Communication Networks and Services Research, Annual Conference on (2006)
Moncton, New Brunswick, Canada
May 24, 2006 to May 25, 2006
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/CNSR.2006.41
Z. Liu , University of Moncton, Canada
J. Almhana , University of Moncton, Canada
V. Choulakian , University of Moncton, Canada
R. McGorman , Lodex International Consulting, Canada
For the purpose of provisioning bandwidth for Internet access, we need to model the trafic at large time scales, over which the trafJic shows evident periodicity, long correlation and a non-Gaussian marginal distribution. To capture these characteristics simultaneously, in this paper we use a periodicity transform to identify the most sign$cant periods of the trafic and use an autoregressive time series to capture the autocorrelation and apply the G-and-H distribution to model the marginal distribution. A prediction-based bandwidth provisioning scheme is proposed and many experimental results on real Internet traces are also provided. Key words: Internet trafic, periodicity
Internet trafic, periodicity transfomt, G-and-H distribution.
J. Almhana, Z. Liu, V. Choulakian and R. McGorman, "Periodic Data Traffic Modeling and Predicition-Based Bandwith Allocation," 2006 Communication Networks and Services Research Conference(CNSR), Moncton, NB, 2006, pp. 131-138.