Gateway Decomposition Call Summary and Next Meeting Request

[Mark Reid]

Gateway Call Summary 10/20/98 11AM-1PM EDT

Attendees registered for call
***********************************
Bellcore, AT&T, Cisco, Nortel, Lucent, Intel, Siemens, Level 3, Ascend,
3COM, IBM, Telia, MCI, NEC, PictureTel, Sun Microsystems, RADVision, Accord,
Selsius, Madge, AGCS, IDT/Net2Phone, VTEL, Dialogic, NetSpeak, Delta
Information, Ricochet, Trillium Digital Systems, Teleogy Networks, Univ
Bremen, Perey Consulting, VideoServer, others

Call purpose summary
***************************
The goal of the audio call was to identify the work items we need to
complete heading into the November Q14 meeting on H.323 Gateways
interworking with switched circuit networks.

Agreements reached on call
**********************************
1) Interface A is a high priority item. It is a new protocol and the first
item of work should be the definition of its scope and requirements. There
was agreement that Interface A should be media related and not be associated
with call setup. It was agreed that Interface A would deal with the
allocation and management of resources used for packet/circuit media
conversion (e.g. DSP's for modem, fax, transcoding, etc.) It was agreed to
keep the scope of Interface A within the terms of reference outlined in
paragraph 3.4.1 of TD89. For example, we are not trying to replace existing
H.323 functionality with a new protocol.
2) We will specify the Interworking of H.323 to SS7/ISUP signaling
(potentially something similar to Q.699) and the Interworking of H.323 to
SCN FAS/NFAS signaling. This work will most likely be contained in an H.246
Annex.
3) We will complete H.246 Annex B - the media conversion of H.323 to
voice/voiceband terminals. (e.g. DTMF to UUI, etc.)
4) We agreed to exclude the MC to MP communication definition from the work
destined for determination at the next Study Group meeting.

Active discussions
**********************
1) What should we do with Interfaces B, C, and D?

Given that the messages passing across these interfaces must reflect all of
the content of the signaling that the respective signaling functions receive
from and send to external entities ... are these protocols modified or
extended ISUP for Interface D, modified or extended Q.931 or QSIG for
Interface C, and modified or extended H.225 for Interface B? ... or are they
a new abstract protocol (ABP) that ISUP, FAS/NFAS, and H.225 are converted
to?

The proponents of the abstract protocol said it would allow the high level
gateway control logic to be shielded from the H.225/FAS-NFAS/ISUP state
machines. Others said that too much information would be lost in a
conversion between H.225/FAS-NFAS/ISUP and an abstract protocol (ABP) and
that the ABP would end up having all of the fields carried on the H.225,
FAS-NFAS, and ISUP interfaces.

No conclusion was reached for moving forward.

Nortel suggested with look at specific call scenarios to drive the
requirements of Interfaces B, C, and D.

General information
**********************
BellCore reported that IPDC and SGCP have merged their protocol proposals as
MGCP. MGCP was introduced into the IETF on 10/19 and according to Nortel, a
working group is being kicked off in the IETF. There was a discussion on how
to get the IETF and ITU-T to work together on these protocols but no
conclusion was reached. MGCP information will be posted to the ITU
reflector.

Next meeting
***************
I would like to invite you to participate in an audio call, Tuesday November
3rd from 11AM to 1PM EST to discuss our continued work plan for gateway
decomposition.

Please send bridge port requests directly to me (mreid at videoserver.com) by
Monday November 2nd at 12:00 Noon EST so that I can make the proper
arrangements. I will send port assignments later that day.

All comments on this summary are welcome!

        Mark