[Iccrg] Soliciting input for cc. bibliography

[Michael Welzl]

On Thu, 2006-06-15 at 17:09, Radek Krzywania wrote:
> > I believe that a short but sharply focused document on approaches to cc  
> > would be extremely valuable for network
> > designers to explore new ways of dealing with congestion in next  
> > generation networks.
> 
> Hi all
> I was tracking the recent discussion on the list, and I agree with Doan.  
> Large number of papers does not improve the current knowledge and are not  
> worth of anylising them in details. It is much better to deal with  
> approaches. I guess there are lot of web pages/papers which tries to  

I like this a lot; indeed, if you look at the GGF document
I co-edited: http://www.welzl.at/research/publications/GFD.55.pdf
it's lengthy, but I'm not so sure if there's much general
congestion control related knowledge that one could learn from
it. I'd hesitate to simply repeat that process here.

I think that the way to do it is to start with an approach
based classification, and mention example mechanisms that
show that this particular approach can work.

e.g.: "methods with additional implicit feedback" / "delay"
      with Vegas and FAST as example schemes.



> compare the approaches and describe them, but I'd like to intorduce yet  
> another page with short algorithms description:
> http://kb.pert.switch.ch/cgi-bin/twiki/view/PERTKB/TransmissionControlProtocol

Thanks, I put it in the wiki.


> You may find there a quite "human readable" descriptions, including main  
> TCP approaches. I'am sure that much details are lack there, but I think it  
> may be a good start anyway.

Stanislav Shalunov pointed me to this document the other day:
http://transport.internet2.edu/transport-design-space.pdf
(also in our wiki)

- and I think that this might be the best starting point. It's
exactly what we need - a classification of approaches, outlining
the design space...

right?

Cheers,
Michael