| Internet-Draft | RTP-mixer format for multi-party RTT | November 2020 |
| Hellstrom | Expires 26 May 2021 | [Page] |
Real-time text mixers for multi-party sessions need to identify the source of each transmitted group of text so that the text can be presented by endpoints in suitable grouping with other text from the same source, while new text from other sources is also presented in readable grouping as received interleaved in real-time.¶
Regional regulatory requirements specify provision of real-time text in multi-party calls. RFC 4103 mixer implementations can use traditional RTP functions for source identification, but the mixer source switching performance is limited when using the default transmission characteristics with redundancy.¶
Enhancements for RFC 4103 real-time text mixing is provided in this document, suitable for a centralized conference model that enables source identification and source switching. The intended use is for real-time text mixers and multi-party-aware participant endpoints. The specified mechanism build on the standard use of the CSRC list in the RTP packet for source identification. The method makes use of the same "text/t140" and "text/red" formats as for two-party sessions.¶
A capability exchange is specified so that it can be verified that a participant can handle the multi-party coded real-time text stream. The capability is indicated by use of a media attribute "rtt-mixer".¶
The document updates RFC 4103[RFC4103]¶
A specifications of how a mixer can format text for the case when the endpoint is not multi-party aware is also provided.¶
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.¶
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.¶
Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."¶
This Internet-Draft will expire on 26 May 2021.¶
Copyright (c) 2020 IETF Trust and the persons identified as the document authors. All rights reserved.¶
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.¶
RFC 4103[RFC4103] specifies use of RFC 3550 RTP [RFC3550] for transmission of real-time text (RTT) and the "text/t140" format. It also specifies a redundancy format "text/red" for increased robustness. RFC 4102 [RFC4102] registers the "text/red" format. Regional regulatory requirements specify provision of real-time text in multi-party calls.¶
Real-time text is usually provided together with audio and sometimes with video in conversational sessions.¶
A requirement related to multi-party sessions from the presentation level standard T.140 for real-time text is: "The display of text from the members of the conversation should be arranged so that the text from each participant is clearly readable, and its source and the relative timing of entered text is visualized in the display."¶
Another requirement is that the mixing procedure must not introduce delays in the text streams that are experienced disturbing the real-time experience of the receiving users.¶
The redundancy scheme of RFC 4103 [RFC4103] enables efficient transmission of redundant text in packets together with new text. However the redundancy header format has no source indicators for the redundant transmissions. The redundant parts in a packet must therefore be from the same source as the new text. The recommended transmission is one new and two redundant generations of text (T140blocks) in each packet and the recommended transmission interval for two-party use is 300 ms.¶
Real-time text mixers for multi-party sessions therefore need to insert the source of each transmitted group of text from a conference participant so that the text can be transmitted interleaved with text groups from different sources in the rate they are created. This enables the text groups to be presented by endpoints in suitable grouping with other text from the same source. The presentation can then be arranged so that text from different sources can be presented in real-time and easily read while it is possible for a reading user to also perceive approximately when the text was created in real time by the different parties. The transmission and mixing is intended to be done in a general way so that presentation can be arranged in a layout decided by the endpoint.¶
There are existing implementations of RFC 4103 without the updates from this document. These will not be able to receive and present real-time text mixed for multi-party aware endpoints.¶
A negotiation mechanism is therefore needed for verification if the parties are able to handle a multi-party coded stream and agreeing on using that method.¶
A fall-back mixing procedure is also needed for cases when the negotiation result indicates that a receiving endpoint is not capable of handling the mixed format. This method is called the mixing procedure for multi-party unaware endpoints. The fall-back method is naturally not expected to meet all performance requirements placed on the mixing procedure for multi-party aware endpoints.¶
The document updates RFC 4103[RFC4103] by introducing an attribute for indicating capability for the multi-party mixing case and rules for source indications and source switching.¶
A number of alternatives were considered when searching an efficient and easily implemented multi-party method for real-time text. This section explains a few of them briefly.¶
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in [RFC2119].¶
The terms SDES, CNAME, NAME, SSRC, CSRC, CSRC list, CC, RTCP, RTP-mixer, RTP-translator are explained in [RFC3550]¶
The term "T140block" is defined in RFC 4103 [RFC4103] to contain one or more T.140 code elements.¶
"TTY" stands for a text telephone type used in North America.¶
"WebRTC" stands for web based communication specified by W3C and IETF.¶
"DTLS-SRTP" stands for security specified in RFC 5764 [RFC5764].¶
"multi-party aware" stands for an endpoint receiving real-time text from multiple sources through a common conference mixer being able to present the text in real-time separated by source and presented so that a user can get an impression of the approximate relative timing of text from different parties.¶
"multi-party unaware" stands for an endpoint not itself being able to separate text from different sources when received through a common conference mixer.¶
The method for multi-party real-time text specified in this document is primarily intended for use in transmission between mixers and endpoints in centralised mixing configurations. It is also applicable between mixers. An often mentioned application is for emergency service calls with real-time text and voice, where a calltaker want to make an attended handover of a call to another agent, and stay observing the session. Multimedia conference sessions with support for participants to contribute in text is another application. Conferences with central support for speech-to-text conversion is yet another mentioned application.¶
