Santiago meeting documents

Sakae OKUBO okubo at GITI.OR.JP
Fri Jun 4 22:38:44 EDT 1999

It seems to be useful to structure Text Telephone support in gateways in
three levels:

1. Simple text-only support.
There is a simple way to use a textphone. That is to connect with the
intention from the beginning to get a pure text call established and stay
in the text-only mode during the whole call.
Many users use this mode only, and it is also used in many cases in
connection to relay services and emergency services. (relay services are
services where the users get help to translate between different modes in
the conversation, most commonly between voice telephony and text telephony.
Relay services are usually manned services.)

2. Alternating text and voice support.
Text telephone suppport can be quite advanced, with switching back and
forth between text and voice during the call. This mechanism is used with
current textphones that were created long before any simultaneous
voice-data technologies were invented.
Ideally, this support would also be wanted on any gateway termination,
monitoring the line during voice operation for textphone tones and at that
moment initiate text mode.

3. Simultaneous text and voice call.
Alternative 2 is quite cumbersome for the users. A simple way to use
simultaneous voice and text is eagerly needed. It is likely that the H.323
Annex G Text Conversation and Text SET will make this happen as the normal
thing in the IP environment.
But will it also eventually evolve in the PSTN? Within V.18 there is an
opportunity to use V.61 for text telephony with voice and text. H.324
(especially without video) is also a possible environment. Market forces
may tell if this enhancement is wanted in the PSTN.

The Total Conversation terminals with simultaneous Video, Text and Voice
should be placed in the simultaneous voice-text level when discussing text
telephone interworking. Since Total Conversation terminals already exist in
the H.320 invironment in ISDN, we have them in SCN as well as in IP networks.


Termination arrangement for text telephone support.
Level 1.
Level 1 can be achieved with a simple termination. On the SCN side it would
contain V.18 modem support and terminate a T.140 text media stream. ( see
eartlier message in the lists)

Level 2.
If level 2 - alternating voice - text is going to be supported, it is
advisable to have a model for how it should be viewed in terms of
terminations, contexts and signaling.
Multiplexor model:
One model would be to regard the termination to contain a kind of
multiplexor. It would then terminate two media streams - one T.140 text
stream and one audio stream. The mechanisms to connect either stream to the
line would be an internal mechanism in the termination.
The need for external signaling in this case is low.

Another model would be to regard the termination to terminate only one type
of stream at a time. It would need to report events, e.g. about the carrier
dropping, so that the voice mode can be started by an MGC command. A whole
set of detailed interactions seems to be needed, between the MGC and the MG
for the purpose of controlling the termination to follow the user's
expectations for mode changes.

I prefer the multiplexor model for the alternating voice - text mode. Some
thoughts should be given to it, to verify if decent support of user
requirements can be achieved.

Level 3.
Level 3, simultaneous voice and text, is seen on the IP side as one T.140
TCP termination and one RTP audio termination.
On the SCN side it is seen as a termination according to the kind of
multiplexor it uses.

I see no problem in mapping this structure within the H.gcp . I just want
the mechanisms to be agreed on and the details around specific termination
parameters and MG behaviour.

It may be useful to take part of the discussion above into the megaco
requirement spec in order to clarify what is meant with text telephone
support and text conversation support.


Gunnar Hellström

At 07:51 1999-06-04 -0600, Don Stanwyck wrote:
>Gunnar Hellstrom wrote:
>> Text Telephone support is very similar to FAX. If we let FAX into gateway
>> specifications, we shall also include text telephony and text conversation
>> in the same move. It saves a lot of efforts.
>Actually, text telephone (e.g., TDD) is quite a bit different from FAX, and
>in the US, at least, mandated to be supported in every network (ADA
>requirements).  TDD conversations can be one-way modem one-way voice,
>two-way modem, part-time modem (what TDD users call voice-over mode), etc.
>My daughter is a TDD user, and we have become quite familiar with it.  The
>difficulties it presents to IP-telephony are considerable, but can be
>solved by allowing that any line, any time, can be instantly converted to a
>modem communication line without any prior notice or any promise that modem
>tones will continue on the line for any particular duration.
>Attachment Converted: "c:\_Omnitor\Attach\stanwyck.vcf"
Gunnar Hellstrom
LM Ericsson

E-mail gunnar.hellstrom at
Tel +46 751 100 501
fax +46 8 556 002 06

More information about the sg16-avd mailing list