2009 | OriginalPaper | Chapter
Blind Maximum-Likelihood Estimation of Traffic Matrices in Long Range Dependent Traffic
Authors : Pier Luigi Conti, Livia De Giovanni, Maurizio Naldi
Published in: Traffic Management and Traffic Engineering for the Future Internet
Publisher: Springer Berlin Heidelberg
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
A method is proposed to estimate traffic matrices in the presence of long-range dependent traffic, while the methods proposed so far for that task have been designed for short-range dependent traffic. The method employs the traffic measurements on links and provides the maximum likelihood estimate of both the traffic matrix and the Hurst parameter. It is “blind”, i.e. it does not exploit any model neither for the traffic intensity values (e.g. the gravity model) nor for the mean-variance relationship (e.g. the power-law model). In the application to a sample network the error on traffic intensities decays rapidly with the traffic intensity down to below 30%. The estimation error of the Hurst parameter can be reduced to a few percentage points with a proper choice of the measurement interval.