Re: Self-similar traffic

[Haobo Yu <haoboy@isi.edu>]

Or if you are using current ns snapshot, look at
tcl/ex/large-scale-web-traffic.tcl. It is from the following paper:

@InProceedings{Traf,
  author = 	 {Anja Feldmann and Anna C. Gilbert and Polly Huang
                  and Walter Willinger},
  title = 	 {Dynamics of {IP} Traffic: A study of the role of
                  variability and the impact of control},
  booktitle = 	 proc # acm-sigcomm,
  year =	 1999,
  month =	 sep
}

- Haobo

On Wed, 20 Oct 1999, Lee Breslau wrote:

> 
> The aggregate traffic produced by a few 10s of Pareto On/Off sources
> should be self-similar I believe (traffic class Pareto).
> 
> Lee
> 
> > 
> > Hi,
> > 
> >  First, I would like to thank all of you who have kindly answered my
> > previous
> > question about "Node throughput and packet delay time". Finally, I wrote
> > a Perl program to extract the parameters I want from the trace-all files.
> > 
> >  I have another question. Currently, I would like to input a self-similar
> > traffic
> > to test the system. In the ns, is there any source generator which can
> > generate the
> > self-similar traffic? Or, should I generate by myself and input
> > to the ns using the attach-traffic Traffic/trace?
> > 
> > Thanks,
> > 
> > Po-Yuen Cheng
> > 
> 
> 
> 
> 
> 
>