the five member institutes of University of Debrecen by getting router data. The paper is about analyizing frame traffic through routers. Higher level protocols are not discussed which is natural considering the goal of the paper and the technical implementation characterizing lower layers. The paper says that after seasonally adjusting data taken from the measurements, the processes have ``long memory in all five cases. The authors estimate the memory parameter in three different ways. They provide the estimations after giving the details of the methods of the estimations. It is interesing that in certain cases all there methods yield nearly the same result (e.g. Institute DRK 0.42, 0.39, 0.4), while in other cases the deviation is greater. The later deviation is not considered significant by the authors. The 0.48 estimation of the third method in case of Institute MFK can be regarded as an interesting result considering the inequalities. NEWLINENEWLINENEWLINEThe paper is easily readable, the results are well illustrated.
- Bispectral analysis of traffic in high-speed networks
- Computer network traffic analysis with the use of statistical self-similarity factor
- An approximation method for capturing complex traffic behavior in high speed networks
- Traffic rate network tomography with higher‐order cumulants
- Long-range and fractal properties of some high speed network traffic data
This page was built for publication: Mathematical statistical characterization of high-speed computer network traffic data
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2707251)
