[Softwires] Per-interface NAT in mobile networks

2010-10-25 Thread mohamed.boucadair 
   Dear all,

   When per-interface NAT is applied to 3GPP networks, a NAT function
   can be embedded in the GGSN/PGW.  The GTP tunnel identifier will be
   used as an identifier in the NAT table to identify sessions belonging
   to each UE.  Packets are then translated and forwarded to their
   destination (either internal or external).  Distinct IPv4 address
   pools may be configured on the GGSN/PGW depending on the APN.  This
   procedure can be implemented in mono stack PDP context or in
   dual stack PDP contexts.

   Embedding a NAT in the PGW/GGSN is one of the scenarios we are
   considering to rationalise the use of IPv4 address in mobile
   networks, while promoting the invocation of IPv6 transfer
   capabilities whenever possible (our recommendation for mobile is to
   go for dual-stack connectivity).  Unlike other solutions where the
   NAT is not embedded in the PGW, the activation of the per-interface
   NAT does not impact the network availability since no new nodes will
   be introduced to embed the NAT function.

   We think per-interface NAT
   (http://datatracker.ietf.org/doc/draft-arkko-dual-stack-extra-lite/ or
   http://tools.ietf.org/html/draft-miles-behave-l2nat-00)
   is a good place to describe this use case.

   Some vendors already support a NAT function in their PGW products,
   but for harmonising the behaviour of such implementations we think it
   would be valuable to see draft-arkko-dual-stack-extra-lite be adopted
   as a WG document in Behave or Softwire.

   Is there any plan to see this I-D progress as Softwire or Behave WG
   document?

   Cheers,

   Med


*
This message and any attachments (the message) are confidential and intended 
solely for the addressees. 
Any unauthorised use or dissemination is prohibited.
Messages are susceptible to alteration. 
France Telecom Group shall not be liable for the message if altered, changed or 
falsified.
If you are not the intended addressee of this message, please cancel it 
immediately and inform the sender.


___
Softwires mailing list
Softwires@ietf.org
https://www.ietf.org/mailman/listinfo/softwires

Re: [Softwires] Per-interface NAT in mobile networks

2010-10-25 Thread Jari Arkko 

Mohamed,

Thanks for your interest on this!

For everyone's information: I have asked the document to be AD sponsored 
through to the IETF to become an RFC. I think Ralph was willing to do 
it, AD reviews and so on still obviously pending. So there may not be a 
need to adopt it at the working group. Review comments would be highly 
appreciated from everyone, of course.


Mohamed: were you thinking that we should have a also some more mobile 
network specific guidance about this? Perhaps that could be a separate 
draft (or even a 3GPP spec). I have been trying to keep my draft at the 
general level. But I do know that there are people who want to use it 
different circumstances, and some of those may benefit from additional 
specification. I know of at least one such case already, maybe yours is 
a second one.


Jari

___
Softwires mailing list
Softwires@ietf.org
https://www.ietf.org/mailman/listinfo/softwires