Network Working Group                                   A. van Wijk, Ed.
Request for Comments: 5194                                G. Gybels, Ed.
Category: Informational                                        June 2008


             Framework for Real-Time Text over IP Using
                the Session Initiation Protocol (SIP)

Status of This Memo

  This memo provides information for the Internet community.  It does
  not specify an Internet standard of any kind.  Distribution of this
  memo is unlimited.

Abstract

  This document lists the essential requirements for real-time Text-
  over-IP (ToIP) and defines a framework for implementation of all
  required functions based on the Session Initiation Protocol (SIP) and
  the Real-Time Transport Protocol (RTP).  This includes interworking
  between Text-over-IP and existing text telephony on the Public
  Switched Telephone Network (PSTN) and other networks.





























van Wijk & Gybels            Informational                      [Page 1]

RFC 5194              Framework for TOIP using SIP             June 2008


Table of Contents

  1. Introduction ....................................................3
  2. Scope ...........................................................4
  3. Terminology .....................................................4
  4. Definitions .....................................................4
  5. Requirements ....................................................6
     5.1. General Requirements for ToIP ..............................6
     5.2. Detailed Requirements for ToIP .............................8
          5.2.1. Session Setup and Control Requirements ..............9
          5.2.2. Transport Requirements .............................10
          5.2.3. Transcoding Service Requirements ...................10
          5.2.4. Presentation and User Control Requirements .........11
          5.2.5. Interworking Requirements ..........................13
                 5.2.5.1. PSTN Interworking Requirements ............13
                 5.2.5.2. Cellular Interworking Requirements ........14
                 5.2.5.3. Instant Messaging Interworking
                          Requirements ..............................14
  6. Implementation Framework .......................................15
     6.1. General Implementation Framework ..........................15
     6.2. Detailed Implementation Framework .........................15
          6.2.1. Session Control and Setup ..........................15
                 6.2.1.1. Pre-Session Setup .........................15
                 6.2.1.2. Session Negotiations ......................16
          6.2.2. Transport ..........................................17
          6.2.3. Transcoding Services ...............................18
          6.2.4. Presentation and User Control Functions ............18
                 6.2.4.1. Progress and Status Information ...........18
                 6.2.4.2. Alerting ..................................18
                 6.2.4.3. Text Presentation .........................19
                 6.2.4.4. File Storage ..............................19
          6.2.5. Interworking Functions .............................19
                 6.2.5.1. PSTN Interworking .........................20
                 6.2.5.2. Mobile Interworking .......................22
                          6.2.5.2.1. Cellular "No-gain" .............22
                          6.2.5.2.2. Cellular Text Telephone
                                     Modem (CTM) ....................22
                          6.2.5.2.3. Cellular "Baudot mode" .........22
                          6.2.5.2.4. Mobile Data Channel Mode .......23
                          6.2.5.2.5. Mobile ToIP ....................23
                 6.2.5.3. Instant Messaging Interworking ............23
                 6.2.5.4. Multi-Functional Combination Gateways .....24
                 6.2.5.5. Character Set Transcoding .................25
  7. Further Recommendations for Implementers and Service
     Providers ......................................................25
     7.1. Access to Emergency Services ..............................25
     7.2. Home Gateways or Analog Terminal Adapters .................25
     7.3. User Mobility .............................................26



van Wijk & Gybels            Informational                      [Page 2]

RFC 5194              Framework for TOIP using SIP             June 2008


     7.4. Firewalls and NATs ........................................26
     7.5. Quality of Service ........................................26
  8. Security Considerations ........................................26
  9. Contributors ...................................................27
  10. References ....................................................27
     10.1. Normative References .....................................27
     10.2. Informative References ...................................29

1.  Introduction

  For many years, real-time text has been in use as a medium for
  conversational, interactive dialogue between users in a similar way
  to how voice telephony is used.  Such interactive text is different
  from messaging and semi-interactive solutions like Instant Messaging
  in that it offers an equivalent conversational experience to users
  who cannot, or do not wish to, use voice.  It therefore meets a
  different set of requirements from other text-based solutions already
  available on IP networks.

  Traditionally, deaf, hard-of-hearing, and speech-impaired people are
  amongst the most prolific users of real-time, conversational, text
  but, because of its interactivity, it is becoming popular amongst
  mainstream users as well.  Real-time text conversation can be
  combined with other conversational media like video or voice.

  This document describes how existing IETF protocols can be used to
  implement a Text-over-IP solution (ToIP).  Therefore, this document
  describes how to use a set of existing components and protocols and
  provides the requirements and rules for that resulting structure,
  which is why it is called a "framework", fitting commonly accepted
  dictionary definitions of that term.

  This ToIP framework is specifically designed to be compatible with
  Voice-over-IP (VoIP), Video-over-IP, and Multimedia-over-IP (MoIP)
  environments.  This ToIP framework also builds upon, and is
  compatible with, the high-level user requirements of deaf, hard-of-
  hearing and speech-impaired users as described in RFC3351 [22].  It
  also meets real-time text requirements of mainstream users.

  ToIP also offers an IP equivalent of analog text telephony services
  as used by deaf, hard-of-hearing, speech-impaired, and mainstream
  users.

  The Session Initiation Protocol (SIP) [2] is the protocol of choice
  for control of Multimedia communications and Voice-over-IP (VoIP) in
  particular.  It offers all the necessary control and signalling
  required for the ToIP framework.




van Wijk & Gybels            Informational                      [Page 3]

RFC 5194              Framework for TOIP using SIP             June 2008


  The Real-Time Transport Protocol (RTP) [3] is the protocol of choice
  for real-time data transmission, and its use for real-time text
  payloads is described in RFC 4103 [4].

  This document defines a framework for ToIP to be used either by
  itself or as part of integrated, multi-media services, including
  Total Conversation [5].

2.  Scope

  This document defines a framework for the implementation of real-time
  ToIP, either stand-alone or as a part of multimedia services,
  including Total Conversation [5].  It provides the:

  a. requirements for real-time text;

  b. requirements for ToIP interworking;

  c. description of ToIP implementation using SIP and RTP;

  d. description of ToIP interworking with other text services.

3.  Terminology

  The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
  "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
  "OPTIONAL" in this document are to be interpreted as described in RFC
  2119 [6] and indicate requirement levels for compliant
  implementations.

4.  Definitions

  Audio bridging: a function of an audio media bridge server, gateway,
  or relay service that sends to each destination the combination of
  audio from all participants in a conference, excluding the
  participant(s) at that destination.  At the RTP level, this is an
  instance of the mixer function as defined in RFC 3550 [3].

  Cellular: a telecommunication network that has wireless access and
  can support voice and data services over very large geographical
  areas.  Also called Mobile.

  Full duplex: media is sent independently in both directions.

  Half duplex: media can only be sent in one direction at a time, or if
  an attempt to send information in both directions is made, errors may
  be introduced into the presented media.




van Wijk & Gybels            Informational                      [Page 4]

RFC 5194              Framework for TOIP using SIP             June 2008


  Interactive text: another term for real-time text, as defined below.

