Re: [Autoconf] draft-clausen-manet-linktype - naming

["Teco Boot" <[e-mail]>]

> > The real problem (or confusion), I think, is that when you say
> > non-transitivity, you assume that all three nodes are on the same
> > link. However, there are others who don't think that they are.
> 
> I don't think that's my assumption.
> 
> It's more fundamental than what you call a link. It's that if
> A's transmissions can reach B, and B's transmissions can reach C,
> but A's transmissions can't reach C, then we have non-transitivity.
> That's the usual case with wireless transmissions (to be precise,
> it's the usual case that A's transmissions may or may not reach C).
> 
> Thus either A, B and C can't be on one link (if your model of a
> link is as Ethernet) or you need a different model of what a link
> is, and can consider A and B to have a link, and B and C to have
> a link (but not A and C). For MANETs the latter may be more useful.
> But the non-transitivity is there, whatever definitions you choose.

Agreed this a definition issue.
In the Autoconf meeting, there was consensus on defining a xxx link type
(name tbd), where different links are on the same interface. I think we can
assume that the medium is shared (we are on the same electro-magnetic
spectrum), but the degree of interference is not known on forehand (e.g.
TDMA vs. CSMA/CA).



The 802 series of standards could be helpful as a reference. The STDs refer
to RFCs, why do we have some asymmetry here?

Useful ones:

IEEE802.1-2001 (section 6.3.3.describes usage of routers and the IP
protocol)

IEEE802.11-2007 (too many sections to list here)

This is an interesting one:
>>> 3.16 basic service set (BSS):
<skip>   Membership in a BSS does not imply
that wireless communication with all other members of the BSS is possible.
<<<

And this one:
>>> 3.76 link:
In the context of an IEEE 802.11 medium access control (MAC) entity, a
physical path consisting
of exactly one traversal of the wireless medium (WM) that is used to
transfer an MAC service data unit
(MSDU) between two stations (STAs).
<<<
(this would exclude broadcast MSDU, which is not the case)



Another comment on naming:
I think we have two task. One of them is describing this "ad hoc wireless
link". 802.11 definitions could be used, maybe some can come up with other
refs. But in a MANET, multiple link types can be used. Think of an Ethernet
cable connecting two radios "back-to-back" or using alternative radios with
lower frequency / lower bandwidth / more Tx power / increased range. I think
it helps to write down something on "MANET topologies".



Just repeating myself:
Lets postpone IP addressing and ND issues until we are familiar with MANETs
that are out there.


Teco.
 

_______________________________________________
Autoconf mailing list
[e-mail]
https://www.ietf.org/mailman/listinfo/autoconf