Re: Call Forward too!... RE: AW: Call hold and transfer in H.323 Anne xF. Too limited??
Karl,
I do not see a limitation in H.450.3.
If a called endpoint B is out of service (which is very likely for a PC based H.323 endpoint but which may be more an exceptional case for other endpoint types like "black phones"), a proxy function has to act on behalf of the endpoint B, provided you want to successfully establish the call. This is not specific to H.450.3 call diversion but applies to H.323 Basic call functions as well.
I disagree. As far as basic call functions are concerned (ie no forwarding is set up) B's gatekeeper will reject any ARQ trying to reach B, and if it is routing H.225.0 and receives a setup message without having had an ARQ, fail the call by some suitable means. The point IS specific to call forwarding. Without any forwarding set up, I don't see what a "proxy function" is going to do.
This proxy function may e.g. be co-located with the gatekeeper or may be located within an other endpoint (with the gk just performing routing to the proxy). The latter one may allow more flexibility and scalability. Anyway, this is an issue of your architecture and does not need to be standardized.
Agreed. The salient point here is that we agree the gatekeeper needs knowledge of what's going on (either to provide the service or to perform the routing to an external proxy).
As an example, FIGURE 5/H.450.3 shows the forwarding function (i.e. the "served signalling entity") possibly being provided by a proxy function (e.g. for cases where the endpoint is out of service). Of course the proxy function needs knowledge about the activated supplementary services for a given user for which the proxy is acting, which again is an implementation issue.
I disagree with your last sentence here. The way the endpoint communicates "all my calls need to be forwarded to Joe in my absence" to its gatekeeper is certainly NOT just an implementation issue. An endpoint implementor ought to be able to implement a dialogue box on his endpoint, which will set up call forwarding whatever gatekeeper it uses (as long as the gatekeeper has this functionality). (S)he will not be able to do this if that particular communication is not standardised in some way - which seems to me to be a pretty good measure to indicate that it is a standardisation, rather than implementation, issue.
As for whether communications between the gatekeeper and the "proxy function" need to be standardised, this is less clear (in my opinion) as it seems a significantly less restrictive requirement for people to need to buy "proxy function" and gatekeeper from the same vendor than to need to buy endpoint, "proxy function" and gatekeeper from the same vendor. However, given all the activity on gateway decomposition, I can imagine that there will be people who disagree.
Regards, Chris
-----Ursprüngliche Nachricht----- Von: Chris Purvis [SMTP:Chris.Purvis@MADGE.COM] Gesendet am: Montag, 22. März 1999 12:07 An: ITU-SG16@mailbag.cps.intel.com Betreff: Call Forward too!... RE: AW: Call hold and
transfer in H.323
Anne xF. Too limited??
Frank, Karl, Anyone else who's interested,
A limitation in H.450.3 (Call Forward) recently became
clear to me, and
the solution may arrive through the sorts of things Frank's suggesting.
Background: What I see as a problem arises from the endpoint-centric
nature of CFU
(Call Forward Unconditional), especially in the case where the
endpoint being
forwarded is PC-based (probably NOT a SET).
The specific problem: I would regard as a prime use for CFU to be the ability to set up forwarding and then switch off my PC, so that all my calls would then go (for example) to a colleague. Everything seems to be there in H.450.3 to
allow this to
happen in the gatekeeper-routed model (obviously the
GK-routed model is
likely to be required, because the GK appears to be the
only "live" entity
that can do the forwarding), with the exception of the
transaction between
my PC and its gatekeeper.
The solution: I don't know. Maybe an H.450.3 extension (which has
dangers because a GK
is not normally a callable entity), maybe a new RAS
transaction type (but we
don't want proliferation of RAS messages)?
The plea: I'd like to see some serious suggestions on this, and
really drive some
solution towards standardisation, either as an Annex to H.450.3 or elsewhere. It seems to me that this is possibly the most important application of Call Forwarding, so it seems a shame that it
is missing.
If consensus works towards a RAS-based solution, we may be
able to do this
in a way that will also suit Frank's ideas?
