On Dec 12, 2007, at 10:32 PM, Simon Flannery wrote:
I would like to check my understanding of Early and Late media also
know as Early (or Late) operating mode:
- Using Early media, the inviting user may talk immediately, and the
core network should buffer the media before the invited user (or
users) accept the session
- The received media is buffered until at least the first invited user
accepts the invitation
- The buffered media is also sent to all users that accept the
invitation
- Conversely, Late media mimics a traditional phone call where the
inviting user must wait until at least one inviting user accepts the
session before being allowed to send media
Is my understanding correct? Are any of the dot points in error?
Welcome to one of the most confusing aspects among the many confusing
aspects of SIP that were introduced by our attempts to compensate for
the broken signaling model of ISUP. Personally, I think we
unwittingly introduced a nightmarish level of complexity and should
have just said "no! If you need two stage signaling, finalize the
first stage with a 200 OK and then re-INVITE if the SDP changes for
the second session." But we didn't do that, so we have what we have.
IETF seems to use the terms "early media" and "early session",
although I'm afraid I don't completely understand the nuances despite
having struggled with them for many years. I suspect I don't want to
understand just because the whole thing hurts my delicate
sensibilities. I'm told that RFC 3960 is a useful reference for this
discussion.
Both uses of "early" refer to the transmission of media before
completion of some phase of the signaling. Sometimes this is mean to
be "before completion of the offer/answer phase" (that is, after an
"offer", but before an "answer", and sometimes (as defined in RFC
3960) before the completion of a final response (generally a 200) to
an INVITE request.
RFC 3960 uses the following definition:
Early media refers to media (e.g., audio and video) that is
exchanged
before a particular session is accepted by the called user.
Within a
dialog, early media occurs from the moment the initial INVITE is
sent
until the User Agent Server (UAS) generates a final response. It
may
be unidirectional or bidirectional, and can be generated by the
caller, the callee, or both. Typical examples of early media
generated by the callee are ringing tone and announcements (e.g.,
queuing status). Early media generated by the caller typically
consists of voice commands or dual tone multi-frequency (DTMF) tones
to drive interactive voice response (IVR) systems.
In IETF and general IMS environments (AFAIK), there is no expectation
that early media will be buffered. Anything not received by a
listener is expected to be discarded.
However, OMA's Push to talk Over Cellular (POC) specification does
(or at least did, the last time I looked) allow a predictive mode
where a POC controller (which is for all practical purposes a SIP
B2BUA with media handling) can "answer" on behalf of a POC terminal
with which it has an active relationship. The POC controller then
buffers media while it activates the radio bearer needed to send the
media on to the terminating POC client. Since the POC controller has
effectively "answered" the INVITE request with a terminal response
and media answer, I don't believe this mode qualifies as "early
media" by any of the definitions in use around the IETF.
Good luck in sorting this all out.
--
Dean
_______________________________________________
Sip mailing list https://www1.ietf.org/mailman/listinfo/sip
This list is for NEW development of the core SIP Protocol
Use [EMAIL PROTECTED] for questions on current sip
Use [EMAIL PROTECTED] for new developments on the application of sip
