Software for the Multifractal Wavelet Model (MWM)

Real WAN Traffic	Gaussian Synthetic Data	MWM Data
The MWM is a multiscale multiplicative data model which is able to capture not only correlation structure but also `spikiness'.  Here, it is shown in comparison with a multiscale additive model which can match only the correlations.

Software

• UNIX (gzip compressed format) (87KB).
  To unpack gzipped files, execute "gunzip mwm.tar.gz" followed by "tar -xvf mwm.tar".
  A new sub-directory called MWM with the matlab files will be generated.
  Start by reading the README file.
  
  
• PC version (zip compressed format) (95KB).
  To unpack the zipped files, execute "unzip mwm.zip" or use your favorite software that can decompress a ".zip" file.
  A new sub-directory called MWM with the matlab files will be generated.
  Start by reading the README file.

Publications
This software synthesizes positive multifractal long-range-dependent data sets as described in

• R. H. Riedi, M. S. Crouse, V. J. Ribeiro, and R. G. Baraniuk,
  A Multifractal Wavelet Model with Application to Network Traffic,
  IEEE Transactions on Information Theory
  (Special Issue on Multiscale Signal Analysis and Modeling),
  Vol 45 (April 1999), 992-1018. [19 pages plus tutorial appendix, 530KB].

Authors:

• Matthew Crouse, DSP Group, Rice University
  
• Vinay Ribeiro, DSP Group, Rice University
  

More on multifractal modeling

• A brief introduction

Please report any bugs or send comments regarding these programs to vinay@rice.edu