[e2e] Determining leaky bucket paramteres

[Oliver Heckmann]

Hi

You will also find an literature overview and some performance measurements
of different algorithms in
http://www.kom.e-technik.tu-darmstadt.de/publications/abstracts/HRS01-1.html
(a paper based on that is submitted to NOSSDAV 2002).

regards

Oliver

----- Original Message -----
From: "Craig Partridge" <craig at aland.bbn.com>
To: "Bhuvan Urgaonkar" <bhuvan at none.cs.umass.edu>
Cc: <end2end-interest at postel.org>
Sent: Friday, February 01, 2002 10:57 PM
Subject: Re: [e2e] Determining leaky bucket paramteres


>
> I would recommend looking for works by Perry Tang.  While I think Mark
> Garrett and I get credit for inventing the algorithm to find the set
> of available token buckets, Perry's done all the hard work to develop
> the ideas (Mark and I just scribbled a solution on a napkin).
>
> Craig
>
> In message <Pine.SOL.4.31.0202011523300.6603-100000 at none.cs.umass.edu>,
Bhuvan
> Urgaonkar writes:
>
> >Hi there,
> >
> >I am looking for work that has been done on the following problem: We are
> >given information about the traffic a source sends over a link (in the
> >form of the number of bytes sent plotted against time). We would like to
> >come up with a leaky bucket description that this source adheres to.
> >Specifically, of the several (rate, burst) pairs that the given traffic
> >confirms to, which is better than others and why.
> >
> >It would be a big help if someone could point me to papers that
> >address either this exact problem or any related problems.
> >
> >regards,
> >-Bhuvan
> >---------------------------------
> >Bhuvan Urgaonkar,
> >http://lass.cs.umass.edu/~bhuvan
> >---------------------------------
> >
>
>