Regards, Chris
-----Original Message----- From: Derks, Frank [mailto:F.Derks@PBC.BE.PHILIPS.COM] Sent: 22 March 1999 10:43 To: ITU-SG16@MAILBAG.INTEL.COM Subject: Re: AW: Call hold and transfer in H.323 AnnexF.
Too limited??
Folks,
when talking about a Simple Endpoint Type, I think we should aim for it to be something that closely resembles a black phone. This way it becomes a lot easier to define what its capabilities are and it makes life easy on the users and on those companies that will actually make (physical) IP-phones. These phones should probably look and act like the normal phones that are currently being used. Looking at how most supplementary
services are
accessed in both the public and the private (PBX) networks, I think it is safe to say that in most cases we are talking about "stimulus protocols". I.e. DTMF digits are sent to an exchange and the exchange interprets certain digit sequences as being the invocation of some service rather than a number to be dialled. The big advantage over functional protocols (like H.450.x) being that services can be added from the exchange side,
without the
terminal having to be modified as well.
Functional protocols never became a success in the ISDN world and this may end up to be the same in the IP world. However, having said this, there is a lot more potential for easy upgrading of e.g. terminal software in this domain, which reduces the side effects of functional protocols.
It does not seem to make sense to define "alternative" mechanisms to provide services, so I would strongly opt for using H.450.x when possible and using a simple stimulus protocol otherwise. The latter would
allow service
providers to easily make services available and I see no reason why this should be standardised. In practice, today, we already see that certain digit sequences for service activation are identical in several countries.
Frank
Frank Derks |Tel +31 35 6893238 | Advanced Development |Fax +31 35 6891030 | Philip Business Communications |P.O. Box 32 | |1200 JD Hilversum | |The Netherlands | ----------------------------------------------------| E-mail: mailto:f.derks@pbc.be.philips.com | WWW: http://www.business-comms.be.philips.com |
-----Original Message----- From: Klaghofer Karl ICN IB NL IP 7 [mailto:Karl.Klaghofer@ICN.SIEMENS.DE] Sent: 18 March 1999 22:36 To: ITU-SG16@MAILBAG.INTEL.COM Subject: AW: AW: Call hold and transfer in H.323 AnnexF.
Too limited??
See comment below.
Karl
-----Ursprüngliche Nachricht----- Von: Paul E. Jones [SMTP:paul.jones@TIES.ITU.INT] Gesendet am: Donnerstag, 18. März 1999 18:57 An: ITU-SG16@mailbag.cps.intel.com Betreff: Re: AW: Call hold and transfer in H.323
AnnexF. Too
limited??
Karl,
Unfortunately, I will have to disagree with your comments.
While it is
true that the H.450 supplementary services could be utilized in
a SET device, I
believe that introducing H.450 into a SET breaks the spirit
of that work.
The goal of Annex F is to define a "Simple Endpoint Type".
There are
simpler ways to put a call on hold and to transfer a call.
Introducing
H.450 introduces a lot more complexity that I believe we
want to have. If
Annex F is not sufficiently clear on how to simply transfer
a call or put
a call on hold, we should work on that text-- I will
absolutely disagree
with introducing H.450 into a SET device.
[Klaghofer, Karl PN VS LP3] Whatever you mean with"introducing" - H.450 as I sayd in my previous mail is a way of providing supplementary services like call hold and call transfer to a SET device. It IS already part of the H.323 Annex F!
Paul
-----Original Message----- From: Klaghofer Karl ICN IB NL IP 7
To: ITU-SG16@MAILBAG.INTEL.COM ITU-SG16@MAILBAG.INTEL.COM Date: Wednesday, March 17, 1999 3:27 PM Subject: AW: Call hold and transfer in H.323 AnnexF.
Too limited??
Gunnar,
You are referring to call hold and transfer in conjunction
with H.323
Annex
F SETs (Audio or Text) and clause 7.6 of H.323 Annex F.
Talking about call hold, clause 7.6 of H.323 Annex F is
not needed for a
SET
at all. Call Hold works for a SET as it is defined in H.450.4.
