Re: [DNA-BOF] Using L2 to provide Instantaneous Movement Detection andNeighborhood Discovery

[Greg Daley <greg.daley@eng.monash.edu.au>]

Hi Paul,

Paul Tan wrote:
> Hi Greg and all,
> 
> since I published the above draft in Feb 03, I have not followed up with 
> it as I was slightly disappointed that the idea was not 'welcome' in the 
> IP community.

I don't think the idea was unwelcome, but I think there
is a fine line between working on L2 and providing L3
information for movement/network attachment detection.

For example, we have to be careful that we don't
specify something which other organizations have
ownership or standardization jurisdiction over
(to be good neighbors).

> Since then, I have only simulated the proposal in various environments 
> (multiple UDP/TCP session, group mobility) with one of the graduate 
> students in my lab.
> 
> The idea is SIMLE, and the results are OBVIOUS. Simulating/Prototyping 
> the idea is trivial, but we need to standardize it in order to see the 
> actual deployment.

I can believe that the reception of the prefix as well as the
router IP address and MAC within a beacon frame or probe response
is a significant boost to performance (it will beat everything else).

Is it something which IETF can standardize on now?

I don't know.

If we were able to get the concept of application specific
information elements which you suggested in your draft to
be accepted by 802 then this may be quite different.

> After re-visiting my draft, I think that the draft can contribute in 
> some ways to this new WG. I see that there are some new drafts which 
> provide similar ideas. That's great to know!
> 
> Lastly, I still believed that using L2 to provide 'L3 information in an 
> extensible manner' is 'a' way to go. However, we must be careful not to 
> 'bloat' the L2 beacons/control-messages. Remember that my original 
> objectives are MD and CARD. It is really up to the IEEE 802.11 to give 
> us an element ID to implement this solution. As in what to put in, it is 
> up to us (e.g. DNA), the L3 guys.

I think that this is what Bernard and Alper are discussing with
strong and weak hints (getting the right information into L2).

The draft you wrote previously (it's still available at:
http://www.ietf.org/internet-drafts/draft-paultan-seamless-ipv6-handoff-802-00.txt 

)
could be used to implement such strong hints, but as you
say, there's work  needed from the 802 (maybe the handoffs
group there??) to get that far.

Whether the development of information elements to populate
beacon frames becomes the domain of the IETF (in a subset
delegated by 802) or another organization would be
dependent on any solution agreed to by 802.

The link-hint work which is proposed for this BoF at the moment
is a simple catalogue of existing systems, though, without
explicit proposals for improving Link-layers.  I'm pretty
sure these issues will come up in the group, and your idea
may play a strong role.  The initial stages of link-hint
improvements may be based on discussion and liaison though,
rather than getting Internet drafts onto standards track..

So please contribute to the discussion. It's not the
wrong IETF-related location to be talking about this.

Greg