Internet Draft draft-ietf-avt-rtp-retransmission- Jose Rey/Matsushita 04.txt David Leon/Nokia Akihiro Miyazaki/Matsushita Viktor Varsa/Nokia Rolf Hakenberg/Matsushita Expires: May 2003 December 2002 RTP retransmission payload format Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC2026. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet- Drafts. 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." The list of current Internet-Drafts can be accessed at http://www.ietf.org/ietf/1id-abstracts.txt The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html. [Note to RFC Editor: This paragraph is to be deleted when this draft is published as an RFC. References in this draft to RFC XXXX should be replaced with the RFC number assigned to this document. References in this draft to RFC YYYY should be replaced with the RFC number assigned the draft-ietf-mmusic-fid when published as RFC. References in this draft to RFC ZZZZ should be replaced with the RFC number assigned the draft-ietf-avt-rtcp-bw when published as RFC. References in this draft to RFC UUUU should be replaced with the RFC number assigned the draft-ietf-avt-srtp when published as RFC. References in this draft to RFC VVVV should be replaced with the RFC number assigned the draft-ietf-avt-rtcp-feedback when published as RFC. References in this draft to RFC WWWW should be replaced with the RFC number of the revision of RFC 1889 being drafted as draft- ietf-avt-rtp-new. Main changes since draft-ietf-avt-rtp- retransmission-02.txt: this document is the result of the merging of draft-ietf-avt-selret-05.txt and draft-ietf-avt-rtp-retransmission- 02.txt. Main changes since draft-ietf-avt-rtp-retransmission-03.txt: RTSP section new drafted.] IETF draft - Expires May 2003 1 Internet Draft RTP Retransmission Payload Format December 2002 Abstract RTP retransmission is an effective packet loss recovery technique for real-time applications with relaxed delay bounds. This document describes an RTP payload format for performing retransmissions. Retransmitted RTP packets are sent in a separate stream from the original RTP stream. It is assumed that feedback from receivers to senders is available. In particular, it is assumed that RTCP feedback as defined in the extended RTP profile for RTCP-based feedback [1] ( denoted AVPF ), is available in this memo. Table of Contents 1. Introduction....................................................3 2. Terminology.....................................................3 3. Requirements and design rationale for a retransmission scheme...4 4. Retransmission payload format...................................6 5. Association of a retransmission stream with its original stream.7 6. Use with the extended RTP profile for RTCP-based feedback.......8 7. Congestion control.............................................10 8. SDP usage......................................................11 9. RTSP considerations............................................14 10. Implementation examples.......................................15 11. IANA considerations...........................................18 12. Security considerations.......................................22 13. Acknowledgements..............................................22 14. References....................................................22 Author's Addresses................................................23 Rey/Leon/Miyazaki/Varsa/Hakenberg 2 Internet Draft RTP Retransmission Payload Format December 2002 1. Introduction Packet losses between an RTP sender and receiver may significantly degrade the quality of the received media. Several techniques, such as forward error correction (FEC), retransmissions or interleaving may be considered to increase packet loss resiliency. RFC 2354 [9] discusses the different options. When choosing a repair technique for a particular application, the tolerable latency of the application has to be taken into account. In the case of multimedia conferencing, the end-to-end delay has to be at most a few hundred milliseconds in order to guarantee interactivity, which usually excludes the use of retransmission. However, in the case of multimedia streaming, the user can tolerate an initial latency as part of the session set-up and thus an end-to- end delay of several seconds may be acceptable. Retransmission may thus be considered for such applications. This document specifies a retransmission method for RTP applicable to unicast and (small) multicast groups: it defines a payload format for retransmitted RTP packets and provides protocol rules for the sender and the receiver involved in retransmissions. Furthermore, this retransmission payload format was designed for use with the extended RTP profile for RTCP-based feedback, AVPF [1]. It may also be used with other RTP profiles defined in the future. The AVPF profile allows for more frequent feedback and for early feedback. It defines a small number of general-purpose feedback messages, e.g. ACKs and NACKs, as well as codec and application- specific feedback messages. See [1] for details. 2. Terminology The following terms are used in this document: Original packet: refers to an RTP packet which carries user data sent for the first time by an RTP sender. Original stream: refers to the RTP stream of original packets. Retransmission packet: refers to an RTP packet whose payload includes the payload of an already sent original packet. Such a retransmission packet is said to be associated with the original RTP packet. Retransmission request: a means by which an RTP receiver is able to request that the RTP sender should send a retransmission packet for a given original packet. Usually, an RTCP NACK message as specified in [1] is used as retransmission request for lost packets. Rey/Leon/Miyazaki/Varsa/Hakenberg 3 Internet Draft RTP Retransmission Payload Format December 2002 Retransmission stream: the stream of retransmission packets associated with an original stream. Session-multiplexing: scheme by which the original stream and the associated retransmission stream are sent into two different RTP sessions. SSRC-multiplexing: scheme by which the original stream and the retransmission stream are sent in the same RTP session with different SSRC values. 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 RFC 2119 [2]. 3. Requirements and design rationale for a retransmission scheme The retransmission scheme is designed to fulfil the following set of requirements: 1. It must not break general RTP and RTCP mechanisms 2. It must be suitable for unicast and small multicast groups. 