[sip-h323] SIP-H.323 IWF protocol requirements

[Paul E. Jones]

Jean-Francois,

Here are a few benefits for using v3:

H.323v3 introduces new fields to allow the GK to signal whether Annex E
should be used or not.  For build large-scale devices, Annex E is definitely
worth considering.  (Annex E was written for V2, but related RAS fields were
added in v3.)

It is possible to use the same TCP connection for multiple calls.

Text was added to explain the use of Lightweight RRQs (this may have been in
the IG, but I can't recall).  A number of things are better explained,
revisions have been made via the IG that may result in interworking issues
if the latest texts (including the IG are not followed).  Indeed, things
should be backward compatible, but the clarifications are important, as they
were obviously points of confusion.

Version 4 has even more.  Here's a few:
  * Gateway decomposition section
  * Support for new services and features
      * Annex L, Annex K, H.450.x, etc.
      * Generic Extensibility Framework
  * Additive Registrations
  * Alternate Gatekeeper usage fully defined
  * Usage Information Reporting
  * Endpoint Capacity Reporting
  * Desired protocol indication
  * Reporting call status with multiple IRRs
  * Call Linkage
  * Early H.245
  * Support for pre-paid calling cards

You can definitely see some advantages to v4.  Probably the most important
introduction in V4 is the Generic Extensibility Framework.  With that, we
can now add new features to the H.323 protocol specification that the
signaling entities in the middle of the call can remain completely ignorant
about.  This allows us to add new features that work end to end without
forcing the intermediate equipment to be updated-- that's one very good
reason for moving to v4.  It's quite possible that v4 may serve as the
minimum specification for some time to come.

Of course, this was possible before with "non-standard" mechanisms, but it
is hoped that the GEF will introduce fewer ripples in the H.323
standardization effort and the software development cycle.

This extensibility framework may also serve well for feature interaction
between SIP and H.323.  I realize that feature interaction is not an
immediate goal, but it's nice to have the foundation in place to try to
utilize it.

That's why I suggested at the meeting to go with the latest: v4.  It's true
that the differences between v2 and v3 are small, but a few scalability
issues were addressed.  Since signaling interworking is not a major CPU
consumer, I would think you'd want to take every possible step to increase
the number of calls handled on a single box. That could easily go over the
number of sockets the box can support.  So using Annex E or carrying
multiple calls over a call signaling channel is desirable.

Of course, I realize that NetMeeting does not support these features, but I
think that heavier consideration should be given to other H.323 entities.  I
cannot recommend trying to offer residential phone service using H.323v2
where SIP interworking is required: it's not nearly as scalable as v3 since
SIP interworking forces all calls to be directed to one box or a few boxes.
SIP works well there, but one really needs Annex E/H.323 to get comparable
scaling ability from H.323.

Paul

----- Original Message -----
From: "Jean-Francois Mule" <jfmule at clarent.com>
To: <paulej at packetizer.com>
Cc: "Wang, Dave" <dwang at nuera.com>; "aHit! IMTC (E-mail)" <ahitag at imtc.org>
Sent: Tuesday, August 29, 2000 12:53 PM
Subject: RE: [sip-h323] SIP-H.323 IWF protocol requirements