In all these applications, normally only one participant at a time will send long text utterances. In some cases, one other participant will occasionally contribute with a longer comment simultaneously. That may also happen in some rare cases when text is interpreted to text in another language in a conference. Apart from these cases, other participants are only expected to contribute with very brief utterings while others are sending text.¶
Text is supposed to be human generated, by some text input means, such as typing on a keyboard or using speech-to-text technology. Occasional small cut-and-paste operations may appear even if that is not the initial purpose of real-time text.¶
The real-time characteristics of real-time text is essential for the participants to be able to contribute to a conversation. If the text is too much delayed from typing a letter to its presentation, then, in some conference situations, the opportunity to comment will be gone and someone else will grab the turn. A delay of more than one second in such situations is an obstacle for good conversation.¶
This section contains a brief introduction of the two methods specified in this document.¶
The main purpose of this document is to specify a method for true multi-party real-time text mixing for multi-party aware endpoints. The method makes use of the current format for real-time text in [RFC4103]. It is an update of RFC 4103 by a clarification on one way to use it in the multi-party situation. It is done by completing a negotiation for this kind of multi-party capability and by indicating source in the CSRC element in the RTP packets. Specific considerations are made to be able to recover text after packet loss.¶
The detailed procedures for the multi-party aware case are specified in Section 3¶
Please use [RFC4103] as reference when reading the specification.¶
A method is also specified in this document for cases when the endpoint participating in a multi-party call does not itself implement any solution for multi-party handling of real-time text. The solution requires the mixer to insert text dividers and readable labels and only send text from one source at a time until a suitable point appears for source change. This solution is a fallback method with functional limitations. It acts on the presentation level.¶
A party performing as a mixer, which has not negotiated the "rtt-mixer" sdp media attribute, but negotiated a "text/red" or "text/t140" format in a session with a participant SHOULD, if nothing else is specified for the application, format transmitted text to that participant to be suitable to present on a multi-party unaware endpoint as further specified in Section 4.2.¶
RFC 4103[RFC4103] specifies use of RFC 3550 RTP[RFC3550], and a redundancy format "text/red" for increased robustness of real-time text transmission. This document updates RFC 4103[RFC4103] by introducing a capability negotiation for handling multi-party real-time text, a way to indicate the source of transmitted text, and rules for efficient timing of the transmissions interleaved from different sources.¶
The capability negotiation is based on use of the sdp media attribute "rtt-mixer".¶
Both parties shall indicate their capability in a session setup or modification, and evaluate the capability of the counterpart.¶
A transmitting party SHALL send text according to the multi-party format only when the negotiation for this method was successful and when the CC field in the RTP packet is set to 1. In all other cases, the packets SHALL be populated and interpreted as for a two-party session.¶
A party which has negotiated the "rtt-mixer" sdp media attribute MUST populate the CSRC-list and format the packets according to Section 3 if it acts as an rtp-mixer and sends multi-party text.¶
A party which has negotiated the "rtt-mixer" sdp media attribute MUST interpret the contents of the "CC" field the CSRC-list and the packets according to Section 3 in received rtp packets in the corresponding RTP stream.¶
A party not performing as a mixer MUST not include the CSRC list.¶
The CC field SHALL show the number of members in the CSRC list, which SHALL be one (1) in transmissions from a mixer involved in a multi-party session, and otherwise 0.¶
When transmitted from a mixer during a multi-party session, a CSRC list SHALL be included in the packet. The single member in the CSRC-list SHALL contain the SSRC of the source of the T140blocks in the packet. When redundancy is used, the recommended level of redundancy is to use one primary and two redundant generations of T140blocks. In some cases, a primary or redundant T140block is empty, but is still represented by a member in the redundancy header.¶
From other aspects, the contents of the RTP packets are equal to what is specified in [RFC4103].¶
As soon as a participant is known to participate in a session with another entity and being available for text reception, a Unicode BOM character SHALL be sent to it by the other entity according to the procedures in this section. If the transmitter is a mixer, then the source of this character SHALL be indicated to be the mixer itself.¶
Note that the BOM character SHALL be transmitted with the same redundancy procedures as any other text.¶
After that, the transmitter SHALL send keep-alive traffic to the receiver(s) at regular intervals when no other traffic has occurred during that interval, if that is decided for the actual connection. Recommendations for keep-alive can be found in [RFC6263].¶
A "text/red" or "text/t140" transmitter in a mixer SHOULD send packets distributed in time as long as there is something (new or redundant T140blocks) to transmit. The maximum transmission interval SHOULD then be 330 ms. It is RECOMMENDED to send next packet to a receiver as soon as new text to that receiver is available, as long as the time after the latest sent packet to the same receiver is more than or equal to 100 ms, and also the maximum character rate ("CPS") to the receiver is not exceeded. The intention is to keep the latency low and network load limited while keeping a good protection against text loss in bursty packet loss conditions.¶
For a transmitter not acting in a mixer, the same transmission interval principles apply, but the maximum transmission interval SHOULD be 300 ms.¶
New and redundant text from one source SHALL be transmitted in the same packet if available for transmission at the same time. Text from different sources MUST NOT be transmitted in the same packet.¶
Text received to a mixer from a participant SHOULD NOT be included in transmission from the mixer to that participant.¶
A mixer SHALL handle reception, recovery from packet loss, deletion of superfluous redundancy, marking of possible text loss and deletion of 'BOM' characters from each participant before queueing received text for transmission to receiving participants.¶
A transmitting party using redundancy SHALL send redundant repetitions of T140blocks already transmitted in earlier packets.¶
