Re: Simulations parameters

[Sean Murphy <murphys@teltec.dcu.ie>]

Max,

On Thu, 2 Sep 1999, Massimo Pegorer - stage wrote:

> Does anybody know which type of TCP (Reno, NewReno...) has been used in
> the draft "Preliminary Simulation Evaluation of an Assured Service"
> (<draft-ibanez-diffserv-assured-eval-00.txt>), and how has been
> configured the ns2 TCP variables (window_...) ???

I tried to reproduce one set of results from that paper to verify my
diffserv implementation - I wrote some code to add some diffserv
functionality to ns. I can send you the code if you like, although I'd
suspect you're writing your own ns mods. I didn't get exactly the same
results, but I got results that were somewhat similar.

For the simple experiment that I wanted to reproduce (sec 3.2 of the
paper), I used Reno TCP with a window size of 64 and got similar results -
I have a short note describing all the other parameters that I used which
I can mail you if you want - it also lists the set of parameters that I
used to reproduce results obtained by Kim, Leland & Thomson in their paper
"evaluation of bandwidth assurance service using red for internet service
differentiation."

HTH,
Sean.