A scenario that cannot be solved without APC-1280 or deferred "Connect"

Scott Petrack Scott_Petrack at VOCALTEC.COM
Tue Sep 9 12:32:36 EDT 1997


Discussion in Sun River brought out that a careful reading of H.225, H.245,
and H.323 implies the following:

1. It is allowed to start H.245 negotiation as soon as the CallSetup
message is received.
2. Once  H.245 negotiation begins, it can proceed to open logical channels
at it's own pace.
3. Once an open logical channel is opened and acknowledged, media streaming
should begin right away.

So it seems to be entirely possible to do media streaming before a CONNECT
message, and the CONNECT message can retain its usual Q.921 meaning.

You can call this "deferred Connect", but it'd be better not to give it any
special name. It's just a compliant H.323 implementation. The CONNECT isn't
deferred, it comes at just the right time -- when the endpoint wants to
signal that the call has connected.... This may or may not be after some
media has been streamed.

Scott Petrack
Principal Technologist
VocalTec Communications Ltd.
petrack at vocaltec.com

btang at LUCENT.COM on 09/09/97 04:50:36 PM

Please respond to ITU-SG16 at mailbag.jf.intel.com

To:   ITU-SG16 at mailbag.jf.intel.com
cc:    (bcc: Scott Petrack)
Subject:  Re: A scenario that cannot be solved without APC-1280 or deferred

Your observation leads to the general question that whether it is necessary
support the opening of a media channel during the Q931 call signaling
procedure. This concerns other scenario as well. For example, when an H323
endpoint calls a POTS user via a gateway, while the Q931 call signaling is
in progress it may be required to have a voice channel opened between the
endpoint and gatweay to forward a PSTN special announcement ("sorry, the
you dialed is not in service") back to the endpoint. As I remebered, a
while ago there was a decision made from this subject group about h323
implementation, that no capability exchange will take place before the
Q931 signaling is completed. I am not sure about the rational behind it,
but maybe that can be brought up again for discussion.
I looked at APC-1280 and found no descriptions for the procedure of using
the Progress-UUIE which is proposed to solve the above-mentioned problem.
Will this be added later on?
Ben Tang
Lucent Technologies
btang at lucent.com

More information about the sg16-avd mailing list