The number of redundant generations of T140blocks to include in transmitted packets SHALL be deduced from the SDP negotiation. It SHOULD be set to the minimum of the number declared by the two parties negotiating a connection. It is RECOMMENDED to declare and transmit one original and two redundant generations of the T140blocks.¶
When text from more than one source is available for transmission from a mixer, the mixer SHALL let the sources take turns in having their text transmitted.¶
The source with the oldest text received in the mixer or oldest redundant text SHOULD be next in turn to get all its available unsent text transmitted. The age of redundant text SHOULD then be considered to be 330 ms after its previous transmission.¶
The mixer SHOULD compose and transmit an RTP packet to a receiver when one of the following conditions has occurred:¶
At time of transmission, the mixer SHALL populate the RTP packet with all T140blocks queued for transmission originating from the source in turn for transmission as long as this is not in conflict with the allowed number of characters per second ("CPS") or the maximum packet size. In this way, the latency of the latest received text is kept low even in moments of simultaneous transmission from many sources.¶
Redundant text SHALL also be included. See Section 3.14¶
The SSRC of the source shall be placed as the only member in the CSRC-list.¶
Note: The CSRC-list in an RTP packet only includes the participant who's text is included in text blocks. It is not the same as the total list of participants in a conference. With audio and video media, the CSRC-list would often contain all participants who are not muted whereas text participants that don't type are completely silent and thus are not represented in RTP packet CSRC-lists.¶
If no unsent T140blocks were available for a source at the time of populating a packet, but T140blocks are available which have not yet been sent the full intended number of redundant transmissions, then the primary T140block for that source is composed of an empty T140block, and populated (without taking up any length) in a packet for transmission. The corresponding SSRC SHALL be placed as usual in its place in the CSRC-list.¶
The first packet in the session, the first after a source switch and the first after a pause SHALL be poulated with the available T140blocks for the source in turn to be sent as primary, and empty T140blocks for the agreed number of redundancy generations.¶
The primary T140block from a source in the latest transmitted packet is saved for populating the first redundant T140block for that source in next transmission of text from that source. The first redundant T140block for that source from the latest transmission is saved for populating the second redundant T140block in next transmission of text from that source.¶
Usually this is the level of redundancy used. If a higher number of redundancy is negotiated, then the procedure SHALL be maintained until all available redundant levels of T140blocks are placed in the packet. If a receiver has negotiated a lower number of "text/red" generations, then that level shall be the maximum used by the transmitter.¶
The T140blocks saved for transmission as redundant data are assigned a planned transmission time 330 ms after the current time, but SHOULD be transmitted earlier if new text for the same source gets in turn for transmission before that time.¶
The timestamp offset values are inserted in the redundancy header, with the time offset from the RTP timestamp in the packet when the corresponding T140block was sent as primary.¶
The timestamp offsets are expressed in the same clock tick units as the RTP timestamp.¶
The timestamp offset values for empty T140blocks have no relevance but SHOULD be assigned realistic values.¶
The number of members in the CSRC list ( 0 or 1) shall be placed in the "CC" header field. Only mixers place value 1 in the "CC" field. A value of "0" indicates that the source is the transmitting device itself and that the source is indicated by the SSRC field. This value is used by endpoints, and by mixers sending data that it is source of itself.¶
The current time SHALL be inserted in the timestamp.¶
The SSRC of the mixer for the RTT session SHALL be inserted in the SSRC field of the RTP header.¶
When there is no new T140block to transmit, and no redundant T140block that has not been retransmitted the intended number of times from any source, the transmission process can stop until either new T140blocks arrive, or a keep-alive method calls for transmission of keep-alive packets.¶
A mixer SHALL send RTCP reports with SDES, CNAME and NAME information about the sources in the multi-party call. This makes it possible for participants to compose a suitable label for text from each source.¶
Integrity considerations SHALL be considered when composing these fields.¶
The "text/red" receiver included in an endpoint with presentation functions will receive RTP packets in the single stream from the mixer, and SHALL distribute the T140blocks for presentation in presentation areas for each source. Other receiver roles, such as gateways or chained mixers are also feasible, and requires consideration if the stream shall just be forwarded, or distributed based on the different sources.¶
If the "CC" field value of a received packet is 1, it indicates that multi-party transmission is active, and the receiver MUST be prepared to act on the source according to its role. If the CC value is 0, the connection is point-to-point.¶
The used level of redundancy generations SHALL be evaluated from the received packet contents. The number of generations (including the primary) is equal to the number of members in the redundancy header.¶
Empty T140blocks are included as fillers for unused primary or redundancy levels in the packets. They just do not provide any contents.¶
The RTP sequence numbers of the received packets SHALL be monitored for gaps and packets out of order. If a sequence number gap appears and still exists after some defined short time for jitter resolution, the packets in the gap SHALL be regarded lost.¶
If it is known that only one source is active in the RTP session, then it is likely that a gap equal to or larger than the agreed number of redundancy generations (including the primary) causes text loss. In that case a t140block SHALL be created with a marker for possible text loss [T140ad1] and assigned to the source and inserted in the reception buffer for that source.¶
If it is known that more than one source is active in the RTP session, then it is not possible in general to evaluate if text was lost when packets were lost. With two active sources and the recommended number of redundancy generations (3), it can take a gap of five consecutive lost packets until any text may be lost, but text loss can also appear if three non-consecutive packets are lost when they contained consecutive data from the same source. A simple method to decide when there is risk for resulting text loss is to evaluate if three or more packets were lost within one second. Then a t140block SHALL be created with a marker for possible text loss [T140ad1] and assigned to the SSRC of the transmitter as a general input from the mixer.¶