  Real-time text: a term for real-time transmission of text in a
  character-by-character fashion for use in conversational services,
  often as a text equivalent to voice-based conversational services.
  Conversational text is defined in the ITU-T Framework for multimedia
  services, Recommendation F.700 [21].

  Text gateway: a function that transcodes between different forms of
  text transport methods, e.g., between ToIP in IP networks and Baudot
  or ITU-T V.21 text telephony in the PSTN.

  Textphone: also "text telephone".  A terminal device that allows
  end-to-end real-time text communication using analog transmission.  A
  variety of PSTN textphone protocols exists world-wide.  A textphone
  can often be combined with a voice telephone, or include voice
  communication functions for simultaneous or alternating use of text
  and voice in a call.

  Text bridging: a function of the text media bridge server, gateway
  (including transcoding gateways), or relay service analogous to that
  of audio bridging as defined above, except that text is the medium of
  conversation.

  Text relay service: a third-party or intermediary that enables
  communications between deaf, hard-of-hearing, and speech-impaired
  people and voice telephone users by translating between voice and
  real-time text in a call.

  Text telephony: analog textphone service.

  Total Conversation: a multimedia service offering real-time
  conversation in video, real-time text and voice according to
  interoperable standards.  All media streams flow in real time.  (See
  ITU-T F.703, "Multimedia conversational services" [5].)

  Transcoding service: a service provided by a third-party User Agent
  that transcodes one stream into another.  Transcoding can be done by
  human operators, in an automated manner, or by a combination of both
  methods.  Within this document, the term particularly applies to
  conversion between different types of media.  A text relay service is
  an example of a transcoding service that converts between real-time
  text and audio.

  TTY: originally, an abbreviation for "teletype".  Often used in North
  America as an alternative designation for a text telephone or
  textphone.  Also called TDD, Telecommunication Device for the Deaf.




van Wijk & Gybels            Informational                      [Page 5]

RFC 5194              Framework for TOIP using SIP             June 2008


  Video relay service: a service that enables communications between
  deaf and hard-of-hearing people and hearing persons with voice
  telephones by translating between sign language and spoken language
  in a call.

  Acronyms:

    2G      Second generation cellular (mobile)
    2.5G    Enhanced second generation cellular (mobile)
    3G      Third generation cellular (mobile)
    ATA     Analog Telephone Adaptor
    CDMA    Code Division Multiple Access
    CLI     Calling Line Identification
    CTM     Cellular Text Telephone Modem
    ENUM    E.164 number storage in DNS (see RFC3761)
    GSM     Global System for Mobile Communications
    ISDN    Integrated Services Digital Network
    ITU-T   International Telecommunications
            Union-Telecommunications Standardisation Sector
    NAT     Network Address Translation
    PSTN    Public Switched Telephone Network
    RTP     Real-Time Transport Protocol
    SDP     Session Description Protocol
    SIP     Session Initiation Protocol
    SRTP    Secure Real Time Transport Protocol
    TDD     Telecommunication Device for the Deaf
    TDMA    Time Division Multiple Access
    TTY     Analog textphone (Teletypewriter)
    ToIP    Real-time Text over Internet Protocol
    URI     Uniform Resource Identifier
    UTF-8   UCS/Unicode Transformation Format-8
    VCO/HCO Voice Carry Over/Hearing Carry Over
    VoIP    Voice over Internet Protocol

5.  Requirements

  The framework described in Section 6 defines a real-time text-based
  conversational service that is the text equivalent of voice-based
  telephony.  This section describes the requirements that the
  framework is designed to meet and the functionality it should offer.

5.1.  General Requirements for ToIP

  Any framework for ToIP must be derived from the requirements of RFC
  3351 [22].  A basic requirement is that it must provide a
  standardized way for offering real-time text-based conversational
  services that can be used as an equivalent to voice telephony by
  deaf, hard-of-hearing, speech-impaired, and mainstream users.



van Wijk & Gybels            Informational                      [Page 6]

RFC 5194              Framework for TOIP using SIP             June 2008


  It is important to understand that real-time text conversations are
  significantly different from other text-based communications like
  email or Instant Messaging.  Real-time text conversations deliver an
  equivalent mode to voice conversations by providing transmission of
  text character by character as it is entered, so that the
  conversation can be followed closely and that immediate interaction
  takes place.

  Store-and-forward systems like email or messaging on mobile networks,
  or non-streaming systems like instant messaging, are unable to
  provide that functionality.  In particular, they do not allow for
  smooth communication through a Text Relay Service.

  In order to make ToIP the text equivalent of voice services, ToIP
  needs to offer equivalent features in terms of conversationality to
  those provided by voice.  To achieve that, ToIP needs to:

  a. offer real-time transport and presentation of the conversation;

  b. provide simultaneous transmission in both directions;

  c. support both point-to-point and multipoint communication;

  d. allow other media, like audio and video, to be used in conjunction
     with ToIP;

  e. ensure that the real-time text service is always available.

  Real-time text is a useful subset of Total Conversation as defined in
  ITU-T F.703 [5].  Total Conversation allows participants to use
  multiple modes of communication during the conversation, either at
  the same time or by switching between modes, e.g., between real-time
  text and audio.

  Deaf, hard-of-hearing, and mainstream users may invoke ToIP services
  for many different reasons:

  - because they are in a noisy environment, e.g., in a machine room of
    a factory where listening is difficult;

  - because they are busy with another call and want to participate in
    two calls at the same time;

  - for implementing text and/or speech recording services (e.g., text
    documentation/audio recording) for legal purposes, for clarity, or
    for flexibility;





van Wijk & Gybels            Informational                      [Page 7]

RFC 5194              Framework for TOIP using SIP             June 2008


  - to overcome language barriers through speech translation and/or
    transcoding services;

  - because of hearing loss, deafness, or tinnitus as a result of the
    aging process or for any other reason, creating a need to replace
    or complement voice with real-time text in conversational sessions.

  In many of the above examples, real-time text may accompany speech.
  The text could be displayed side by side, or in a manner similar to
  subtitling in broadcasting environments, or in any other suitable
  manner.  This could occur with users who are hard of hearing and also
  for mixed media calls with both hearing and deaf people participating
  in the call.

  A ToIP user may wish to call another ToIP user, join a conference
  session involving several users, or initiate or join a multimedia
  session, such as a Total Conversation session.

  A common scenario for multipoint real-time text is conference calling
  with many participants.  Implementers could, for example, use
  different colours to render different participants' text, or could
  create separate windows or rendering areas for each participant.

5.2.  Detailed Requirements for ToIP

  The following sections list individual requirements for ToIP.  Each
  requirement has been given a unique identifier (R1, R2, etc.).
  Section 6 (Implementation Framework) describes how to implement ToIP
  based on these requirements by using existing protocols and
  techniques.

