[Iccrg] Proposal for ICCRG operations

[Steven Low]

Hi Keshav

Your proposed steps make a lot of sense.  It seems a little too
aggressive - great if we can make it.  In particular, Phases 2
and 3 on developing a solutions will likely take more time if
the goal is not just to sort out what proposals are available
today and understand their tradeoffs, but to come up with something
that is significantly better than the various proposals available
today (each of which has limitations in some scenarios).

I think a TCP benchmarking suite will be a very helpful tool to
guide us in this process.  It will be useful especially in your
Phase 3 to evaluate various proposals, but just thinking through
the design and implementation of a good benchmark lends clarity
to issues of Internet congestion control and metrics to optimize
against.

Aaron Falk is chairing a panel at a conference to discuss this
very issue.  Hopefully, we wll hear from the participants soon
on their thoughts.

Steven


S. Keshav wrote:

> A Proposal for ICCRG Operations
> ---------------------------------------------
> 
> The ICCRG Charter Document states (lightly paraphrased): "The key goal of
> ICCRG is to move towards consensus on which technologies are viable
> long-term solutions for the Internet congestion control architecture and
> what constitutes an appropriate cost/benefit tradeoff. ... Unfortunately, it
> is unclear at this stage whether any single solution is viable, or whether
> there needs to be a synthesis of ideas from many places."
> 
> To achieve the requisite clarity I propose the following phases.
> 
> Phase 1: Challenge document
> Target date: Preferably by end of Feb, but no later than March 20th (IETF
> Dallas date)
> 
> This document will outline the challenges faced by current TCP congestion
> control, and the requirements of any long-term solution for Internet
> congestion control.
> 
> Exit criterion: Consensus in this list that the document covers the bases.
> 
> Phase 2: Solution proposals
> Target date: By Fall IETF
> 
> We will broadcast a call to all contenders to submit a congestion control
> proposal. To make this reasonably efficient, the format of the submission
> will need to be agreed to and specified by this community.
> 
> Exit criterion: A document that surveys all contenders and presents them in
> standard format
> 
> Phase 3: Discussion
> Target date: Spring IETF 2007
>   
> The goal is to identify common themes and best practices in the proposed
> solutions as well as holes in coverage. Perhaps, we can use these to run a
> competition or bakeoff for the top spot, (which may require teams to
> collaborate or synthesize ideas).
> 
> Exit criterion:  A document that identifies the form of the best solution.
> 
> Phase 4: Consensus formation
> Target date: Fall IETF 2007
> 
> We will use this phase to clean up the preliminary solution and handle any
> issues than may come up in the previous phase.
> 
> Exit criterion: Creation of a document that recommends a 'best known
> solution'
> 
> I hope that this approach will result in a solution that is not only widely
> accepted by the community but has also been critically evaluated.
> 
> 
> Comments on this proposal are very welcome.
> 
> thanks
> 
> keshav
> 
> 
> 
> _______________________________________________
> Iccrg mailing list
> Iccrg at cs.ucl.ac.uk
> http://oakham.cs.ucl.ac.uk/mailman/listinfo/iccrg
> 

-- 
_____________________________________________
Steven Low                assoc prof, cs & ee
netlab.caltech.edu        tel: (626) 395-6767
slow at caltech.edu          fax: (626) 568-3603