Handover Keying (hokey) Working Group MINUTES  

Chairs  : Glen Zorn (gwz@net-zen.net)

          Tina Tsou (tena@huawei.com)

Jabber  : hokey@jabber.ietf.org

URL     : http://tools.ietf.org/wg/hokey/

Agenda  : Version 0.1 (draft)

Meeting : IETF 78; Maastricht, Netherlands

Date    : Thursday, 29 July 2010

Time    : 15:10-16:10

Location: 0.1 London

=========================================================

Welcome & Administrivia (5 minutes)

Document Status

Glen Zorn

Draft: http://www.ietf.org/id/draft-ietf-hokey-ldn-discovery-01.txt

 

Discussion

Glen Zorn:

WGLC ends tomorrow

Need to send in comments immediately

Zhen Cao:

Noted his comment on the list that this might be better be an RA option.

Glen Zorn:

Suggested that could be done, but as another draft.

      It is correct that the title should be "DHCPv6 Option For ..."

 

Zhen Cao

     Draft: http://datatracker.ietf.org/doc/draft-ietf-hokey-erp-aak/

 

Discussion

Simon Mizikovsky:

What is the issue with CAP determination?

      Zhen Cao:

The problem is which CAP will receive the master key?

Tom Taylor:

Could not see what the problem was either, and thought it was a question of cryptographic separation of keys issued to different CAPs.

Glen Zorn:

Clarified that this was a question of whether a given CAP was authorized to provide service to the mobile node. The EAP/AAK server could not do the job, particularly if multiple AAA hops separate the server from the CAP.

 

Current Status of IEEE 802.21a

     Yoshihiro Ohba

 

Discussion

Qin Wu:

Asked whether Option III is related to AAA.

Yoshi:

Explained that AAA has to do with access control, and that may require centralized control. So there is a relationship between Option III and AAA.

Simon Mizikowsky:

Both schemes show key establishment on the target system.

Simon Mizikowsky:

Referring to chart 2, how is association established between serving and target PoS?

Yoshi:

Single domain situation is clear. Not so clear for inter-domain.

Simon:

But reference point is specified between Serving and Target PoS.  Surely this is acceptable only in when a security association is established between them.

Discussion cut short by Chair: belongs in 802.21a.

 

Qin Wu:

Asked how discovery is carried out.

Yoshi:

Clarified that this is part of 802.21a work, but just as an information element.  Main task is for 802.21.

 

Proposed New Charter Items

§        EAP Extensions for EAP Re-authentication Protocol (ERP)

Qin Wu

 

Discussion

No changes from 5296 so far.

Three key issues raised by errata.

 - requirement that ER server lie in the path of AAA messages

 - assumption that ER server can edit AAA messages

 - unnecessarily restrictive in deployment options

Open issues

 - Terminology issue

   Glen Zorn suggested simply calling it "re-authentication protocol".

 

 - Need to clarify description of NAI-keyname

 

 - Editorial issues with section 3.1. Need clear text to address ERP with home

   ER server.

 

 - Qin: re-authentication can only happen after attachment.

   Yoshi concerned that removal of text referring to re-authentication before

   attachment could hurt 802.21a work. Referred to list for discussion.

 

 - other issues

 

Follow up: resolve errata, issue new version, look for feedback from WG.

 

§        EAP Extensions for Early Authentication Protocol (EEAP) for Inter-AAA-Domain handover

Qin Wu

Out of time -- presentation limited -- discussion to the list

 

=========================================================