  The requirements are organized under the following headings:

  - session setup and session control;

  - transport;

  - use of transcoding services;

  - presentation and user control;

  - interworking.









van Wijk & Gybels            Informational                      [Page 8]

RFC 5194              Framework for TOIP using SIP             June 2008


5.2.1.  Session Setup and Control Requirements

  Conversations could be started using a mode other than real-time
  text.  Simultaneous or alternating voice and real-time text is used
  by a large number of people who can send voice but must receive text
  (due to a hearing impairment), or who can hear but must send text
  (due to a speech impairment).

  R1: It SHOULD be possible to start conversations in any mode (real-
  time text, voice, video) or combination of modes.

  R2: It MUST be possible for the users to switch to real-time text, or
  add real-time text as an additional modality, during the
  conversation.

  R3: Systems supporting ToIP MUST allow users to select any of the
  supported conversation modes at any time, including in mid-
  conversation.

  R4: Systems SHOULD allow the user to specify a preferred mode of
  communication in each direction, with the ability to fall back to
  alternatives that the user has indicated are acceptable.

  R5: If the user requests simultaneous use of real-time text and
  audio, and this is not possible because of constraints in the
  network, the system SHOULD try to establish text-only communication
  if that is what the user has specified as his/her preference.

  R6: If the user has expressed a preference for real-time text,
  establishment of a connection including real-time text MUST have
  priority over other outcomes of the session setup.

  R7: It MUST be possible to use real-time text in conferences both as
  a medium of discussion between individual participants (for example,
  for sidebar discussions in real-time text while listening to the main
  conference audio) and for central support of the conference with
  real-time text interpretation of speech.

  R8: Session setup and negotiation of modalities MUST allow users to
  specify the language of the real-time text to be used.  (It is
  RECOMMENDED that similar functionality be provided for the video part
  of the conversation, i.e., to specify the sign language being used).

  R9: Where certain session services are available for the audio media
  part of a session, these functions MUST also be supported for the
  real-time text media part of the same session.  For example, call
  transfer must act on all media in the session.




van Wijk & Gybels            Informational                      [Page 9]

RFC 5194              Framework for TOIP using SIP             June 2008


5.2.2.  Transport Requirements

  ToIP will often be used to access a relay service [24], allowing
  real-time text users to communicate with voice users.  With relay
  services, as well as in direct user-to-user conversation, it is
  crucial that text characters are sent as soon as possible after they
  are entered.  While buffering may be done to improve efficiency, the
  delays SHOULD be kept minimal.  In particular, buffering of whole
  lines of text will not meet character delay requirements.

  R10: Characters must be transmitted soon after entry of each
  character so that the maximum delay requirement can be met.  An end-
  to-end delay time of one second is regarded as good, while users note
  and appreciate shorter delays, down to 300ms.  A delay of up to two
  seconds is possible to use.

  R11: Real-time text transmission from a terminal SHALL be performed
  character by character as entered, or in small groups of characters,
  so that no character is delayed from entry to transmission by more
  than 300 milliseconds.

  R12: It MUST be possible to transmit characters at a rate sufficient
  to support fast human typing as well as speech-to-text methods of
  generating real-time text.  A rate of 30 characters per second is
  regarded as sufficient.

  R13: A ToIP service MUST be able to deal with international character
  sets.

  R14: Where it is possible, loss or corruption of real-time text
  during transport SHOULD be detected and the user should be informed.

  R15: Transport of real-time text SHOULD be as robust as possible, so
  as to minimize loss of characters.

  R16: It SHOULD be possible to send and receive real-time text
  simultaneously.

5.2.3.  Transcoding Service Requirements

  If the User Agents of different participants indicate that there is
  an incompatibility between their capabilities to support certain
  media types, e.g., one User Agent only offering T.140 over IP, as
  described in RFC 4103 [4], and the other one only supporting audio,
  the user might want to invoke a transcoding service.

  Some users may indicate their preferred modality to be audio while
  others may indicate real-time text.  In this case, transcoding



van Wijk & Gybels            Informational                     [Page 10]

RFC 5194              Framework for TOIP using SIP             June 2008


  services might be needed for text-to-speech (TTS) and speech-to-text
  (STT).  Other examples of possible scenarios for including a relay
  service in the conversation are: text bridging after conversion from
  speech, audio bridging after conversion from real-time text, etc.

  A number of requirements, motivations, and implementation guidelines
  for relay service invocation can be found in RFC 3351 [22].

  R17: It MUST be possible for users to invoke a transcoding service
  where such service is available.

  R18: It MUST be possible for users to indicate their preferred
  modality (e.g., ToIP).

  R19: It MUST be possible to negotiate the requirements for
  transcoding services in real time in the process of setting up a
  call.

  R20: It MUST be possible to negotiate the requirements for
  transcoding services in mid-call, for the immediate addition of those
  services to the call.

  R21: Communication between the end participants SHOULD continue after
  the addition or removal of a text relay service, and the effect of
  the change should be limited in the users' perception to the direct
  effect of having or not having the transcoding service in the
  connection.

  R22: When setting up a session, it MUST be possible for a user to
  specify the type of relay service requested (e.g., speech to text or
  text to speech).  The specification of a type of relay SHOULD include
  a language specifier.

  R23: It SHOULD be possible to route the session to a preferred relay
  service even if the user invokes the session from another region or
  network than that usually used.

  R24: It is RECOMMENDED that ToIP implementations make the invocation
  and use of relay services as easy as possible.

5.2.4.  Presentation and User Control Requirements

  A user should never be in doubt about the status of the session, even
  if the user is unable to make use of the audio or visual indication.
  For example, tactile indications could be used by deaf-blind
  individuals.





van Wijk & Gybels            Informational                     [Page 11]

RFC 5194              Framework for TOIP using SIP             June 2008


  R25: User Agents for ToIP services MUST have alerting methods (e.g.,
  for incoming sessions) that can be used by deaf and hard-of-hearing
  people or provide a range of alternative, but equivalent, alerting
  methods that can be selected by all users, regardless of their
  abilities.

  R26: Where real-time text is used in conjunction with other media,
  exposure of user control functions through the User Interface needs
  to be done in an equivalent manner for all supported media.  For
  example, it must be possible for the user to select between audio,
  visual, or tactile prompts, or all must be supplied.

  R27: If available, identification of the originating party (e.g., in
  the form of a URI or a Calling Line Identification (CLI)) MUST be
  clearly presented to the user in a form suitable for the user BEFORE
  the session invitation is answered.

  R28: When a session invitation involving ToIP originates from a
  Public Switched Telephone Network (PSTN) text telephone (e.g.,
  transcoded via a text gateway), this SHOULD be indicated to the user.
  The ToIP client MAY adjust the presentation of the real-time text to
  the user as a consequence.

  R29: An indication SHOULD be given to the user when real-time text is
  available during the call, even if it is not invoked at call setup
  (e.g., when only voice and/or video is used initially).

  R30: The user MUST be informed of any change in modalities.

  R31: Users MUST be presented with appropriate session progress
  information at all times.

  R32: Systems for ToIP SHOULD support an answering machine function,
  equivalent to answering machines on telephony networks.

