[e2e] packet-pair probe implementation

[Cottrell, Les]

Older implementations of pathload, pathchirp, pipechar appear to work pretty well up to OC3 limits. There is a new version of pathload (see http://www.cc.gatech.edu/fac/Constantinos.Dovrolis/pathload.html <http://www.cc.gatech.edu/fac/Constantinos.Dovrolis/pathload.html> ) that I believe is aimed at working at higher speeds, but we have not had a chance to try it. There is also ABwE (see http://moat.nlanr.net/PAM2003/PAM2003papers/3781.pdf <http://moat.nlanr.net/PAM2003/PAM2003papers/3781.pdf> ) which appears to work in most cases up to many hundreds of Mbits/s, however it is not packaged to be a downloadable tool yet. Beyond the Gbits/s rates, inter packet timing becomes difficult and interrupt coalescing can cause problems.

-----Original Message-----
From: Atsuo J. [mailto:atsuo_j at yahoo.com] 
Sent: Thursday, May 01, 2003 2:36 PM
To: end2end-interest at postel.org
Subject: [e2e] packet-pair probe implementation


Is the packet-pair probe (bandwidth estimation) technique (introduced by Keshav and revised by Paxson) really useful in the realistic network?
Does anybody test it on the real testbed? 
Also, are there any real implementation/usage of this technique at all?
 
Your suggestions will be very appreciated.
 
Atsuo



  _____  

Do you Yahoo!?
The New  <http://us.rd.yahoo.com/search/mailsig/*http://search.yahoo.com> Yahoo! Search - Faster. Easier. Bingo. 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.postel.org/pipermail/end2end-interest/attachments/20030501/3ad71e87/attachment.html