eap Mailing List: some comments to draft-adrangi-eap-network-discovery-and-selectio n-01.txt

[Tomas Goldbeck-Löwe (KI/EAB) (tomas.goldbeck-lowe]

Hi Farid and all,

Reading the new version of the draft about EAP based network discovery and 
selection. Sending this email to let you know that I'm quite happy with draft 
as it looks now, and I believe it will fit into the 3GPP-WLAN architecture.


Some questions though for my clarification:
1. Page 7, Decorated NAI; The text talks about that "It may include information 
for several Mediating Networks to be indicated on the route to the Home Service 
Network."
However, I don't seen anywhere else in the draft support when the Decorated NAI 
includes multiple MN's. I assume we then use the syntax 
'mediating-net-2!home-realm!username [at] mediating-net-1'. Is this correct?
Also, I my understanding we then must specify that a AAA supporting this 
functionality only moves the first realm in the username (i.e. if the decorated 
NAI looks like 'mediating-net-2!home-realm!username [at] mediating-net-1' the 
AAA in mediating-net-1 must "re-shuffle" the NAI to 'home-realm!username [at] 
mediating-net-2') before the packet is passed on. Or am I missing something?

2. Solution option 1 and 2 includes the con that "It MAY introduce a contention 
problem if space in the Type-Data field has already been used up for other 
purposes.  "
Can you explain why this is not true also for option 3?


and some editorials:
The following sections/paragraphs includes strange characters both on my screen 
and in my printout:
a. Page 2, Introduction, second paragraph; the sentence in the parenthesis 
starting with  "(i.e., öRoaming Partnerö ..."

b. Page 2, Section 1.1; "(referred to as the öEAP over RADIUSö [4]) "

c. Page 7, Access Point; "öA station that ..."
           RADIUS Server; "öThis is a server ..."
           Service Set ID; "öan identifier attached ..."

d. Page 13, "[Option 3] û Use a subsequent ..."

e. Page 17, Section 2.3, NAI Decoration; lots of funny characters...

f. Reference [10]; I believe it is missing the file name and 'work in 
progress', no?


Kind regards,
        --> Tomas