  R33: If an answering machine function is supported, it MUST support
  at least 160 characters for the greeting message.  It MUST support
  incoming text message storage of a minimum of 4096 characters,
  although systems MAY support much larger storage.  It is RECOMMENDED
  that systems support storage of at least 20 incoming messages of up
  to 16000 characters per message.

  R34: When the answering machine is activated, user alerting SHOULD
  still take place.  The user SHOULD be allowed to monitor the auto-
  answer progress, and where this is provided, the user SHOULD be
  allowed to intervene during any stage of the answering machine
  procedure and take control of the session.




van Wijk & Gybels            Informational                     [Page 12]

RFC 5194              Framework for TOIP using SIP             June 2008


  R35: It SHOULD be possible to save the text portion of a
  conversation.

  R36: The presentation of the conversation SHOULD be done in such a
  way that users can easily identify which party generated any given
  portion of text.

  R37: ToIP SHOULD handle characters such as new line, erasure, and
  alerting during a session as specified in ITU-T T.140 [8].

5.2.5.  Interworking Requirements

  There is a range of existing real-time text services.  There is also
  a range of network technologies that could support real-time text
  services.

  Real-time/interactive texting facilities exist already in various
  forms and on various networks.  In the PSTN, they are commonly
  referred to as text telephony.

  Text gateways are used for converting between different protocols for
  text conversation.  They can be used between networks or within
  networks where different transport technologies are used.

  R38: ToIP SHOULD provide interoperability with text conversation
  features in other networks, for instance the PSTN.

  R39: When communicating via a gateway to other networks and
  protocols, the ToIP service SHOULD support the functionality for
  alternating or simultaneous use of modalities as offered by the
  interworking network.

  R40: Calling party identification information, such as CLI, MUST be
  passed by gateways and converted to an appropriate form, if required.

  R41: When interworking with other networks and services, the ToIP
  service SHOULD provide buffering mechanisms to deal with delays in
  call setup and with differences in transmission speeds, and/or to
  interwork with half-duplex services.

5.2.5.1.  PSTN Interworking Requirements

  Analog text telephony is used in many countries, mainly by deaf,
  hard-of-hearing and speech-impaired individuals.

  R42: ToIP services MUST provide interworking with PSTN legacy text
  telephony devices.




van Wijk & Gybels            Informational                     [Page 13]

RFC 5194              Framework for TOIP using SIP             June 2008


  R43: When interworking with PSTN legacy text telephony services,
  alternating text and voice function MAY be supported.  (Called "voice
  carry over (VCO) and hearing carry over (HCO)").

5.2.5.2.  Cellular Interworking Requirements

  As mobile communications have been adopted widely, various solutions
  for real-time texting while on the move were developed.  ToIP
  services should provide interworking with such services as well.

  Alternative means of transferring the text telephony data have been
  developed when TTY services over cellular were mandated by the FCC in
  the USA.  They are the a) "No-gain" codec solution, and b) the
  Cellular Text Telephony Modem (CTM) solution [7], both collectively
  called "Baudot mode" solution in the USA.

  The GSM and 3G standards from 3GPP make use of the CTM modem in the
  voice channel for text telephony.  However, implementations also
  exist that use the data channel to provide such functionality.
  Interworking with these solutions should be done using text gateways
  that set up the data channel connection at the GSM side and provide
  ToIP at the other side.

  R44: a ToIP service SHOULD provide interworking with mobile text
  conversation services.

5.2.5.3.  Instant Messaging Interworking Requirements

  Many people use Instant Messaging to communicate via the Internet
  using text.  Instant Messaging usually transfers blocks of text
  rather than streaming as is used by ToIP.  Usually a specific action
  is required by the user to activate transmission, such as pressing
  the ENTER key or a send button.  As such, it is not a replacement for
  ToIP; in particular, it does not meet the needs for real-time
  conversations including those of deaf, hard-of-hearing, and speech-
  impaired users as defined in RFC 3351 [22].  It is less suitable for
  communications through a relay service [24].

  The streaming nature of ToIP provides a more direct conversational
  user experience and, when given the choice, users may prefer ToIP.

  R45: a ToIP service MAY provide interworking with Instant Messaging
  services.








van Wijk & Gybels            Informational                     [Page 14]

RFC 5194              Framework for TOIP using SIP             June 2008


6.  Implementation Framework

  This section describes an implementation framework for ToIP that
  meets the requirements and offers the functionality as set out in
  Section 5.  The framework presented here uses existing standards that
  are already commonly used for voice-based conversational services on
  IP networks.

6.1.  General Implementation Framework

  This framework specifies the use of the Session Initiation Protocol
  (SIP) [2] to set up, control, and tear down the connections between
  ToIP users whilst the media is transported using the Real-Time
  Transport Protocol (RTP) [3] as described in RFC 4103 [4].

  RFC 4504 describes how to implement support for real-time text in SIP
  telephony devices [23].

6.2.  Detailed Implementation Framework

6.2.1.  Session Control and Setup

  ToIP services MUST use the Session Initiation Protocol (SIP) [2] for
  setting up, controlling, and terminating sessions for real-time text
  conversation with one or more participants and possibly including
  other media like video or audio.  The Session Description Protocol
  (SDP) used in SIP to describe the session is used to express the
  attributes of the session and to negotiate a set of compatible media
  types.

  SIP [2] allows participants to negotiate all media, including real-
  time text conversation [4].  ToIP services can provide the ability to
  set up conversation sessions from any location as well as provision
  for privacy and security through the application of standard SIP
  techniques.

6.2.1.1.  Pre-Session Setup

  The requirements of the user to be reached at a consistent address
  and to store preferences for evaluation at session setup are met by
  pre-session setup actions.  That includes storing of registration
  information in the SIP registrar to provide information about how a
  user can be contacted.  This will allow sessions to be set up rapidly
  and with proper routing and addressing.

  The need to use real-time text as a medium of communications can be
  expressed by users during registration time.  Two situations need to
  be considered in the pre-session setup environment:



van Wijk & Gybels            Informational                     [Page 15]

RFC 5194              Framework for TOIP using SIP             June 2008


  a. User Preferences: It MUST be possible for a user to indicate a
     preference for real-time text by registering that preference with
     a SIP server that is part of the ToIP service.

  b. Server Support of User Preferences: SIP servers that support ToIP
     services MUST have the capability to act on calling user
     preferences for real-time text in order to accept or reject the
     session.  The actions taken can be based on the called users
     preferences defined as part of the pre-session setup registration.
     For example, if the user is called by another party, and it is
     determined that a transcoding server is needed, the session should
     be re-directed or otherwise handled accordingly.

  The ability to include a transcoding service MUST NOT require user
  registration in any specific SIP registrar, but MAY require
  authorisation of the SIP registrar to invoke the service.

  A point-to-point session takes place between two parties.  For ToIP,
  one or both of the communicating parties will indicate real-time text
  as a possible or preferred medium for conversation using SIP in the
  session setup.

