RE: [802.21] 802.16 Requirements Teleconference
Phil and all,
Just like GPCS in 80216g-05_008r1, MIH CS is one of many CS options.
MIH messages going over CS is actually not management and control traffic
from 802.16's point of view. It is just data to be delivered to MIH function
which is sitting above CS.
In 802.21 draft, there are two mechanisms defined to deliver MIH related
messages over the air interface one is over data plane (over CS) and the
other is over management plane (C-SAP or M-SAP). CS extension is just to
satisfy one of two options.
Transmitting MIH messages (new MIH ethertype) will not, I believe, impose
much complexity to CS. It is for right service flow setup and
classification.
Best Regards,
Ronny
-----Original Message-----
From: Phillip Barber [mailto:pbarber@BROADBANDMOBILETECH.COM]
Sent: Monday, November 07, 2005 10:50 AM
To: STDS-802-21@LISTSERV.IEEE.ORG
Subject: Re: [802.21] 802.16 Requirements Teleconference
16g was created to bring management plane into scope and absorb and apply
recommendations of 802.21 (one of 16g tasks).
802.16 identified Management and Control traffic don't go over the CS_SAP,
ever. Management and Control traffic go over their respective SAPs in the
Management Plane (M_SAP & C_SAP). This does not add new complexity. This has
always been intended.
Thanks,
Phillip Barber
Huawei
----- Original Message -----
From: "Junghoon Jee" <jhjee@ETRI.RE.KR>
To: <STDS-802-21@listserv.ieee.org>
Sent: Sunday, November 06, 2005 9:43 PM
Subject: RE: [802.21] 802.16 Requirements Teleconference
> So sorry for the previous mistaken message.
> Here is what I intended to express.
>
> --------------------------------
> Hello,
>
> Question:
> Why are we specifying a new CS extension? in 3.6?
> In 802.21 draft we are defining two methods to transfer MIH messages. One
> is
> using data plane (new Ethertype) and the other is using control plane. CS
> extension in 3.6 is to cover one of two options. Even though we have new
> ethertype we should have CS extension for right service flow setup.
>
> Why not use secondary management connection that does not cross the
> Convergence Sublayer.
> [Ronny] You are right. We can, of course, use not only secondary
> management
> connection but also primary management connection or basic connection in
> order not to cross Convergence Sublayer. It is more likely to transport
> MIH
> messages through Primary or Basic connection than Secondary management
> connection. In 802.16 requirement and amendments document, we have a way
> to
> transfer MIH messages through primary and secondary management connection.
>
> [Junghoon] During the previous conference call for 802.16 requirement
> works,
> it is known that amending the CS SAP is fallback case when amending the
> C_SAP/M_SAP does not be done immediately and successfully. Thus, I believe
> that we do not need to make explicit statement of requiring amendment of
> both the C_SAP/M_SAP and the CS SAP.
> Requiring amending both SAPs might result in burden for IEEE 802.16g folks
> to take the 21 requirements.
> Is there any opinion who thinks that we need to still require amending CS
> SAP when amending C_SAP/M_SAP is done successful ? Please let me know what
> I
> am missing.
>
>
> I am saying it because I know the dynamics in .16 and particularly in
> WiMAX
> profiles which will drive the product certification. Currently the forum
> can't even agree on having more than one CS type. Just one CS type for
> now,
> imagine? CS is a very sensitive area and we should avoid it if we can.
>
> [Junghoon] I agree with Peretz's concern here.
>
> Regards,
> -Junghoon
>
> [Ronny] I understand your point. However, we can not think of WiMAX
> profile
> yet because simply 802.21 technology is not ready to discuss WiMAX
> profile.
> Maybe in WiMAX profile stage 2 or in later stage we may be able to start
> discussion about 802.21 technology. Just like we have different CS types
> in
> 802.16 specification, we need to define MIH CS type in 802.16 spec and we
> have to see what we can do in WiMAX.
>
> Besides, CS will be used for packet classification. Do we need to classify
> MIH and assign it a unique CID? why not stick to SMC?
> Peretz Feder
>
>
>
> Regards,
> -Junghoon
>