Talking about Call Transfer, clause 7.6 of H.323 Annex F
is not needed
for a
SET, if the transfer is executed by the endpoints as
defined in H.450.2.
Codec re-negotiation you are referring to is no problem
and takes place
between the transferred and the transferred-to endpoint.
This may cover
your
case with wireless endpoints being involved.
For call transfer, section 7.6 of H.323 Annex F is only
needed if the
gatekeeper or a proxy acts on behalf of the transferred
SET endpoint B.
However, media re-negotiation also should work here as
part of the
fastStart
method.
Regards, Karl
Karl Klaghofer, Siemens AG, Dpmt. ICN IB NL IP7 Hofmannstr. 51, D-81359 Munich, Germany Tel.: +49 89 722 31488, Fax.: +49 89 722 37629 e-mail: karl.klaghofer@icn.siemens.de
-----Ursprüngliche Nachricht----- Von: Gunnar Hellstrom [SMTP:gunnar.hellstrom@OMNITOR.SE] Gesendet am: Dienstag, 16. März 1999 23:01 An: ITU-SG16@mailbag.cps.intel.com Betreff: Call hold and transfer in H.323 AnnexF.
Too limited??
Dear multimedia experts.
In my efforts to establish the simple IP voice and text
telephone Text
SET, I came across a section in H.323 Annex F (Simple
Endpoint Type, TD11 in
Monterey) that I feel is causing a functional obstacle
also to the
voice
users. Can anyone clarify if I am correct and why it is
specified the
way
it is.
In section 7.6.1 and 7.6.2 it is specified:" The Audio
SET device
shall
then resume transmitting its media stream(s) to the transport
address(es)
newly indicated in the OpenLogicalChannel structures." I understand that this means that you cannot
re-negotiate audio coding,
and you cannot add text conversation after rerouting the
call from a Voice
only SET to a Text SET.
Re-negotiating the audio coding will probably be a
desired function,
e.g.
when rerouting from a fixed to a wireless IP phone. Adding a data channel for text will also be a desired
function, after
answering a call in an audio-only SET, and then
rerouting it to a
text-capable SET. That action is very common in today's text telephone
usage, and I would
expect it to be just as common in the IP telephony
world. You first
receive the call in the terminal that is closest to you, and
then you get a
reason
to start text mode. Then you transfer the call to
another device with
text
capabilities, where you can switch mode.
Questions:
- Is that kind of call transfer that is handled by the
mechanisms in
7.61
and 7.6.2?
Are my conclusion right about the limitations?
Is this limitation a consequence of using Fast Connect?
Do you see any possibility to avoid the negative
effects of it - to
make re-negotiation possible?
- Is the specified functionality acceptable in the
voice world? If two
devices have agreed on a voice coder, is it likely
that the third
device
supports it? Will this not create a lot of unsuccessful
call transfers
where the users will have a no chance to understand why
they fail?
Another question area:
- When selecting the transport protocol for the text
conversation, the
current draft (APC 1504) specifies TCP or UDP. I realize
that there are
situations where TCP must be avoided. One such situation is a
sub-titled
H.332 transmission. Also other multi-casting situations
is better off
with
a UDP based transport protocol. I am therfore now leaning towards using RTP as the
transport for text
conversation. With RTP we can discover dropped frames
and possibly
invent
a mechanism to mark that event in the text stream for
T.140 to display.
If
we have less than 3 % dropped frames, I think the users
would accept it.
6.1 Do you agree that there are situations when TCP
should be avoided,
and
a UDP based protocol preferred?
6.2 Do you agree that RTP is a good alternative, with a
thin protocol
for
error indications to the user?
6.3 Most packets will carry only 1-4 characters . Can
anyone give me an
indication on the expected packet loss rates in
different situations
for
such packets. Or a document giving such figures. Is
max 3% loss
reachable?
Please give your view on these questions.
Best regards
Gunnar Hellström
Gunnar Hellstrom Representing Ericsson in ITU-T
E-mail gunnar.hellstrom@omnitor.se Tel +46 751 100 501 fax +46 8 556 002 06
participants (1)
-
Chris Purvis