[e2e] packet-pair probe implementation

[Constantine Dovrolis]

Joseph,

I agree that some rough bandwidth estimate using packet trains
may be useful in modifying slow-start, as you did in Swift Start.
However, strictly speaking, it has been shown before that packet
trains do not measure "available bandwidth", at least if the
latter is defined as "bandwidth not used by other traffic".

Related references:

1. http://www.sics.se/cna/connected/conn-globecom.ps
@INPROCEEDINGS{melander:globecom00,
     AUTHOR = "B. Melander and M. Bjorkman and P. Gunningberg",
     TITLE = "{A New End-to-End Probing and Analysis Method for
              Estimating Bandwidth Bottlenecks}",
     BOOKTITLE = "IEEE Global Internet Symposium",
     YEAR = "2000"
}

2. http://www.cc.gatech.edu/~dovrolis/Papers/infocom01.ps
@INPROCEEDINGS{dovrolis:infocom01,
     AUTHOR = "C. Dovrolis and P. Ramanathan and D. Moore",
     TITLE = "{What do Packet Dispersion Techniques Measure?}",
     BOOKTITLE = "Proceedings of IEEE INFOCOM",
     YEAR = "2001"
}


Constantinos

--------------------------------------------------------------
Constantinos Dovrolis | 218 GCATT | 404-385-4205
Assistant Professor | Networking and Telecommunications Group
College of Computing | Georgia Institute of Technology
dovrolis at cc.gatech.edu
http://www.cc.gatech.edu/fac/Constantinos.Dovrolis/

On Mon, 5 May 2003, Joseph Ishac wrote:

> Another implementation of packet-pair probes can be found within the
> Swift Start Proposal (BBN/NASA).  Swift Start uses an initial burst of 4
> packets to estimate the available bandwidth, and then paces a portion of
> the estimate onto the wire over the course of a round trip time.
>
> The project page is at:
> http://www.ir.bbn.com/projects/pace/cap/index.html
>
> There is a FreeBSD kernel implementation available at the above site, as
> well as a few supporting technical reports.
>
> -Joseph
>
> On Thu, May 01, 2003 at 02:36:02PM -0700, Atsuo J. wrote:
> > 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 Yahoo! Search - Faster. Easier. Bingo.
>
>