RE: comments : draft-ietf-v6ops-3gpp-analysis-00.txt

[<juha.wiljakka@nokia.com>]

Hi, Luc!

Thanks for your comments!  However, the UE really must be IPv6-only to the IMS direction. IMS case 1 can not be solved using DSTM - that solution would be "dual stacked" to the IMS direction.

Of course, otherwise the mobile can be dual stack. IPv4 stack can be used for e.g. Internet access / web surfing (using a separate IPv4 PDP context), i.e. for non-IMS-related traffic.

Best Regards,
	           -Juha-

-----Original Message-----
From: ext BELOEIL Luc FTRD/DMI/CAE
[mailto:luc.beloeil@rd.francetelecom.com]
Sent: 13 December, 2002 10:58
To: Wiljakka Juha (NMP/Tampere)
Cc: V6ops (E-mail)
Subject: comments : draft-ietf-v6ops-3gpp-analysis-00.txt


Hi Juha, 

Thanks to have update the draft !

I still have the same comments on 4.2 paragraph. 

1- a UE that uses IMS services must contact IMS devices over IPv6
(because IMS is IPv6-only). I agree.

2- Nevertheless a UE that uses IMS services could also be dual-stack
(even if it does not use its IPv4 stack to contact IMS devices)

3- As pointed out in section 2.2 and 3.1, DSTM could also be used so as
to send IPv4 datagrams over IPv6 in IPv6 PDP Contexts.

4- when that dual-stack UE is connecting to an IPv4-only node in a
IPv4-only network through IMS, I argue that NAT-PT can be avoided by
using DSTM.

I understand that DSTM might not be easy to set-up inside 3GPP networks
but that is an alternative to NAT-PT. 

Perhaps that 3GPP does not want that such transition mechanisms involve
UE but only network ?

Regards,

Luc