Le 01/04/2013 21:35, Manfredi, Albert E a écrit :
Scott,
If the only goal is in-vehicle, then clearly ULAs are the answer, and
there's nothing more to be discussed. And for that matter, RFC 1918
addresses and IPv4 would be more than adequate. Limiting comms to
in-vehicle does not require any global uniqueness of addresses.
I agree (except the IPv4 part).
I think the model to follow here is more like E.164 uniform resource
identifiers (ENUM, RFC 3761). In enum, the telephone numbers are not
hard-coded in the IP address. Rather, they become a DNS entry. Seems
to me that any plan on using the uniqueness of the VIN number for
cars should follow this same approach.
Not saying that this DNS approach would answer all vehicle comms
issues to be resolved, though. For example, for self-driving
vehicles, there would be a need for more real-time (perhaps at layer
2) comms between adjacent cars and between cars and the roadway. E.g.
some way of a car with an impending problem to be shuttled off to the
emergency lanes, to get out of the way of the moving traffic.
Alex
Bert
-----Original Message----- From: Scott Brim
[mailto:s...@internet2.edu] Sent: Monday, April 01, 2013 9:23 AM To:
Manfredi, Albert E Cc: Alexandru Petrescu; fg...@si6networks.com;
6man Subject: Re: I-D Action:
draft-imadali-its-vinipv6-viid-00.txt
The scope of the draft was more or less restricted to in-vehicle
communications because of the privacy concerns ("The focus of this
work is to enable in-vehicle networks to exchange packets with
VIN-based IPv6 addresses." -- although inter-vehicle communications
is still a big deal under use cases), so I don't think use of DNS,
etc., apply. If the draft isn't really restricted to in-vehicle
communications I have a number of architectural arguments about why
it's a bad idea. VINs are neither topological locators nor IP
endpoint identifiers, and trying to map them onto either causes
confusion. They're fine as higher layer tokens.
Scott
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