Implementations MAY apply more refined methods for more reliable detection of if text was lost or not. Any refined method SHOULD rather falsely mark possible loss when there was no loss instead of not marking possible loss when there was loss.¶
When applying the following procedures, the effects MUST be considered of possible timestamp wrap around and the RTP session possibly changing SSRC.¶
When a packet is received in an RTP session using the packetization for multi-party aware endpoints, its T140blocks SHALL be extracted in the following way. The description is adapted to the default redundancy case using the original and two redundant generations.¶
The source SHALL be extracted from the CSRC-list if available, otherwise from the SSRC.¶
If the received packet is the first packet received from the source, then all T140blocks in the packet SHALL be retrieved and assigned to a receive buffer for the source beginning with the second generation redundancy, continuing with the first generation redundancy and finally the primary.¶
Note: The normal case is that in the first packet, only the primary data has contents. The redundant data has contents in the first received packet from a source only after initial packet loss.¶
If the packet is not the first packet from a source, then if the second generation redundant data is available, its timestamp SHALL be created by subtracting its timestamp offset from the RTP timestamp. If the resulting timestamp is later than the latest retrieved data from the same source, then the redundant data SHALL be retrieved and appended to the receive buffer. The process SHALL be continued in the same way for the first generation redundant data. After that, the primary data SHALL be retrieved from the packet and appended to the receive buffer for the source.¶
Unicode character 'BOM' is used as a start indication and sometimes used as a filler or keep alive by transmission implementations. These SHALL be deleted after extraction from received packets.¶
This solution has good performance for up to five participants simultaneously sending text. At higher numbers of participants simultaneously sending text, a jerkiness is visible in the presentation of text. With ten participants simultaneously transmitting text, the jerkiness is about one second. Even so, the transmission of text catches up, so there is limited delay of new text. The solution is therefore suitable for emergency service use, relay service use, and small or well-managed larger multimedia conferences. Only in large unmanaged conferences with a high number of participants there may on very rare occasions appear situations when many participants happen to send text simultaneously, resulting in unpleasantly jerky presentation of text from each sending participant. It should be noted that it is only the number of users sending text within the same moment that causes jerkiness, not the total number of users with RTT capability.¶
Security SHOULD be applied on both session control and media. In applications where legacy endpoints without security may exist, a negotiation SHOULD be performed to decide if security by encryption will be applied. If no other security solution is mandated for the application, then RFC 8643 OSRTP [RFC8643] SHOULD be applied to negotiate SRTP media security with DTLS. Most SDP examples below are for simplicity expressed without the security additions. The principles (but not all details) for applying DTLS-SRTP [RFC5764] security is shown in a couple of the following examples.¶
This sections shows some examples of SDP for session negotiation of the real-time text media in SIP sessions. Audio is usually provided in the same session, and sometimes also video. The examples only show the part of importance for the real-time text media.¶
Offer example for "text/red" format and multi-party support:
m=text 11000 RTP/AVP 100 98
a=rtpmap:98 t140/1000
a=rtpmap:100 red/1000
a=fmtp:100 98/98/98
a=rtt-mixer
¶
Answer example from a multi-party capable device
m=text 14000 RTP/AVP 100 98
a=rtpmap:98 t140/1000
a=rtpmap:100 red/1000
a=fmtp:100 98/98/98
a=rtt-mixer
¶
Offer example for "text/red" format including multi-party
and security:
a=fingerprint: (fingerprint1)
m=text 11000 RTP/AVP 100 98
a=rtpmap:98 t140/1000
a=rtpmap:100 red/1000
a=fmtp:100 98/98/98
a=rtt-mixer
¶
The "fingerprint" is sufficient to offer DTLS-SRTP, with the media line still indicating RTP/AVP.¶
Note: For brevity, the entire value of the SDP fingerprint attribute is not shown in this and the following example.¶
Answer example from a multi-party capable device with security
a=fingerprint: (fingerprint2)
m=text 16000 RTP/AVP 100 98
a=rtpmap:98 t140/1000
a=rtpmap:100 red/1000
a=fmtp:100 98/98/98
a=rtt-mixer
¶
With the "fingerprint" the device acknowledges use of SRTP/DTLS.¶
Answer example from a multi-party unaware device that also
does not support security:
m=text 12000 RTP/AVP 100 98
a=rtpmap:98 t140/1000
a=rtpmap:100 red/1000
a=fmtp:100 98/98/98
¶
This example shows a symbolic flow of packets from a mixer including loss and recovery. The sequence includes interleaved transmission of text from two RTT sources A and B. P indicates primary data. R1 is first redundant generation data and R2 is the second redundant generation data. A1, B1, A2 etc are text chunks (T140blocks) received from the respective sources and sent on to the receiver by the mixer. X indicates dropped packet between the mixer and a receiver. The session is assumed to use original and two redundant generations of RTT.¶
|-----------------------| |Seq no 101, Time=20400 | |CC=1 | |CSRC list A | |R2: A1, Offset=600 | |R1: A2, Offset=300 | |P: A3 | |-----------------------|¶
Assuming that earlier packets ( with text A1 and A2) were received in sequence, text A3 is received from packet 101 and assigned to reception area A. The mixer is now assumed to have received text from source B and will send that text 100 ms after packet 101. Transmission of A2 and A3 as redundancy is planned for 330 ms after packet 101 if no new text from A is ready to be sent before that.¶
|-----------------------| |Seq no 102, Time=20500 | |CC=1 | |CSRC list B | |R2 Empty, Offset=600 | |R1: Empty, Offset=300 | |P: B1 | |-----------------------| Packet 102 is received. B1 is retrieved from this packet. Redundant transmission of B1 is planned 330 ms after packet 102.¶
X------------------------| X Seq no 103, Timer=20730| X CC=1 | X CSRC list A | X R2: A2, Offset=630 | X R1: A3, Offset=330 | X P: Empty | X------------------------| Packet 103 is assumed to be dropped in network problems. It contains redundancy for A. Sending A3 as second level redundancy is planned for 330 ms after packet 104.¶
