>In general there were lots of questions and some issues
identified as well around Information Service and IEs.
>Which IEs are really applicable to different access
networks and which are more independent of different access networks?
[Cheng]
I guess this has to be sort out by the .21 before passing
any requirements to the media specific groups. Also it relateds to the question
below. If the "support" of IE is just about providing transport means
from PoA to MT, the question is a non-issue. However, if the
"support" is also about the media specific technology also
providing such info to the .21 IS Server, different requirements are implied.
[/Cheng]
[Vivek G Gupta] Communication between PoA and .21 server is
probably gonna happen at L3 and anyway such requirements are out of scope of
802.21 and probably even access technology standard such as 802.11. The
requirements for providing transport between PoA to Mobile Terminal need to be
clearly articulated and addressed in specific access technology standard (such
as 802.11).
As for availability of IEs, the information represented by
IE should be available from that access technology standard in some way. For
example in case of 802.11, whether an 802.11 network supports .11i/.11e etc. is
already defined and hence these probably don’t impose any new
requirements on 802.11. The 802.21 IS can use these already defined methods to
gather this information. My guess is that from 802.21 perspective we are mostly
unlikely to come up with link technology related IEs for which information is
not already available from different access networks.
>If IEs are retrieved at L2 from PoA, is there an
implicit assumption that the PoA knows how to communicate with 802.21
Information Server? Is this a
>reasonable assumption?
[Cheng]
Does this also mean an interface from .21 IS server to be
defined so that PoAs or others can access? Sounds this requires some more work
at the .21 side.
[/Cheng]
[Vivek G Gupta] After talking to some other folks, this may
turn out to be a reasonable assumption. The definition of interface between
PoAs and .21 server is out of scope of both 802.21 and possibly even the PoA
specific technology standard.
L2 access for IEs between Mobile
Terminal and PoA still seems reasonable because the MT may not have established
a L3 connection and may need this information before establishing a L3
connection.
>Do we need to specify which IEs can be supported in
unauthenticated state and which cannot?
[Cheng]
Probably we need to specifyat which stage an IE may be
exchanged, and which type of protection should be provided, e.g.
authenticated. It is up to the specific media group to decide how to provide
such protection. (or they may reply that such thing cannot be supported, then
.21 probably needs to change the assumptions)
[/Cheng]
[Vivek G Gupta] All IEs should be supported by .21 server
and not by PoA (the request terminates at and is eventually satisfied by .21
server and not by the PoA which merely forwards the request to get the value of
an IE…)
There may be security issues associated with transport and
these may apply to all IEs.
What IEs cannot be retrieved in unauthenticated state needs
to be specified by .21 as well. The rules for this may well be based on how the
specific access technology makes available the information related to specific IEs.
Thanx for your comments and responses.
Best Regards
-Vivek
From: Gupta,
Vivek G [mailto:vivek.g.gupta@intel.com]
Sent: Wednesday, August 17, 2005
11:01 AM
To: Cheng Hong;
STDS-802-21@listserv.ieee.org
Subject: RE: 802.21
Teleconference: L2 Requirements for 802.11
From: Cheng Hong
[mailto:Hong.Cheng@sg.panasonic.com]
Sent: Tuesday, August 16, 2005
7:33 PM
To: Gupta, Vivek G;
STDS-802-21@listserv.ieee.org
Subject: RE: 802.21
Teleconference: L2 Requirements for 802.11
Thanks a lot for the document. Sorry I
could not make for the teleconf yesterday, but I got a few questions (sorry if
they are already clarified during teleconf)
1) In Table 1, what is the use case for
event 9 to 11? What is the defintion of Handoff? Is this a lower layer thing or
application layer thing?
[Vivek G Gupta] In my own opinion events 9-11 seem
redundant/duplicate and may be eliminated from later version of 802.21 drafts.
2) Regarding table 2, in the column of
"MIHF<>MIHF", are those indicating bi-directional relationship
or unidirectional?
[Vivek G Gupta] believe these are Bidirectional..
E.g. when it is
"Client<>Network PoA", does it mean "From Client to
Network PoA", or "between Client and Network PoA"?
[Vivek G Gupta] Between client and network PoA. The
command could be sent by either the client or network PoA depending on client
initiated or network initiated handovers. Current ongoing architectural
discussions may have impact on these.
3) The table 3, are these intend to be the
payload to be carried by the 11 protocol, or are they the IEs to be
defined in 11?
[Vivek G Gupta] We discussed these in detail and most of the
meeting time was spent on Information Service and IEs. These are IEs and some
of these (those relevant to 802.11 MAC/PHY need to be defined by 802.11).
Others are more general and independent of MAC/PHY…
4) In table 3, item 2, why the 11 network
needs to indicate support of other type of networks,e.g. 16 and GSM, etc. And,
is the intension to have all the network types in the same namespace instead of
subgroups, e.g. 11-subtypes, 16-subtypes, cellular-subtypes...etc?
[Vivek G Gupta] We may need to organize this information
better.
This particular information may not directly come from .11
access network but from 802.21 Information Server. 802.11 may not need to do
any thing in this case.
5) In table 3, item 6, what does this
mean? Is this the location of the PoA or the Terminal? In 11, if this is the
terminal location, it may mean a different requirement on the 11 specification,
ie. to provide meansurement capability for an accurate terminal location.
[Vivek G Gupta] It’s the location of PoA…
6) Table 3, item 7, can this support both
IP versions?
[Vivek G Gupta] Not sure what you mean here..maybe the Type
field (Enumerated) is incorrect, but again 802.11 may not need to do anything
for this IE.
7) Table 3, item 9, for a 11 network,
probably the most important type is "Generic IP Service". Also, not
sure about how the "supported" means, e.g. "VPN" has many
types..
[Vivek G Gupta] Yes we did talk about these and again this
information may come from 802.21 Information server and 802.11 may not need to
define anything.
In general there were lots of questions and some issues
identified as well around Information Service and IEs.
Which IEs are really applicable to different access networks
and which are more independent of different access networks?
What does it mean for an access network to provide support
for an IE?
If IEs are retrieved at L2 from PoA, is there an implicit
assumption that the PoA knows how to communicate with 802.21 Information
Server? Is this a reasonable assumption?
Do we need to specify which IEs can be supported in
unauthenticated state and which cannot?
From:
stds-802-21@ieee.org [mailto:stds-802-21@ieee.org] On Behalf Of Gupta, Vivek G
Sent: Tuesday, August 16, 2005
3:24 AM
To: STDS-802-21@listserv.ieee.org
Subject: RE: 802.21
Teleconference: L2 Requirements for 802.11
Attached is an initial document listing
some of the requirements for 802.11 media type.
We shall discuss these requirements
tomorrow and then update them further.
Please feel free to send in your comments
and suggested amendments.
Best Regards
-Vivek
From:
stds-802-21@ieee.org [mailto:stds-802-21@ieee.org] On Behalf Of Gupta, Vivek G
Sent: Monday, August 15, 2005 6:54
AM
To: STDS-802-21@listserv.ieee.org
Subject: 802.21 Teleconference: L2
Requirements for 802.11
The next 802.21 telecon on L2 Requirements for 802.11 will
be held as per below.
Please plan on joining us then.
Date: August 16, 2005, 9:00 AM EST
Phone: 916-356-2663, Bridge: 3, Passcode:
0566976
Lines: 25
Best Regards
-Vivek