Re: [ns] Pareto's "shape_" parameter values ?

["Peter Rabinovitch" <peter.rabinovitch@alcatel.com>]

Whoa there.....

"The shape = 3 - 2 * Hurst_parameter ........."

If what you are talking about is the aggregation of heavy tailed
processes to get a long range dependent process, then you have to
remember to aggregate them and to scale them.

See "Proof of a Fundamental Result in Self-Similar Traffic Modeling"
(Taqqu, Willinger and Sherman) and "The Mystery of the Missing Scales:
Pitfalls in the Use of Fractal Renewal Processes to Simulate LRD
Processes" (M. Roughan, J. Yates and D. Veitch) before you do.

As far as 0.5 being Poisson, NO. H=0.5 implies independent increments of
the traffic. Thats all. It can be brownian motion, bernoulli, etc. 
As far as "1<H<0.5 => Traffic with some selfsimilar properties" I assume
you meant 0.5<H<1. Still not precisely correct, because, for example,
brownian motion H=0.5, is self-similar, i.e. H=0.5 can be SS too.



Debojyoti Dutta wrote:
> 
> This has been discussed 2 days back too. You may want to read the
> archives. The shape parameter determines teh tail of the distribution..
> 
> The shape = 3 - 2 * Hurst_parameter .........
> 
> H=0.5 => poisson traffic.
> 1<H<0.5 => Traffic with some selfsimilar properties
> 
> For more details you can look up Stalling's High Speed Networks .....
> 
> Debo
> 
> On Wed, 5 Sep 2001, [iso-8859-1] Andrea Josi wrote:
> 
> >
> > Hi,
> >
> > i would like to know what values can be given to the
> > Pareto traffic "shape_" parameter .
> > Does it represent a specific parameter of the Pareto
> > distribution ?
> > Is it limited to a specific interval of values ?
> >
> > Thanks in advance.
> >
> > ______________________________________________________________________
> > Do You Yahoo!?
> > Il tuo indirizzo gratis e per sempre @yahoo.it su http://mail.yahoo.it
> >
> >