X------------------------| X Seq no 104, Timer=20820| X CC=1 | X CSRC list B | X R2: Empty, Offset=600 | X R1: B1, Offset=300 | X P: B2 | X------------------------| Packet 104 contains text from B, including new B2 and redundant B1. It is assumed dropped in network problems. The mixer has A3 redundancy to send but no new text appears from A and therefore the redundancy is sent 330 ms after the previous packet with text from A.¶
|------------------------| | Seq no 105, Timer=21060| | CC=1 | | CSRC list A | | R2: A3, Offset=660 | | R1: Empty, Offset=330 | | P: Empty | |------------------------| Packet 105 is received. A gap for lost 103, and 104 is detected. Assume that no other loss was detected the last second. Then it can be concluded that nothing was totally lost. R2 is checked. Its original time was 21040-660=20400. A packet with text from A was received with that timestamp, so nothing needs to be recovered. B1 and B2 still needs to be transmitted as redundancy. This is planned 330 ms after packet 105. But that would be at 21150 which is only 90 ms after the latest packet. It is instead transmitted at time 21160.¶
|-----------------------| |Seq no 106, Timer=21160| |CC=1 | |CSRC list B | | R2: B1, Offset=660 | | R1: B2, Offset=340 | | P: Empty | |-----------------------|¶
Packet 106 is received.¶
The second level redundancy in packet 106 is B1 and has timestamp offset 660 ms. The timestamp of packet 106 minus 660 is 20500 which is the timestamp of packet 101 THAT was received. So B1 does not need to be retrieved. The first level redundancy in packet 106 has offset 340. The timestamp of packet 106 minus 340 is 20820. That is later than the latest received packet with source B. Therefore B2 is retrieved and assigned to the input buffer for source B. No primary is available in packet 106¶
After this sequence, A3 and B1 and B2 have been received. In this case no text was lost.¶
The default maximum rate of reception of "text/t140" real-time text is in RFC 4103 [RFC4103] specified to be 30 characters per second. The value MAY be modified in the CPS parameter of the FMTP attribute in the media section for the "text/t140" media. A mixer combining real-time text from a number of sources may occasionally have a higher combined flow of text coming from the sources. Endpoints SHOULD therefore specify a suitable higher value for the CPS parameter, corresponding to its real reception capability. A value for "CPS" of 90 is the default for the "text/t140" stream in the "text/red" format when multi-party real-time text is negotiated. See RFC 4103 [RFC4103] for the format and use of the CPS parameter. The same rules apply for the multi-party case except for the default value.¶
ITU-T T.140 [T140] provides the presentation level requirements for the RFC 4103 [RFC4103] transport. T.140 [T140] has functions for erasure and other formatting functions and has the following general statement for the presentation:¶
"The display of text from the members of the conversation should be arranged so that the text from each participant is clearly readable, and its source and the relative timing of entered text is visualized in the display. Mechanisms for looking back in the contents from the current session should be provided. The text should be displayed as soon as it is received."¶
Strict application of T.140 [T140] is of essence for the interoperability of real-time text implementations and to fulfill the intention that the session participants have the same information of the text contents of the conversation without necessarily having the exact same layout of the conversation.¶
T.140 [T140] specifies a set of presentation control codes to include in the stream. Some of them are optional. Implementations MUST be able to ignore optional control codes that they do not support.¶
There is no strict "message" concept in real-time text. The Unicode Line Separator character SHALL be used as a separator allowing a part of received text to be grouped in presentation. The characters "CRLF" may be used by other implementations as replacement for Line Separator. The "CRLF" combination SHALL be erased by just one erasing action, just as the Line Separator. Presentation functions are allowed to group text for presentation in smaller groups than the line separators imply and present such groups with source indication together with text groups from other sources (see the following presentation examples). Erasure has no specific limit by any delimiter in the text stream.¶
A multi-party aware receiving party, presenting real-time text MUST separate text from different sources and present them in separate presentation fields. The receiving party MAY separate presentation of parts of text from a source in readable groups based on other criteria than line separator and merge these groups in the presentation area when it benefits the user to most easily find and read text from the different participants. The criteria MAY e.g. be a received comma, full stop, or other phrase delimiters, or a long pause.¶
When text is received from multiple original sources, the presentation SHOULD provide a view where text is added in multiple presentation fields.¶
If the presentation presents text from different sources in one common area, the presenting endpoint SHOULD insert text from the local user ended at suitable points merged with received text to indicate the relative timing for when the text groups were completed. In this presentation mode, the receiving endpoint SHALL present the source of the different groups of text. This presentation style is called the "chat" style here.¶
A view of a three-party RTT call in chat style is shown in this example .¶
_________________________________________________
| |^|
|[Alice] Hi, Alice here. |-|
| | |
|[Bob] Bob as well. | |
| | |
|[Eve] Hi, this is Eve, calling from Paris. | |
| I thought you should be here. | |
| | |
|[Alice] I am coming on Thursday, my | |
| performance is not until Friday morning.| |
| | |
|[Bob] And I on Wednesday evening. | |
| | |
|[Alice] Can we meet on Thursday evening? | |
| | |
|[Eve] Yes, definitely. How about 7pm. | |
| at the entrance of the restaurant | |
| Le Lion Blanc? | |
|[Eve] we can have dinner and then take a walk |-|
|______________________________________________|v|
| <Eve-typing> But I need to be back to |^|
| the hotel by 11 because I need |-|
| | |
| <Bob-typing> I wou |-|
|______________________________________________|v|
| of course, I underst |
|________________________________________________|
¶
Figure 3: Example of a three-party RTT call presented in chat style seen at participant 'Alice's endpoint.¶
Other presentation styles than the chat style may be arranged.¶
This figure shows how a coordinated column view MAY be presented.¶