  The following features MAY be implemented to facilitate the session
  establishment using ToIP:

  a. Caller Preferences: SIP headers (e.g., Contact) [10] can be used
     to show that real-time text is the medium of choice for
     communications.

  b. Called Party Preferences [11]: The called party being passive can
     formulate a clear rule indicating how a session should be handled,
     either using real-time text as a preferred medium or not, and
     whether this session needs to be handled by a designated SIP proxy
     or the SIP User Agent.

  c. SIP Server Support for User Preferences: It is RECOMMENDED that
     SIP servers also handle the incoming sessions in accordance with
     preferences expressed for real-time text.  The SIP server can also
     enforce ToIP policy rules for communications (e.g., use of the
     transcoding server for ToIP).

6.2.1.2.  Session Negotiations

  The Session Description Protocol (SDP) used in SIP [2] provides the
  capabilities to indicate real-time text as a medium in the session
  setup.  RFC 4103 [4] uses the RTP payload types "text/red" and
  "text/t140" for support of ToIP, which can be indicated in the SDP as
  a part of the SIP INVITE, OK, and SIP/200/ACK media negotiations.  In



van Wijk & Gybels            Informational                     [Page 16]

RFC 5194              Framework for TOIP using SIP             June 2008


  addition, SIP's offer/answer model [12] can also be used in
  conjunction with other capabilities, including the use of a
  transcoding server for enhanced session negotiations [28,29,13].

6.2.2.  Transport

  ToIP services MUST support the Real-Time Transport Protocol (RTP) [3]
  according to the specification of RFC 4103 [4] for the transport of
  real-time text between participants.

  RFC 4103 describes the transmission of T.140 [8] real-time text on IP
  networks.

  In order to enable the use of international character sets, the
  transmission format for real-time text conversation SHALL be UTF-8
  [14], in accordance with ITU-T T.140.

  If real-time text is detected to be missing after transmission, there
  SHOULD be a "text loss" indication in the real-time text as specified
  in T.140 Addendum 1 [8].

  The redundancy method of RFC 4103 [4] SHOULD be used to significantly
  increase the reliability of the real-time text transmission.  A
  redundancy level using 2 generations gives very reliable results and
  is therefore strongly RECOMMENDED.

  In order to avoid exceeding the capabilities of the sender, receiver,
  or network (congestion), the transmission rate SHOULD be kept at or
  below 30 characters per second, which is the default maximum rate
  specified in RFC 4103 [4].  Lower rates MAY be negotiated when needed
  through the "cps" parameter as specified in RFC 4103 [4].

  Real-time text capability is announced in SDP by a declaration
  similar to this example:

  m=text 11000 RTP/AVP 100 98
  a=rtpmap:98 t140/1000
  a=rtpmap:100 red/1000
  a=fmtp:100 98/98/98

  By having this single coding and transmission scheme for real-time
  text defined in the SIP session control environment, the opportunity
  for interoperability is optimized.  However, if good reasons exist,
  other transport mechanisms MAY be offered and used for the T.140-
  coded text, provided that proper negotiation is introduced, but the
  RFC 4103 [4] transport MUST be used as both the default and the
  fallback transport.




van Wijk & Gybels            Informational                     [Page 17]

RFC 5194              Framework for TOIP using SIP             June 2008


6.2.3.  Transcoding Services

  Invocation of a transcoding service MAY happen automatically when the
  session is being set up based on any valid indication or negotiation
  of supported or preferred media types.  A transcoding framework
  document using SIP [28] describes invoking relay services, where the
  relay acts as a conference bridge or uses the third-party control
  mechanism.  ToIP implementations SHOULD support this transcoding
  framework.

6.2.4.  Presentation and User Control Functions

6.2.4.1.  Progress and Status Information

  Session progress information SHOULD use simple language so that as
  many users as possible can understand it.  The use of jargon or
  ambiguous terminology SHOULD be avoided.  It is RECOMMENDED that text
  information be used together with icons to symbolise the session
  progress information.

  In summary, it SHOULD be possible to observe indicators about:

  - Incoming session

  - Availability of real-time text, voice, and video channels

  - Session progress

  - Incoming real-time text

  - Any loss in incoming real-time text

  - Typed and transmitted real-time text

6.2.4.2.  Alerting

  For users who cannot use the audible alerter for incoming sessions,
  it is RECOMMENDED to include a tactile, as well as a visual,
  indicator.

  Among the alerting options are alerting by the User Agent's User
  Interface and specific alerting User Agents registered to the same
  registrar as the main User Agent.

  It should be noted that external alerting systems exist and one
  common interface for triggering the alerting action is a contact
  closure between two conductors.




van Wijk & Gybels            Informational                     [Page 18]

RFC 5194              Framework for TOIP using SIP             June 2008


6.2.4.3.  Text Presentation

  Requirement R32 states that, in the display of text conversations,
  users must be able to distinguish easily between different speakers.
  This could be done using color, positioning of the text (i.e.,
  incoming real-time text and outgoing real-time text in different
  display areas), in-band identifiers of the parties, or a combination
  of any of these techniques.

6.2.4.4.  File Storage

  Requirement R31 recommends that ToIP systems allow the user to save
  text conversations.  This SHOULD be done using a standard file
  format.  For example: a UTF-8 text file in XHTML format [15],
  including timestamps, party names (or addresses), and the
  conversation text.

6.2.5.  Interworking Functions

  A number of systems for real-time text conversation already exist as
  well as a number of message-oriented text communication systems.
  Interoperability is of interest between ToIP and some of these
  systems.

  Interoperation of half-duplex and full-duplex protocols, and between
  protocols that have different data rates, may require text buffering.
  Some intelligence will be needed to determine when to change
  direction when operating in half-duplex mode.  Identification may be
  required of half-duplex operation either at the "user" level (i.e.,
  users must inform each other) or at the "protocol" level (where an
  indication must be sent back to the gateway).  However, special care
  needs to be taken to provide the best possible real-time performance.

  Buffering schemes SHOULD be dimensioned to adjust for receiving at 30
  characters per second and transmitting at 6 characters per second for
  up to 4 minutes (i.e., less than 3000 characters).

  When converting between simultaneous voice and text on the IP side,
  and alternating voice and text on the other side of a gateway, a
  conflict can occur if the IP user transmits both audio and text at
  the same time.  In such situations, text transmission SHOULD have
  precedence, so that while text is transmitted, audio is lost.

  Transcoding of text to and from other coding formats may need to take
  place in gateways between ToIP and other forms of text conversation,
  for example, to connect to a PSTN text telephone.





van Wijk & Gybels            Informational                     [Page 19]

RFC 5194              Framework for TOIP using SIP             June 2008


  Session setup through gateways to other networks may require the use
  of specially formatted addresses or other mechanisms for invoking
  those gateways.

  ToIP interworking requires a method to invoke a text gateway.  These
  text gateways act as User Agents at the IP side.  The capabilities of
  the gateway during the call will be determined by the call
  capabilities of the terminal that is using the gateway.  For example,
  a PSTN textphone is generally only able to receive voice and real-
  time text, so the gateway will only allow ToIP and audio.

