[support] IKE (v2) and IPsec implementation in a NEPL context

[Sebastien Decugis]

Hi,

I add just a few thoughts for the pleasure of the discussion :-)

> A person having its own lightweight
> MAR creating a PAN for all his devices (Phone, laptop, maybe gps device
> sending data to a server? etc)
>   
I agree this is the usual use case for nested scenario. My point was
that I don't think we are remotely close to have this kind of
connectivity nowadays ;)

Even before you can start any mobility signaling, you must acquire
network access. Unless you are working in an open environment (research
testbed) this is not so simple. The delay of authentication (EAP
exchange and so on) is usually not so short, so basically the benefit of
IP mobility (the continuity of the session) will not be so obvious
because all application timers will have expired anyway when you change
your point of attachment...

Sorry if the previous paragraph is not clear, let me rephrase like this:
I am not sure that having a router that supports NEMO makes a big
difference compared to a simple router, for the MAR, in terms of user
experience. On the other side, for moving infrastructures such as
trains, we know in advance which access networks will be encountered,
and in this case the authentication can be "smart" (pre-authentication
and so on) and NEMO "really" useful.

Once again, this is only my personal opinion :-)

> Or even better a rescuer out in the rescue field getting connectivity from
> another rescuer or a vehicle, again by carrying its own personal MAR
> supporting his camera, gps device, smart phone, sensors etc...
>   
In this case, does the rescuer really need a "mobile" router ? During a
mission, he probably will not change its point of attachment much...


> Well, no I don't think that we are that close to real-life nested scenarios,
> but we are not that far either! From another point of view, we are
> researchers we should cater for the not so unlike to happen scenarios ;-)
>   

That's true :) But if we don't mind the operational aspects of the
mechanism, it is likely to never be used. Nowadays, Internet is far from
being an open network, IMHO. So, the concept of mobility where you can
use any infrastructure to get connectivity seems a bit odd to me, since
there are so many other conditions to consider to get this connectivity
(interoperability of technology, credentials, knowledge of the
parameters, ...). Of course, a lot of work is on-going in this area
also, but I think there are some stop-blockers anyway : for example, we
cannot really assume that all network providers will ever come to an
agreement to let all users roam through their networks.

> Well, what's your opinion now? Because you say that "two years ago" you were
> with that impression...
>   
Yes, for two years I have been working in a slightly different field:
AAA (authentication, authorization, accounting). Hence my speech about
network access in the previous paragraphs :) I have not monitored much
what is happening in IP mobility field, that's why I put this "2 years
ago". I know that the ITS for example is very active field of research,
I know also that connectivity is now available in some trains and
planes, but I don't know the technologies that are used for this. Is it
NEMO ?

> I think id/loc solutions are even further down the road...!
>   

I am not sure it is *that* far away either, but on the other hand the
goal is probably slightly different, and the problems it will encounter
are probably the same as mipv6 :)

Best regards,
Sebastien.