_____________________________________________________________________ | Bob | Eve | Alice | |____________________|______________________|_______________________| | | |I will arrive by TGV. | |My flight is to Orly| |Convenient to the main | | |Hi all, can we plan |station. | | |for the seminar? | | |Eve, will you do | | | |your presentation on| | | |Friday? |Yes, Friday at 10. | | |Fine, wo | |We need to meet befo | |___________________________________________________________________|¶
Figure 4: An example of a coordinated column-view of a three-party session with entries ordered vertically in approximate time-order.¶
When the mixer has indicated multi-party capability by the "rtt-mixer" sdp attribute in an SDP negotiation, but the multi-party capability negotiation fails with an endpoint, then the agreed "text/red" or "text/t140" format SHALL be used and the mixer SHOULD compose a best-effort presentation of multi-party real-time text in one stream intended to be presented by an endpoint with no multi-party awareness.¶
This presentation format has functional limitations and SHOULD be used only to enable participation in multi-party calls by legacy deployed endpoints implementing only RFC 4103 without any multi-party extensions specified in this document.¶
The principles and procedures below do not specify any new protocol elements. They are instead composed from the information in ITU-T T.140 [T140] and an ambition to provide a best effort presentation on an endpoint which has functions only for two-party calls.¶
The mixer mixing for multi-party unaware endpoints SHALL compose a simulated limited multi-party RTT view suitable for presentation in one presentation area. The mixer SHALL group text in suitable groups and prepare for presentation of them by inserting a new line between them if the transmitted text did not already end with a new line. A presentable label SHOULD be composed and sent for the source initially in the session and after each source switch. With this procedure the time for source switching is depending on the actions of the users. In order to expedite source switch, a user can for example end its turn with a new line.¶
When text is received by the mixer from the different participants, the mixer SHALL recover text from redundancy if any packets are lost. The mark for lost text [T140ad1] SHOULD be inserted in the stream if unrecoverable loss appears. Any Unicode "BOM" characters, possibly used for keep-alive shall be deleted. The time of creation of text (retrieved from the RTP timestamp) SHALL be stored together with the received text from each source in queues for transmission to the recipients.¶
The following procedure SHOULD be applied for each recipient of multi-part text from the mixer.¶
The text for transmission SHOULD be formatted by the mixer for each receiving user for presentation in one single presentation area. Text received from a participant SHOULD NOT be included in transmission to that participant. When there is text available for transmission from the mixer to a receiving party from more than one participant, the mixer SHOULD switch between transmission of text from the different sources at suitable points in the transmitted stream.¶
When switching source, the mixer SHOULD insert a line separator if the already transmitted text did not end with a new line (line separator or CRLF). A label SHOULD be composed from information in the CNAME and NAME fields in RTCP reports from the participant to have its text transmitted, or from other session information for that user. The label SHOULD be delimited by suitable characters (e.g. '[ ]') and transmitted. The CSRC SHOULD indicate the selected source. Then text from that selected participant SHOULD be transmitted until a new suitable point for switching source is reached.¶
Integrity considerations SHALL be taken when composing the label.¶
Seeking a suitable point for switching source SHOULD be done when there is older text waiting for transmission from any party than the age of the last transmitted text. Suitable points for switching are:¶
When switching source, the source which has the oldest text in queue SHOULD be selected to be transmitted. A character display count SHOULD be maintained for the currently transmitted source, starting at zero after the label is transmitted for the currently transmitted source.¶
The status SHOULD be maintained for the latest control code for Select Graphic Rendition (SGR) from each source. If there is an SGR code stored as the status for the current source before the source switch is done, a reset of SGR shall be sent by the sequence SGR 0 [009B 0000 006D] after the new line and before the new label during a source switch. See SGR below for an explanation. This transmission does not influence the display count.¶
If there is an SGR code stored for the new source after the source switch, that SGR code SHOULD be transmitted to the recipient before the label. This transmission does not influence the display count.¶
Text from a source sent to the recipient SHOULD increase the display count by one per transmitted character.¶
The following control codes specified by T.140 require specific actions. They SHOULD cause specific considerations in the mixer. Note that the codes presented here are expressed in UCS-16, while transmission is made in UTF-8 transform of these codes.¶
A mixer transmitting to a multi-party unaware terminal SHOULD send primary data only from one source per packet. The SSRC SHOULD be the SSRC of the mixer. The CSRC list SHOULD contain one member and be the SSRC of the source of the primary data.¶
When a multi-party unaware endpoint presents a conversation in one display area in a chat style, it inserts source indications for remote text and local user text as they are merged in completed text groups. When an endpoint using this layout receives and presents text mixed for multi-party unaware endpoints, there will be two levels of source indicators for the received text; one generated by the mixer and inserted in a label after each source switch, and another generated by the receiving endpoint and inserted after each switch between local and remote source in the presentation area. This will waste display space and look inconsistent to the reader.¶
New text can be presented only from one source at a time. Switch of source to be presented takes place at suitable places in the text, such as end of phrase, end of sentence, line separator and inactivity. Therefore the time to switch to present waiting text from other sources may become long and will vary and depend on the actions of the currently presented source.¶
Erasure can only be done up to the latest source switch. If a user tries to erase more text, the erasing actions will be presented as letter X after the label.¶
Text loss because of network errors may hit the label between entries from different parties, causing risk for misunderstanding from which source a piece of text is.¶
These facts makes it strongly RECOMMENDED to implement multi-party awareness in RTT endpoints. The use of the mixing method for multi-party-unaware endpoints should be left for use with endpoints which are impossible to upgrade to become multi-party aware.¶