  Examples of possible scenarios for invocation of the text gateway
  are:

  a. PSTN textphone users dial a prefix number before dialing out.

  b. Separate real-time text subscriptions, linked to the phone number
     or terminal identifier/ IP address.

  c. Real-time text capability indicators.

  d. Real-time text preference indicators.

  e. Listen for V.18 modem modulation text activity in all PSTN calls
     and routing of the call to an appropriate gateway.

  f. Call transfer request by the called user.

  g. Placing a call via the Web, and using one of the methods described
     here

  h. A text gateway with its own telephone number and/or SIP address
     (this requires user interaction with the gateway to place a call).

  i. ENUM address analysis and number plan.

  j. Number or address analysis leads to a gateway for all PSTN calls.

6.2.5.1.  PSTN Interworking

  Analog text telephony is cumbersome because of incompatible national
  implementations where interworking was never considered.  A large
  number of these implementations have been documented in ITU-T V.18
  [16], which also defines the modem detection sequences for the
  different text protocols.  In rare cases, the modem type
  identification may take considerable time, depending on user actions.





van Wijk & Gybels            Informational                     [Page 20]

RFC 5194              Framework for TOIP using SIP             June 2008


  To resolve analog textphone incompatibilities, text telephone
  gateways are needed to transcode incoming analog signals into T.140
  and vice versa.  The modem capability exchange time can be reduced by
  the text telephone gateways initially assuming the analog text
  telephone protocol used in the region where the gateway is located.
  For example, in the USA, Baudot [25] might be tried as the initial
  protocol.  If negotiation for Baudot fails, the full V.18 modem
  capability exchange will take place.  In the UK, ITU-T V.21 [26]
  might be the first choice.

  In particular, transmission of real-time text on PSTN networks takes
  place using a variety of codings and modulations, including ITU-T
  V.21 [26], Baudot [25], dual-tone multi-frequency (DTMF), V.23 [27],
  and others.  Many difficulties have arisen as a result of this
  variety in text telephony protocols and the ITU-T V.18 [16] standard
  was developed to address some of these issues.

  ITU-T V.18 [16] offers a native text telephony method, plus it
  defines interworking with current protocols.  In the interworking
  mode, it will recognise one of the older protocols and fall back to
  that transmission method when required.

  Text gateways MUST use the ITU-T V.18 [16] standard at the PSTN side.
  A text gateway MUST act as a SIP User Agent on the IP side and
  support RFC 4103 real-time text transport.

  While ToIP allows receiving and sending real-time text simultaneously
  and is displayed on a split screen, many analog text telephones
  require users to take turns typing.  This is because many text
  telephones operate strictly half duplex.  Only one can transmit text
  at a time.  The users apply strict turn-taking rules.

  There are several text telephones which communicate in full duplex,
  but merge transmitted text and received text in the same line in the
  same display window.  Here too the users apply strict turn taking
  rules.

  Native V.18 text telephones support full duplex and separate display
  from reception and transmission so that the full duplex capability
  can be used fully.  Such devices could use the ToIP split screen as
  well, but almost all text telephones use a restricted character set
  and many use low text transmission speeds (4 to 7 characters per
  second).

  That is why it is important for the ToIP user to know that he or she
  is connected with an analog text telephone.  The session description
  [9] SHOULD contain an indication that the other endpoint for the call




van Wijk & Gybels            Informational                     [Page 21]

RFC 5194              Framework for TOIP using SIP             June 2008


  is a PSTN textphone (e.g., connected via an ATA or through a text
  gateway).  This means that the textphone user may be used to formal
  turn taking during the call.

6.2.5.2.  Mobile Interworking

  Mobile wireless (or cellular) circuit switched connections provide a
  digital real-time transport service for voice or data.  The access
  technologies include GSM, CDMA, TDMA, iDen, and various 3G
  technologies, as well as WiFi or WiMAX.

  ToIP may be supported over the cellular wireless packet-switched
  service.  It interfaces to the Internet.

  The following sections describe how mobile text telephony is
  supported.

6.2.5.2.1.  Cellular "No-gain"

  The "No-gain" text telephone transporting technology uses specially
  modified Enhanced Full Rate (EFR) [17] and Enhanced Variable Rate
  (EVR) [18] speech vocoders in mobile terminals used to provide a text
  telephony call.  It provides full duplex operation and supports
  alternating between voice and text ("VCO/HCO").  It is dedicated to
  CDMA and TDMA mobile technologies and the US Baudot (i.e., 45 bit/s)
  type of text telephones.

6.2.5.2.2.  Cellular Text Telephone Modem (CTM)

  CTM [7] is a technology-independent modem technology that provides
  the transport of text telephone characters at up to 10 characters/sec
  using modem signals that can be carried by many voice codecs and uses
  a highly redundant encoding technique to overcome the fading and cell
  changing losses.

6.2.5.2.3.  Cellular "Baudot mode"

  This term is often used by cellular terminal suppliers for a cellular
  phone mode that allows TTYs to operate into a cellular phone and to
  communicate with a fixed-line TTY.  Thus it is a common name for the
  "No-Gain" and the CTM solutions when applied to the Baudot-type
  textphones.









van Wijk & Gybels            Informational                     [Page 22]

RFC 5194              Framework for TOIP using SIP             June 2008


6.2.5.2.4.  Mobile Data Channel Mode

  Many mobile terminals allow the use of the circuit-switched data
  channel to transfer data in real time.  Data rates of 9600 bit/s are
  usually supported on the 2G mobile network.  Gateways provide
  interoperability with PSTN textphones.

6.2.5.2.5.  Mobile ToIP

  ToIP could be supported over mobile wireless packet-switched services
  that interface to the Internet.  For 3GPP 3G services, ToIP support
  is described in 3G TS 26.235 [19].

6.2.5.3.  Instant Messaging Interworking

  Text gateways MAY be used to allow interworking between Instant
  Messaging systems and ToIP solutions.  Because Instant Messaging is
  based on blocks of text, rather than on a continuous stream of
  characters like ToIP, gateways MUST transcode between the two
  formats.  Text gateways for interworking between Instant Messaging
  and ToIP MUST apply a procedure for bridging the different
  conversational formats of real-time text versus text messaging.  The
  following advice may improve user experience for both parties in a
  call through a messaging gateway.

  a. Concatenate individual characters originating at the ToIP side
     into blocks of text.

  b. When the length of the concatenated message becomes longer than 50
     characters, the buffered text SHOULD be transmitted to the Instant
     Messaging side as soon as any non-alphanumerical character is
     received from the ToIP side.

  c. When a new line indicator is received from the ToIP side, the
     buffered characters up to that point, including the carriage
     return and/or line-feed characters, SHOULD be transmitted to the
     Instant Messaging side.

  d. When the ToIP side has been idle for at least 5 seconds, all
     buffered text up to that point SHOULD be transmitted to the
     Instant Messaging side.

  e. Text Gateways must be capable of maintaining the real-time
     performance for ToIP while providing the interworking services.

  It is RECOMMENDED that during the session, both users be constantly
  updated on the progress of the text input.  Many Instant Messaging
  protocols signal that a user is typing to the other party in the



van Wijk & Gybels            Informational                     [Page 23]

RFC 5194              Framework for TOIP using SIP             June 2008


  conversation.  Text gateways between such Instant Messaging protocols
  and ToIP MUST provide this signalling to the Instant Messaging side
  when characters start being received, or at the beginning of the
  conversation.

