[e2e] Can feedback be generated more fast in ECN?

[Qiong Li]

On Wed, 14 Feb 2001, Michael Welzl wrote:

> > We need not limit to SQ.
> > The question is:
> > (1) whether backward ECN(SQ like) is better than forward ECN(ECN like)
>
> All BECN-for-the-Internet efforts that I know of used SQ. It's handy
> because it's available and designed for that purpose. And whatever
> you use doesn't really matter.
>
> And you can't limit it to this simple question; it's just a little
> more complex. But if you absolutely WANT it as simple as that
> ("could it replace ECN because it's plainly better"), the answer is no.
>
>
> > (2) can we design a "cheap" mechanism using backward ECN
>
> Now that's a relevant question.   :)
>
> Cheers,
> Michael
> >

We have done a preliminary simulation in NS-2 to test backward ECN. In the
simulation, RED-like calculation is done in the forward path, and the
calculation results are then used as references to mark packets in the
backward path in the same router. For obvious reasons the queues in the
simulation are more stable than that produced by normal RED+ECN. The
strong presumption of this scheme is that the traffic of both directions
pass through the same node, which may not always be true in the Internet.


Cheers
Qiong