The following pictures are examples of the view on a participant's display for the multi-party-unaware case.¶
_________________________________________________ | Conference | Alice | |________________________|_________________________| | |I will arrive by TGV. | |[Bob]:My flight is to |Convenient to the main | |Orly. |station. | |[Eve]:Hi all, can we | | |plan for the seminar. | | | | | |[Bob]:Eve, will you do | | |your presentation on | | |Friday? | | |[Eve]:Yes, Friday at 10.| | |[Bob]: Fine, wo |We need to meet befo | |________________________|_________________________|¶
Figure 5: Alice who has a conference-unaware client is receiving the multi-party real-time text in a single-stream. This figure shows how a coordinated column view MAY be presented on Alice's device.¶
_________________________________________________ | |^| |[Alice] Hi, Alice here. |-| | | | |[mix][Bob] Bob as well. | | | | | |[Eve] Hi, this is Eve, calling from Paris | | | I thought you should be here. | | | | | |[Alice] I am coming on Thursday, my | | | performance is not until Friday morning.| | | | | |[mix][Bob] And I on Wednesday evening. | | | | | |[Eve] we can have dinner and then walk | | | | | |[Eve] But I need to be back to | | | the hotel by 11 because I need | | | |-| |______________________________________________|v| | of course, I underst | |________________________________________________|¶
Figure 6: An example of a view of the multi-party unaware presentation in chat style. Alice is the local user.¶
The SIP conferencing framework, mainly specified in RFC 4353[RFC4353], RFC 4579[RFC4579] and RFC 4575[RFC4575] is suitable for coordinating sessions including multi-party RTT. The RTT stream between the mixer and a participant is one and the same during the conference. Participants get announced by notifications when participants are joining or leaving, and further user information may be provided. The SSRC of the text to expect from joined users MAY be included in a notification. The notifications MAY be used both for security purposes and for translation to a label for presentation to other users.¶
In managed conferences, control of the real-time text media SHOULD be provided in the same way as other for media, e.g. for muting and unmuting by the direction attributes in SDP [RFC4566].¶
Note that floor control functions may be of value for RTT users as well as for users of other media in a conference.¶
Multi-party RTT sessions may involve gateways of different kinds. Gateways involved in setting up sessions SHALL correctly reflect the multi-party capability or unawareness of the combination of the gateway and the remote endpoint beyond the gateway.¶
One case that may occur is a gateway to PSTN for communication with textphones (e.g. TTYs). Textphones are limited devices with no multi-party awareness, and it SHOULD therefore be suitable for the gateway to not indicate multi-party awareness for that case. Another solution is that the gateway indicates multi-party capability towards the mixer, and includes the multi-party mixer function for multi-party unaware endpoints itself. This solution makes it possible to make adaptations for the functional limitations of the textphone (TTY).¶
More information on gateways to textphones (TTYs) is found in RFC 5194[RFC5194]¶
Gateway operation to real-time text in WebRTC may also be required. In WebRTC, RTT is specified in [I-D.ietf-mmusic-t140-usage-data-channel].¶
A multi-party bridge may have functionality for communicating by RTT both in RTP streams with RTT and WebRTC t140 data channels. Other configurations may consist of a multi-party bridge with either technology for RTT transport and a separate gateway for conversion of the text communication streams between RTP and t140 data channel.¶
In WebRTC, it is assumed that for a multi-party session, one t140 data channel is established for each source from a gateway or bridge to each participant. Each participant also has a data channel with two-way connection with the gateway or bridge.¶
The t140 channel used both ways is for text from the WebRTC user and from the bridge or gateway itself to the WebRTC user. The label parameter of this t140 channel is used as NAME field in RTCP to participants on the RTP side. The other t140 channels are only for text from other participants to the WebRTC user.¶
When a new participant has entered the session with RTP transport of rtt, a new t140 channel SHOULD be established to WebRTC users with the label parameter composed from the NAME field in RTCP on the RTP side.¶
When a new participant has entered the multi-party session with RTT transport in a WebRTC t140 data channel, the new participant SHOULD be announced by a notification to RTP users. The label parameter from the WebRTC side SHOULD be used as the NAME RTCP field on the RTP side, or other available session information.¶
This document updates RFC 4103[RFC4103] by introducing an sdp media attribute "rtt-mixer" for negotiation of multi-party mixing capability with the [RFC4103] format, and by specifying the rules for packets when multi-party capability is negotiated and in use.¶
The congestion considerations and recommended actions from RFC 4103 [RFC4103] are valid also in multi-party situations.¶
The first action in case of congestion SHOULD be to temporarily increase the transmission interval up to two seconds.¶
If the unlikely situation appears that more than 20 participants in a conference send text simultaneously, it will take more than 7 seconds between presentation of text from each of these participants. More time than that can cause confusion in the session. It is therefore RECOMMENDED that the mixer discards such text in excess inserts a general indication of possible text loss [T140ad1] in the session. If the main text contributor is indicated in any way, the mixer MAY avoid deleting text from that participant.¶
James Hamlin for format and performance aspects.¶
[RFC EDITOR NOTE: Please replace all instances of RFCXXXX with the RFC number of this document.]¶
IANA is asked to register the new sdp attribute "rtt-mixer".¶
The RTP-mixer model requires the mixer to be allowed to decrypt, pack and encrypt secured text from the conference participants. Therefore the mixer needs to be trusted. This is similar to the situation for central mixers of audio and video.¶
The requirement to transfer information about the user in RTCP reports in SDES, CNAME and NAME fields, and in conference notifications, for creation of labels may have privacy concerns as already stated in RFC 3550 [RFC3550], and may be restricted of privacy reasons. The receiving user will then get a more symbolic label for the source.¶
Participants with malicious intentions may appear and e.g. disturb the multi-party session by a continuous flow of text, or masquerading as text from other participants. Counteractions should be to require secure signaling, media and authentication, and to provide higher level conference functions e.g. for blocking and expelling participants.¶
