Re: [Autoconf] Question regarding use of LLs

[Henning Rogge <henning.rogge at fkie.fraunhofer.de>]

On Mon November 16 2009 11:13:30 Alexandru Petrescu wrote:
> > As I understand "4861 links" no... 4861 would assume that the two links
> > between A-B and B-C are not connected directly.
> 
> Yes, I believe 4861 would assume B is a router between the two links.
In a wireless network it's just a matter of radio range and connectivity, 
which might change through external influences and movement.
 
> > The typical case in a MANET (especially if every node has only one
> > interface) is that there are LOTS of nodes (in this case B) that can
> > reach two nodes (in this case A and C) on the same interface which cannot
> > hear each other.
> 
> Could we define the ABC links in terms of 4861 links or not?
In wireless networks links between nodes don't have disjoined groups of 
interfaces. The 'is connected to' condition is not always transitive in 
wireless networks.
 
> If not, I am afraid we can not use the term "link" in MANET at all.
I disagree.

Link is a connection between two devices on layer 2. It's the right  term for 
the connection between two wireless devices too.

> On another hand, people have run "MANET" with links and link-local
> addresses in the past too.
MOST times it will work. Just by trying to get a "linklocal unique" IP, you 
will get "mesh unique" IPs. Or maybe the same IPs are two far away to trouble 
you.

But you have no guarantee for this and DAD as described in IPv6 cannot detect 
some of the problems.

Henning Rogge

-- 
Diplom-Informatiker Henning Rogge , Fraunhofer-Institut für
Kommunikation, Informationsverarbeitung und Ergonomie FKIE
Kommunikationssysteme (KOM)
Neuenahrer Straße 20, 53343 Wachtberg, Germany
Telefon +49 228 9435-263,   Fax +49 228 9435 685
mailto:henning.rogge at fkie.fraunhofer.de http://www.fkie.fraunhofer.de
GPG: E1C6 0914 490B 3909 D944 F80D 4487 C67C 55EC CFE0

Attachment: signature.asc
Description: This is a digitally signed message part.