  At the ToIP side, an indicator of writing the Instant Message MUST be
  present where the Instant Messaging protocol provides one.  For
  example, the real-time text user MAY see ". . . waiting for replying
  IM. . . " and when 5 seconds have passed another . (dot) can be
  shown.

  Those solutions will reduce the difficulties between streaming and
  blocked text services.

  Even though the text gateway can connect Instant Messaging and ToIP,
  the best solution is to take advantage of the fact that the user
  interfaces and the user communities for instant messaging and ToIP
  telephony are very similar.  After all, the character input,
  character display, Internet connectivity, and SIP stack can be the
  same for Instant Messaging (SIMPLE) and ToIP.  Thus, the user may
  simply use different applications for ToIP and text messaging in the
  same terminal.

  Devices that implement Instant Messaging SHOULD implement ToIP as
  described in this document so that a more complete text communication
  service can be provided.

6.2.5.4.  Multi-Functional Combination Gateways

  In practice, many interworking gateways will be implemented as
  gateways that combine different functions.  As such, a text gateway
  could be built to have modems to interwork with the PSTN and support
  both Instant Messaging as well as ToIP.  Such interworking functions
  are called combination gateways.

  Combination gateways could provide interworking between all of their
  supported text-based functions.  For example, a text gateway that has
  modems to interwork with the PSTN and that support both Instant
  Messaging and ToIP could support the following interworking
  functions:

  - PSTN text telephony to ToIP

  - PSTN text telephony to Instant Messaging

  - Instant Messaging to ToIP





van Wijk & Gybels            Informational                     [Page 24]

RFC 5194              Framework for TOIP using SIP             June 2008


6.2.5.5.  Character Set Transcoding

  Gateways between the ToIP network and other networks MAY need to
  transcode text streams.  ToIP makes use of the ISO 10646 character
  set.  Most PSTN textphones use a 7-bit character set, or a character
  set that is converted to a 7-bit character set by the V.18 modem.

  When transcoding between character sets and T.140 in gateways,
  special consideration MUST be given to the national variants of the
  7-bit codes, with national characters mapping into different codes in
  the ISO 10646 code space.  The national variant to be used could be
  selectable by the user on a per-call basis, or be configured as a
  national default for the gateway.

  The indicator of missing text in T.140, specified in T.140 amendment
  1, cannot be represented in the 7-bit character codes.  Therefore the
  indicator of missing text SHOULD be transcoded to the ' (apostrophe)
  character in legacy text telephone systems, where this character
  exists.  For legacy systems where the ' character does not exist, the
  .  (full stop) character SHOULD be used instead.

7.  Further Recommendations for Implementers and Service Providers

7.1.  Access to Emergency Services

  It must be possible to place an emergency call using ToIP and it must
  be possible to use a relay service in such a call.  The emergency
  service provided to users utilising the real-time text medium must be
  equivalent to the emergency service provided to users utilising
  speech or other media.

  A text gateway must be able to route real-time text calls to
  emergency service providers when any of the recognised emergency
  numbers that support text communications for the country or region
  are called, e.g., "911" in the USA and "112" in Europe.  Routing
  real-time text calls to emergency services may require the use of a
  transcoding service.

  A text gateway with cellular wireless packet-switched services must
  be able to route real-time text calls to emergency service providers
  when any of the recognized emergency numbers that support real-time
  text communication for the country is called.

7.2.  Home Gateways or Analog Terminal Adapters

  Analog terminal adapters (ATA) using SIP-based IP communication and
  RJ-11 connectors for connecting traditional PSTN devices SHOULD
  enable connection of legacy PSTN text telephones [23].



van Wijk & Gybels            Informational                     [Page 25]

RFC 5194              Framework for TOIP using SIP             June 2008


  These adapters SHOULD contain V.18 modem functionality, voice
  handling functionality, and conversion functions to/from SIP-based
  ToIP with T.140 transported according to RFC 4103 [4], in a similar
  way as it provides interoperability for voice sessions.

  If a session is set up and text/t140 capability is not declared by
  the destination endpoint (by the endpoint terminal or the text
  gateway in the network at the endpoint), a method for invoking a
  transcoding server SHALL be used.  If no such server is available,
  the signals from the textphone MAY be transmitted in the voice
  channel as audio with a high quality of service.

  NOTE: It is preferred that such analog terminal adaptors do use RFC
  4103 [4] on board and thus act as a text gateway.  Sending textphone
  signals over the voice channel is undesirable due to possible
  filtering and compression and packet loss between the endpoints.
  This can result in character loss in the textphone conversation or
  even not allowing the textphones to connect to each other.

7.3.  User Mobility

  ToIP User Agents SHOULD use the same mechanisms as other SIP User
  Agents to resolve mobility issues.  It is RECOMMENDED that users use
  a SIP address, resolved by a SIP registrar, to enable basic user
  mobility.  Further mechanisms are defined for all session types for
  3G IP multimedia systems.

7.4.  Firewalls and NATs

  ToIP uses the same signalling and transport protocols as VoIP.
  Hence, the same firewall and NAT solutions and network functionality
  that apply to VoIP MUST also apply to ToIP.

7.5.  Quality of Service

  Where Quality of Service (QoS) mechanisms are used, the real-time
  text streams should be assigned appropriate QoS characteristics, so
  that the performance requirements can be met and the real-time text
  stream is not degraded unfavourably in comparison to voice
  performance in congested situations.

8.  Security Considerations

  User confidentiality and privacy need to be met as described in SIP
  [2].  For example, nothing should reveal in an obvious way the fact
  that the ToIP user might be a person with a hearing or speech
  impairment.  It is up to the ToIP user to make his or her hearing or
  speech impairment public.  If a transcoding server is being used,



van Wijk & Gybels            Informational                     [Page 26]

RFC 5194              Framework for TOIP using SIP             June 2008


  this SHOULD be as transparent as possible.  However, it might still
  be possible to discern that a user might be hearing or speech
  impaired based on the attributes present in SDP, although the
  intention is that mainstream users might also choose to use ToIP.
  Encryption SHOULD be used on an end-to-end or hop-by-hop basis as
  described in SIP [2] and SRTP [20].

  Authentication MUST be provided for users in addition to message
  integrity and access control.

  Protection against Denial-of-Service (DoS) attacks needs to be
  provided, considering the case that the ToIP users might need
  transcoding servers.

9.  Contributors

  The following people contributed to this document: Willem Dijkstra,
  Barry Dingle, Gunnar Hellstrom, Radhika R. Roy, Henry Sinnreich, and
  Gregg C. Vanderheiden.

  The content and concepts within are a product of the SIPPING Working
  Group.  Tom Taylor (Nortel) acted as independent reviewer and
  contributed significantly to the structure and content of this
  document.

10.  References

10.1.  Normative References

  [1]   Bradner, S., Ed., "Intellectual Property Rights in IETF
        Technology", BCP 79, RFC 3979, March 2005.