Timestamps and timestamp offsets added to the packet examples in section 3.23, and the description corrected.¶
A number of minor corrections added in sections 3.10 - 3.23.¶
The packet composition was modified for interleaving packets from different sources.¶
The packet reception was modified for the new interleaving method.¶
The packet sequence examples was adjusted for the new interleaving method.¶
Modifications according to responses to Brian Rosen of 2020-11-03¶
Changed name on the SDP media attribute to "rtt-mixer"¶
Restructure of section 2 for balance between aware and unaware cases.¶
Moved conference control to own section.¶
Improved clarification of recovery and loss in the packet sequence example.¶
A number of editorial corrections and improvements.¶
Deleted the method requiring a new packet format "text/rex" because of the longer standardization and implementation period it needs.¶
Focus on use of RFC 4103 text/red format with shorter transmission interval, and source indicated in CSRC.¶
Added a method based on the "text/red" format and single source per packet, negotiated by the "rtt-mixer" sdp attribute.¶
Added reasoning and recommendation about indication of loss.¶
The highest number of sources in one packet is 15, not 16. Changed.¶
Added in information on update to RFC 4103 that RFC 4103 explicitly allows addition of FEC method. The redundancy is a kind of forward error correction..¶
Improved definitions list format.¶
The format of the media subtype parameters is made to match the requirements.¶
The mapping of media subtype parameters to sdp is included.¶
The CPS parameter belongs to the t140 subtype and does not need to be registered here.¶
nomenclature and editorial improvements¶
"this document" used consistently to refer to this document.¶
'Redundancy header' renamed to 'data header'.¶
More clarifications added.¶
Language and figure number corrections.¶
Mention possible need to mute and raise hands as for other media. ---done ----¶
Make sure that use in two-party calls is also possible and explained. - may need more wording -¶
Clarify the RTT is often used together with other media. --done--¶
Tell that text mixing is N-1. A users own text is not received in the mix. -done-¶
In 3. correct the interval to: A "text/rex" transmitter SHOULD send packets distributed in time as long as there is something (new or redundant T140blocks) to transmit. The maximum transmission interval SHOULD then be 300 ms. It is RECOMMENDED to send a packet to a receiver as soon as new text to that receiver is available, as long as the time after the latest sent packet to the same receiver is more than 150 ms, and also the maximum character rate to the receiver is not exceeded. The intention is to keep the latency low while keeping a good protection against text loss in bursty packet loss conditions. -done-¶
In 1.3 say that the format is used both ways. -done-¶
In 13.1 change presentation area to presentation field so that reader does not think it shall be totally separated. -done-¶
In Performance and intro, tell the performance in number of simultaneous sending users and introduced delay 16, 150 vs requirements 5 vs 500. -done --¶
Clarify redundancy level per connection. -done-¶
Timestamp also for the last data header. To make it possible for all text to have time offset as for transmission from the source. Make that header equal to the others. -done-¶
Mixer always use the CSRC list, even for its own BOM. -done-¶
Combine all talk about transmission interval (300 ms vs when text has arrived) in section 3 in one paragraph or close to each other. -done-¶
Documents the goal of good performance with low delay for 5 simultaneous typers in the introduction. -done-¶
Describe better that only primary text shall be sent on to receivers. Redundancy and loss must be resolved by the mixer. -done-¶
SDP and better description and visibility of security by OSRTP RFC 8634 needed.¶
The description of gatewaying to WebRTC extended.¶
The description of the data header in the packet is improved.¶
2,5,6 More efficient format "text/rex" introduced and attribute a=rtt-mix deleted.¶
3. Brief about use of OSRTP for security included- More needed.¶
4. Brief motivation for the solution and why not rtp-translator is used added to intro.¶
7. More limitations for the multi-party unaware mixing method inserted.¶
8. Updates to RFC 4102 and 4103 more clearly expressed.¶
9. Gateway to WebRTC started. More needed.¶
Changed file name to draft-ietf-avtcore-multi-party-rtt-mix-00¶
Replaced CDATA in IANA registration table with better coding.¶
Converted to xml2rfc version 3.¶
Changed company and e-mail of the author.¶
Changed title to "RTP-mixer formatting of multi-party Real-time text" to better match contents.¶
Check and modification where needed of use of RFC 2119 words SHALL etc.¶
More about the CC value in sections on transmitters and receivers so that 1-to-1 sessions do not use the mixer format.¶
Enhanced section on presentation for multi-party-unaware endpoints¶
A paragraph recommending CPS=150 inserted in the performance section.¶
In Abstract and 1. Introduction: Introduced wording about regulatory requirements.¶
In section 5: The transmission interval is decreased to 100 ms when there is text from more than one source to transmit.¶
In section 11 about SDP negotiation, a SHOULD-requirement is introduced that the mixer should make a mix for multi-party unaware endpoints if the negotiation is not successful. And a reference to a later chapter about it.¶
The presentation considerations chapter 14 is extended with more information about presentation on multi-party aware endpoints, and a new section on the multi-party unaware mixing with low functionality but SHOULD a be implemented in mixers. Presentation examples are added.¶
A short chapter 15 on gateway considerations is introduced.¶
Clarification about the text/t140 format included in chapter 10.¶
This sentence added to the chapter 10 about use without redundancy. "The text/red format SHOULD be used unless some other protection against packet loss is utilized, for example a reliable network or transport."¶
Note about deviation from RFC 2198 added in chapter 4.¶
In chapter 9. "Use with SIP centralized conferencing framework" the following note is inserted: Note: The CSRC-list in an RTP packet only includes participants who's text is included in one or more text blocks. It is not the same as the list of participants in a conference. With audio and video media, the CSRC-list would often contain all participants who are not muted whereas text participants that don't type are completely silent and so don't show up in RTP packet CSRC-lists.¶
Editorial cleanup.¶
Changed capability indication from fmtp-parameter to SDP attribute "rtt-mix".¶
Swapped order of redundancy elements in the example to match reality.¶
Increased the SDP negotiation section¶