> Paul,
> Based on the current aHit IWF requirements, could help the group identify
> what particular h323v3 functionality would be mandatory for the IWF.
> If we can meet the requirements with v2, then we should mandate v2 which
> means also that v3 implementations will meet the requirements since there
is
> some backward compatibility.
>
> Jean-Francois Mule
> Clarent Corp.
>
> > -----Original Message-----
> > From: Wang, Dave [mailto:dwang at nuera.com]
> > Sent: Tuesday, August 29, 2000 8:59 AM
> > To: aHit! IMTC (E-mail)
> > Subject: FW: [sip-h323] SIP-H.323 interworking Requirements Draft
> >
> >
> > > -----Original Message-----
> > > From: Paul E. Jones [mailto:paulej at packetizer.com]
> > > Sent: Monday, August 28, 2000 11:53 PM
> > > To: Roy, Radhika R, ALCOO; sip-h323 at egroups.com;
> > > ITU-SG16 at mailbag.cps.intel.com
> > > Subject: Re: [sip-h323] SIP-H.323 interworking Requirements Draft
> > >
> > >
> > > Radhika,
> > >
> > > A few months ago, I published the final set of ASN.1
> > > corrections for H.323v3
> > > to the ITU-T SG16 mailing list and to the IMTC H.323 mailing
> > > list.  I also
> > > contact 11 companies that have H.323v3 products or products
> > > in development
> > > that I was aware of.
> > >
> > > There are certainly plenty of v3 implementations in
> > progress and it's
> > > important to be prepared for those implementations.  H.323v4 will be
> > > approved in November and I expect even quicker adoption of that
> > > Recommendation, because it introduces a lot of new functionality.
> > >
> > > I think what the group should try to strive for is a
> > > specification that will
> > > work with the most recent version of H.323-- H.323v4.  Of
> > > course there will
> > > be few implementations, but there are not a large number of
> > > new mandatory
> > > features.  (I don't have a list handy.)
> > >
> > > For a "bigger picture" view of what's new in the various
> > > versions of H.323,
> > > see:
> > > http://www.packetizer.com/iptel/h323/
> > >
> > > Best Regards,
> > > Paul
> > >
> > > ----- Original Message -----
> > > From: "Roy, Radhika R, ALCOO" <rrroy at att.com>
> > > To: "Roy, Radhika R, ALCOO" <rrroy at att.com>; <sip-h323 at egroups.com>
> > > Sent: Monday, August 28, 2000 10:32 PM
> > > Subject: RE: [sip-h323] SIP-H.323 interworking Requirements Draft
> > >
> > >
> > > > Hi, Everyone:
> > > >
> > > > Here is my view as stated below:
> > > >
> > > > It is very difficult to find any vendor to support H.323
> > > version 3.  Many
> > > > companies are now testing vendor implementations of H.323
> > > version 2, and
> > > it
> > > > is also reported that multivendor interoperability at the
> > > version 2 level
> > > is
> > > > still problematic.  In many cases, no promise has been
> > > obtained from the
> > > > major vendors when they will support version 3.  It may be
> > > very nice to
> > > have
> > > > SIP/H.323 v3 interworking, but we think SIP/H.323 v2
> > > interworking should
> > > > come first.
> > > >
> > > > Considering all factors, I think that we should keep our first
> > > interworking
> > > > requirements to support H.323 version 2 as defined in the
> > > present draft
> > > > (while we may consider for H.323 version 3 in the future).
> > > >
> > > > So, I recommend that we should make any changes in the
> > > present draft.
> > > >
> > > > Best regards,
> > > > Radhika R. Roy
> > > > AT&T
> > > > +11 732 420 1580
> > > >
> > > > -----Original Message-----
> > > > From: Roy, Radhika R, ALCOO [mailto:rrroy at att.com]
> > > > Sent: Sunday, August 27, 2000 10:32 PM
> > > > To: sip-h323 at egroups.com
> > > > Subject: [sip-h323] SIP-H.323 interworking Requirements Draft
> > > >
> > > >
> > > > Hi, Everyone:
> > > >
> > > > I have just returned from the ITU-T SG16 Q.13 (H.323)
> > > meeting where our
> > > > SIP-H.323 Interworking requirement draft was also discussed.
> > > >
> > > > The Q.13 members have strongly recommended that we should
> > > consider H.323
> > > > version 3 (instead of version 2) for our SIP-H.323
> > > interworking work.
> > > >
> > > > We would appreciate if you would kindly provide your
> > comments on the
> > > point.
> > > >
> > > > Best regards,
> > > > Radhika R. Roy
> > > > AT&T
> > > > +1 732 420 1580
> > > > rrroy at att.com
> > > >
> > > >
> > > > To Post a message, send it to:   sip-h323 at eGroups.com
> > > >
> > > > To Unsubscribe, send a blank message to:
> > > sip-h323-unsubscribe at eGroups.com
> > > >
> > > >
> > > > To Post a message, send it to:   sip-h323 at eGroups.com
> > > >
> > > > To Unsubscribe, send a blank message to:
> > > sip-h323-unsubscribe at eGroups.com
> > > >
> > > >
> > > > To Post a message, send it to:   sip-h323 at eGroups.com
> > > >
> > > > To Unsubscribe, send a blank message to:
> > > sip-h323-unsubscribe at eGroups.com
> > > >
> > >
> > > -------------------------- eGroups Sponsor
> > > -------------------------~-~>
> > > Need EDA tools on a short term or peak load basis?
> > > Take a free 7 day trial!
> > > http://click.egroups.com/1/8464/7/_/302437/_/967548702/
> > > --------------------------------------------------------------
> > > -------_->
> > >
> > > To Post a message, send it to:   sip-h323 at eGroups.com
> > >
> > > To Unsubscribe, send a blank message to:
> > > sip-h323-unsubscribe at eGroups.com
> > >
> >
>

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
For help on this mail list, send "HELP ITU-SG16" in a message to
listserv at mailbag.intel.com