Re: [nemo] NEMOv4 vs DS-MIPv6

[Alexandru Petrescu <[e-mail]>]

Henrik Levkowetz wrote:
> Hi George,
> 
> on 2006-05-10 17:51 Tsirtsis, George said the following: ...
>>> I'm not sure what your scenario is, and in which situation NEMOv4
>>>  support would be necessary. If someone argue that NEMOv4 is 
>>> necessary because there are existing vehicles - for instance - 
>>> with IPv4 capabilities and the customer doesn't want to upgrade 
>>> the vehicles to IPv6, then I guess the customer could be adviced 
>>> to deployed DS-MIPv6
>> on
>>> the MR.
>>> 
>> First of all I am not sure that DS-MIPv6 currently supports IPv4 
>> network prefixes allocation (maybe it should). But even if it does,
>>  what you are suggesting is that the network operator providing 
>> services to that car is IPv6 capable. Is it not clear to everyone 
>> that in most of the world this is still not the case?
> 
> Mmm??  I believe that is incorrect.  I don't see why the network 
> operator has to be IPv6 capable.  The only part in the network that 
> has to be IPv6 is the home link, which can be virtual.  So you'll 
> need a MIPv6 HA, but the box it sits in doesn't even need a single 
> physical interface which has an IPv6 address.

Sorry for asking this again, but I can't really understand this.

I understand that some IPv6 home links can be virtual.  But how about
all the IPv6 real links that host IPv6 hosts.  If these IPv6 hosts need
to become mobile there is a need for a HA on that link with a real
physical interface, no?  I think in the case of the real IPv6 links
hosting IPv6 hosts that don't have IPv4 stacks DS-MIPv6 is little
adapted, no?

There seems to me to be two main concepts of using Mobile IPv6: one with
virtual home links (3gpp, mobile operators, HoA derived from MNP, etc)
and one with real home links.  I think both concepts should be supported
by any MIP6-related protocol.

Alex


Alex

> 
>>> It would be interesting to list down in which situation a 
>>> transition mechanims (DM-MIPv6 or another one) could apply or not
>>>  so that we
>> could
>>> clarify how many important scenarios would require a plain IPv4
>> solution. NEMOv4 would be useful for network operators that support
>>  mobile equipment (cars, trains, home routers or whatever) and
>> that, for whatever reason, do not support IPv6 but still want to
>> support network mobility.
> 
> (As indicated above) I believe this doesn't hold.
> 
>> I really do not like the fact that lately, to do any IPv4 work in 
>> the IETF, we are getting pressured to release customer information 
>> and deployment details. I will say the following which I hope will 
>> satisfy you: We have a number of customers (network operators) in 
>> various parts of the world that have deployed our FLASH-OFDM 
>> system. This is basically a MobileIPv4 based cellular data system 
>> (search for Flarion press releases if you want specifics...and 
>> sorry for the commercial).
>> 
>> Our customers would like to support network mobility. Some 
>> customers would also like to move to IPv6 some day, which is why we
>>  proposed DS-MIPv4 and DS-MIPv6 solutions. The two decisions, 
>> however, are not related and should not have to be coupled.
> 
> Fine.  But nevertheless, NEMOv6 with DSMIPv6 will let you support 
> network mobility without the operator having to support IPv6.
> 
> 
> Henrik
>