On Oct 14, 2009, at 17:05, Julien Abeille (jabeille) wrote:
Hi Carsten,
I got the mesh under point.
Do we agree that
- in route over
- considering duty cycled nodes with efficient duty cycling techniques
(smapled listening, centralized scheduling)
(RFC4861) AR works (I can resolve addresses of one hop neighbors),
Well, 4861 explicitly says that it does not specify AR for NBMAs.
If you want to use the NS/NA pair of messages anyway, you first have
to make an on-link determination.
You would use an NHDP for that, and that could also carry the LLA
(possibly implicitly), so there is never a *need* for AR.
Then, of course, you could run the NHDP using messages that look like
4861 NS/NA messages, but they wouldn't have the same semantics; You
still have to specify the *protocol* of your NHDP, and that would best
be done in the context of the routing protocol.
(I don't mind a routing protocol using ND messages, in particular if
there are piggy-backing opportunities; it just complicates things a
bit.)
NUD works,
I'd like to discuss NUD separately, as the term is used for a number
of things in 4861, some of which are really about routing. (Also,
with 802.15.4 ACKs, there is a bit less of a need for NUD than on
Ethernet.)
router and prefix discovery... Work
Yes.
(In a LoWPAN, you would add context dissemination here, see section
4.3 and 4.5.2/4.5.3.)
A remaining problem is that the lifetime of the RA may not be useful,
as the router may go out of the node's radio range earlier.
You may call the part of the NHDP that detects this NUD, but see above.
DAD fails
Yes, because DAD is not a one-hop feature in a LoWPAN.
In other terms features related to reachability of one hop neighbors,
resolution of their address, router/parameter/prefix discovery work?
Addressing related features fail?
See above.
Gruesse, Carsten
_______________________________________________
6lowpan mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6lowpan