  [2]   Rosenberg, J., Schulzrinne, H., Camarillo, G., Johnston, A.,
        Peterson, J., Sparks, R., Handley, M., and E. Schooler, "SIP:
        Session Initiation Protocol", RFC 3261, June 2002.

  [3]   Schulzrinne, H., Casner, S., Frederick, R., and V. Jacobson,
        "RTP: A Transport Protocol for Real-Time Applications", STD 64,
        RFC 3550, July 2003.

  [4]   Hellstrom, G. and P. Jones, "RTP Payload for Text
        Conversation", RFC 4103, June 2005.

  [5]   ITU-T Recommendation F.703,"Multimedia Conversational
        Services", November 2000.

  [6]   Bradner, S., "Key words for use in RFCs to Indicate Requirement
        Levels", BCP 14, RFC 2119, March 1997.



van Wijk & Gybels            Informational                     [Page 27]

RFC 5194              Framework for TOIP using SIP             June 2008


  [7]   3GPP TS 26.226, "Cellular Text Telephone Modem Description"
        (CTM).

  [8]   ITU-T Recommendation T.140, "Protocol for Multimedia
        Application Text Conversation" (February 1998) and Addendum 1
        (February 2000).

  [9]   Handley, M., Jacobson, V., and C. Perkins, "SDP: Session
        Description Protocol", RFC 4566, July 2006.

  [10]  Rosenberg, J., Schulzrinne, H., and P. Kyzivat, "Indicating
        User Agent Capabilities in the Session Initiation Protocol
        (SIP)", RFC 3840, August 2004.

  [11]  Rosenberg, J., Schulzrinne, H., and P. Kyzivat, "Caller
        Preferences for the Session Initiation Protocol (SIP)", RFC
        3841, August 2004.

  [12]  Rosenberg, J. and H. Schulzrinne, "An Offer/Answer Model with
        Session Description Protocol (SDP)", RFC 3264, June 2002.

  [13]  Camarillo, G., Burger, E., Schulzrinne, H., and A. van Wijk,
        "Transcoding Services Invocation in the Session Initiation
        Protocol (SIP) Using Third Party Call Control (3pcc)", RFC
        4117, June 2005.

  [14]  Yergeau, F., "UTF-8, a transformation format of ISO 10646", STD
        63, RFC 3629, November 2003.

  [15]  "XHTML 1.0: The Extensible HyperText Markup Language: A
        Reformulation of HTML 4 in XML 1.0", W3C Recommendation,
        Available at http://www.w3.org/TR/xhtml1.

  [16]  ITU-T Recommendation V.18, "Operational and Interworking
        Requirements for DCEs operating in Text Telephone Mode",
        November 2000.

  [17]  TIA/EIA/IS-823-A, "TTY/TDD Extension to TIA/EIA-136-410
        Enhanced Full Rate Speech Codec (must used in conjunction with
        TIA/EIA/IS-840)"

  [18]  TIA/EIA/IS-127-2, "Enhanced Variable Rate Codec, Speech Service
        Option 3 for Wideband Spread Spectrum Digital Systems, Addendum
        2."

  [19]  "IP Multimedia default codecs", 3GPP TS 26.235





van Wijk & Gybels            Informational                     [Page 28]

RFC 5194              Framework for TOIP using SIP             June 2008


  [20]  Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K.
        Norrman, "The Secure Real-time Transport Protocol (SRTP)", RFC
        3711, March 2004.

  [21]  ITU-T Recommendation F.700, "Framework Recommendation for
        Multimedia Services", November 2000.

10.2.  Informative References

  [22]  Charlton, N., Gasson, M., Gybels, G., Spanner, M., and A. van
        Wijk, "User Requirements for the Session Initiation Protocol
        (SIP) in Support of Deaf, Hard of Hearing and Speech-impaired
        Individuals", RFC 3351, August 2002.

  [23]  Sinnreich, H., Ed., Lass, S., and C. Stredicke, "SIP Telephony
        Device Requirements and Configuration", RFC 4504, May 2006.

  [24]  European Telecommunications Standards Institute (ETSI), "Human
        Factors (HF); Guidelines for Telecommunication Relay Services
        for Text Telephones". TR 101 806, June 2000.

  [25]  TIA/EIA/825 "A Frequency Shift Keyed Modem for Use on the
        Public Switched Telephone Network." (The specification for
        45.45 and 50 bit/s TTY modems.)

  [26]  International Telecommunication Union (ITU), "300 bits per
        second duplex modem standardized for use in the general
        switched telephone network". ITU-T Recommendation V.21,
        November 1988.

  [27]  International Telecommunication Union (ITU), "600/1200-baud
        modem standardized for use in the general switched telephone
        network", ITU-T Recommendation V.23, November 1988.

  [28]  Camarillo, G., "Framework for Transcoding with the Session
        Initiation Protocol", Work in Progress, May 2006.

  [29]  Camarillo, G., "The SIP Conference Bridge Transcoding Model",
        Work in Progress, January 2006.












van Wijk & Gybels            Informational                     [Page 29]

RFC 5194              Framework for TOIP using SIP             June 2008


Authors' Addresses

  Guido Gybels
  Department of New Technologies
  RNID, 19-23 Featherstone Street
  London EC1Y 8SL, UK

  Tel +44-20-7294 3713
  Txt +44-20-7296 8001 Ext 3713
  Fax +44-20-7296 8069
  EMail: [email protected]
  http://www.ictrnid.org.uk


  Arnoud A. T. van Wijk
  Real-Time Text Taskforce (R3TF)

  EMail: [email protected]
  http://www.realtimetext.org
































van Wijk & Gybels            Informational                     [Page 30]

RFC 5194              Framework for TOIP using SIP             June 2008


Full Copyright Statement

  Copyright (C) The IETF Trust (2008).

  This document is subject to the rights, licenses and restrictions
  contained in BCP 78, and except as set forth therein, the authors
  retain all their rights.

  This document and the information contained herein are provided on an
  "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
  OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND
  THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS
  OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF
  THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
  WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Intellectual Property

  The IETF takes no position regarding the validity or scope of any
  Intellectual Property Rights or other rights that might be claimed to
  pertain to the implementation or use of the technology described in
  this document or the extent to which any license under such rights
  might or might not be available; nor does it represent that it has
  made any independent effort to identify any such rights.  Information
  on the procedures with respect to rights in RFC documents can be
  found in BCP 78 and BCP 79.

  Copies of IPR disclosures made to the IETF Secretariat and any
  assurances of licenses to be made available, or the result of an
  attempt made to obtain a general license or permission for the use of
  such proprietary rights by implementers or users of this
  specification can be obtained from the IETF on-line IPR repository at
  http://www.ietf.org/ipr.

  The IETF invites any interested party to bring to its attention any
  copyrights, patents or patent applications, or other proprietary
  rights that may cover technology that may be required to implement
  this standard.  Please address the information to the IETF at
  [email protected].












van Wijk & Gybels            Informational                     [Page 31]