[Tmrg] test suite: preliminary traces

[Lachlan Andrew]

Greetings all,

We finally have some traces which I submit as a first draft for use in
the TCP evaluation suite.  They're available at
<http://wil.cs.caltech.edu/suite/>.

Our aims were:
- The traces should preserve the correlation structure of the original
traffic traces kindly provided by UNC.
- Dependence of file size distribution on RTTs should be preserved, as
much as possible.
- Short (100s) experiments should see similar mean load as full 1hr
experiements.
- For the NIST people:  Very long simulations can be run without the
aggregate traffic becoming perfectly periodic.

Please read the description of how the traces were generated/massaged,
and comment or criticise :)


Wang Gang, is your NS2 framework well developed enough to check what
per-packet RTT distribution is actually achieved when this traffic is
used on the dumbbell topology with the delays we discussed?

Cheers,
Lachlan

-- 
Lachlan Andrew  Dept of Computer Science, Caltech
1200 E California Blvd, Mail Code 256-80, Pasadena CA 91125, USA
Ph: +1 (626) 395-8820    Fax: +1 (626) 568-3603
http://netlab.caltech.edu/lachlan