[manet] Re: MANET WG Charter and MANET Internetworking

Hi Juliusz,

> -----Original Message-----
> From: Juliusz Chroboczek <[email protected]>
> Sent: Wednesday, July 09, 2025 11:43 AM
> To: Templin (US), Fred L <[email protected]>
> Cc: Donald Eastlake <[email protected]>; manet <[email protected]>; [email protected]
> Subject: Re: [manet] Re: MANET WG Charter and MANET Internetworking
> 
> > The phrase "operate autonomously when isolated" means that there is
> > a need for some form of unique IP addressing that is somehow gotten
> > without the aide of supporting network infrastructure and has no
> > Internet topological reference.  So that, when the MANET
> > (opportunistically) attaches to the Internet, it can't simply inject an
> > aggregated IP prefix into BGP. So, there are different classes of IP
> > addresses needed - those gotten from infrastructure, and those made
> > available for autonomous operation when isolated.
> 
> There are some requirements here that you're not making explicit, and I'm
> struggling to understand what they are.  Are you assuming that the MANET
> starts up without having been assigned a globally-routable network prefix,
> that it then connects to the Internet

Yes, for many MANETs (disaster relief, vehicle-to-vehicle communications, etc.) that
is exactly what I am assuming. Each MANET node must have a unique IP address
even if there is no common shared globally routable network prefix common to
all nodes. The MANET then forms spontaneously based on the discovered network
neighborhood according to a MANET routing protocol. Multihop routing is then
possible outside of the context of any Internet infrastructure.

> and that it cannot renumber at that point?

When the MANET comes in contact with an Internet attachment point, one or
a few nodes may be within wireless communications range of the attachment
point while the majority of nodes may still be multiple hops away from the
access point and still in need of their non-aggregated unique addresses. So,
MANET-wide renumbering is possible but would be an onerous task and if the
MANET moves again and comes in contact with a different access point then
the whole process would have to be done all over again. Other processes that
could be considered would be tunneling or translation - or possibly even
tunneling together with translation.

> Where do these requirements come from?  Do they come from an actual usage
> scenario?  Or have they been crafted in order to force the use of
> a specific solution?

Not crafted in any rigged way, no. There have been many decades of investigation
into MANET use cases (disaster relief, manned/unmanned vehicular communications,
tactical military, many others) and they all have the commonality that a MANET can
form based on an arbitrary collection of nodes coming together in a common operating
region to spontaneously form a multi-hop connected network even outside of the
presence of supporting Internet infrastructure.

Thank you - Fred

> 
> -- Juliusz

_______________________________________________
manet mailing list -- [email protected]
To unsubscribe send an email to [email protected]