Raj, > > Changes: > > > > - Revised definition of "Serving Network". > > > > - Added an informational reference for 802.21 as an example mechanism > > for discovering PaC and CPAA to discover each other. > > > > Basavaraj questioned about the need for PAA-initiated pre-auth, and I > > agreed with his comment. But > > PAA-initiated pre-auth still remains, based on Alper's comment "PANA > > already supports PAA-initiated auth. So, unless we do anything > special > > (to eliminate this case), it comes for free." > > Even though PANA supports PAA-initiated auth, I do not believe this is > applicable in the case where the target network to which the host is > not > even attached would be able to trigger authentication. > So while the PAA-initiated auth makes sense in the regular PANA use > case, I > don't see this as being valid in the scenario described in the pre-auth > I-D.
As you know, there are examples of wireless network architectures that support network-initiated handovers. Such architectures are aware of the mobile's target network elements (BS, authenticator, etc.) and perform a number of signaling in the infra-side and also dictate the mobile to move. Such architectures are potentially capable of using network-initiated pre-auth. I'm inclined to keep PAA-initiated pre-auth in the spec because it has some potential use, it is harmless, removing it from spec is more work than keeping it, and it comes for free due to its built-in nature to the base spec. So, I don't see the benefit of removing it, or cost/complexity of keeping it. Maybe I'm missing something. Alper _______________________________________________ Pana mailing list Pana@ietf.org https://www.ietf.org/mailman/listinfo/pana