[Sigtran] MTP3b over M2UA?

["Brett Wallis" <[e-mail]>]

I'm new to this list, so perhaps this is a closed issue, but I was hoping to
get some feedback...

There are some current arguments amongst UMTS vendors as to whether M2UA was
meant to support MTP3b.  In our scenario, we have the UMTS Iu interface
which supports MTP3b/SSCF-NNI/SSCOP/AAL5/ATM and we want to terminate this
L2 (Q.2140) at the SG component of a MGW and backhaul this to the MGC via
MTP3b/M2UA/SCTP (more on this below).  So the argument is also whether
termination of Q.2140 by the SG component of a MGW is also valid in the M2UA
scenario.

I-D
http://www.ietf.org/internet-drafts/draft-ietf-sigtran-signalling-over-sctp-
applic-09.txt
<http://www.ietf.org/internet-drafts/draft-ietf-sigtran-signalling-over-sctp
-applic-09.txt>  supports M2UA transport for MTP3b (section 4.2.1.1), but
wording in RFC 3331 is being used against us.  This is mainly b/c RFC 3331
mentions MTP3b only in a section on SCTP management.  There is a reference
to Q.2140, but that's all.  Also, this I-D above has expired and I don't
know what its current validity or future status is.  

An argument used against us is that the primitives b/w MTP3b and Q.2140 are
different from the primitives between MTP3 and MTP2.  I'm not even aware
that MTP3-MTP2 interface primitives have been defined.  From what I can
gather, the only real difference between these two data link layers is that
some of the optional procedures of MTP2 are not supported by Q.2140.

Can anybody please provide opinions here?  It seems this role is exactly
what M2UA was created for, and I fail to see how Q.2140 on one side of the
MGW would prevent M2UA from being used here. 

Below is the specific argument that I had previously made for our case:


	RFC3331 clearly references MTP3b, which is for broadband transport.
RFC3331 as well as the other IETF document I referenced
(http://www.ietf.org/internet-drafts/draft-ietf-sigtran-signalling-over-sctp
-applic-09.txt
<http://www.ietf.org/internet-drafts/draft-ietf-sigtran-signalling-over-sctp
-applic-09.txt> ) clearly indicate that M2UA is applicable for MTP3b
transport.  RFC3331 even adds an informative reference to Q.2140 (B-ISDN ATM
Adaptation Layer), which is exactly what is used on the Iu interface.

	If so, then it should be agreed that the following is a supported
configuration for M2UA:

	<MTP3b user>/MTP3b/L2/L1 <---> SG component of MGW <---> <MTP3b
user>/MTP3b/M2UA/SCTP/IP/L2/L1

	The only remaining question is what are the allowed protocols for L2
on the left.  RFC3331 makes frequent reference to "MTP layer 2" which I
believe is a generic reference to the data link layer, and not specifically
to narrowband MTP2 as defined by Q.70x.  Even if they disagree with this
position, they must be able to identify the allowed protocols for L2 since
they have recognized that M2UA can carry MTP3b.

	MTP3b (Q.2210) is defined specifically to use the services of
Q.2140, which is the exact L2 supported on Iu.  So given that MTP3b/M2UA is
perfectly allowed by the IETF, and MTP3b is defined to run on
SSCF-NNI/SSCOP/AAL5/ATM, how can HW argue that this is not a supported
configuration?  L2 beneath MTP3b above must be Q.2140.  What other L2
protocols does HW believe MTP3b can use on non-SIGTRAN interface?  As Dan
pointed out, SAAL-NNI could just have easily been named MTP2b since it is
the MTP layer 2 protocol that is used with MTP3b.


I would greatly appreciate your feedback.

Thanks,

Brett Wallis
GSM/UMTS BICN Architecture
phone: [phone] (ESN 444)
email: [e-mail]