3. It must work with mixers and translators. 4. It must work with all known payload types. 5. It must not prevent the use multiple payload types in a session. 6. In order to support the largest variety of payload formats the RTP receiver must be able to indicate how many and which RTP packets were lost. This requirement is referred to as sequence number preservation. Without such a requirement, it would be impossible to use retransmission with payload formats, such as conversational text [10] or most audio/video streaming applications, that use the RTP sequence number to detect lost packets. When designing a solution for RTP retransmission, several approaches may be considered for the multiplexing of the original RTP packets and the retransmitted RTP packets. One approach may be to retransmit the RTP packet with its original sequence number and send original and retransmission packets in the same stream. The retransmission packet would then be identical to the original RTP packet, i.e. the same header (and thus same sequence number) and the same payload. However, such an approach is not acceptable because it would corrupt the RTCP statistics. As a consequence, requirement 1 would not be met. Correct RTCP statistics require that for every RTP packet within the RTP stream, the sequence number be increased by one. Another approach may be to multiplex original RTP packets and retransmission packets in the same stream using the payload type field. With such an approach the original stream and the Rey/Leon/Miyazaki/Varsa/Hakenberg 4 Internet Draft RTP Retransmission Payload Format December 2002 retransmission stream would share the same sequence number space. As a result, the RTP receiver would not be able to infer how many and which original packets (i.e. with which sequence number) were lost. In other words, this feature does not satisfy the sequence number preservation requirement (requirement 6). This in turn implies that requirement 4 would not be met. Interoperability with mixers and translators would also be more difficult if they do not understand this new payload type in a sender RTP stream. For these reasons, a solution based on payload type multiplexing of original packets and retransmission packets in the same RTP stream is excluded. Finally, the original and retransmission packets may be sent in two separate streams. These two streams may be multiplexed either by sending them in two different sessions , i.e. session-multiplexing, or in the same session using different SSRCs, i.e. SSRC-multi- plexing. Since original and retransmission packets carry media of the same type, the objections in Section 5.2 of RTP, RFC WWWW [3] to RTP multiplexing do not apply. Using two separate streams satisfies all the requirements listed in this section. Mixers and translators may process the original stream and simply discard the retransmission stream if they are unable to utilise it. 3.1 Multiplexing scheme choice Session-multiplexing and SSRC-multiplexing have different pros and cons: Session-multiplexing is based on sending the retransmission stream in a different RTP session (as defined in RTP [3]) from that of the original stream, i.e. the original and retransmission streams are sent to different network addresses and/or port numbers. Having a separate session allows more flexibility. In multicast, using two sessions for retransmission allows a receiver to choose whether to subscribe or not to the RTP session carrying the retransmission stream. It is also possible for the original session to be single- source multicast and have separate unicast sessions to convey retransmissions to each of the receivers, which will then receive only the retransmission packets they requested. The use of separate sessions also allows differential treatment by the network and may simplify processing in mixers, translators and packet caches. With SSRC-multiplexing, a single session is needed for the original and the retransmission stream. This allows streaming servers and middleware which are involved in a high number of concurrent sessions to minimise their port usage. This retransmission payload format allows both session-multiplexing and SSRC-multiplexing. From an implementation point of view, there Rey/Leon/Miyazaki/Varsa/Hakenberg 5 Internet Draft RTP Retransmission Payload Format December 2002 is little difference between the two approaches. Hence, in order to maximise interoperability, both multiplexing approaches SHOULD be supported. 4. Retransmission payload format The format of a retransmission packet is shown below: +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | RTP Header | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | OSN | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | Original RTP Packet Payload | | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ The RTP header usage is as follows: If the original and the retransmission streams are sent in separate RTP sessions, the same SSRC value MUST be used for the original stream and the retransmission stream. In case of an SSRC collision, an RTCP BYE packet MUST be sent for the original RTP session. After a new SSRC identifier is obtained, the SSRC of the retransmission session MUST be set to this value. If the original stream and the retransmission stream are sent in the same RTP session, two different SSRC values MUST be used for the original stream and the retransmission stream as required by RTP. For either multiplexing scheme, the sequence number has the standard definition, i.e. it MUST be one higher than the sequence number of the preceding packet sent in the retransmission stream. The retransmission packet timestamp is set to the original timestamp, i.e. to the timestamp of the original packet. As a consequence, the initial RTP timestamp for the first packet of the retransmission stream is not random but equal to the original timestamp of the first packet requested for retransmission. See the security considerations section in this document for security implications. Implementers have to be aware that the RTCP jitter value for the retransmission stream does not reflect the actual network jitter since there could be little correlation between the time a packet is retransmitted and its original timestamp. The payload type is dynamic. Each payload type of the original stream MUST map to a different payload type value in the retransmission stream. Therefore, when multiple payload types are Rey/Leon/Miyazaki/Varsa/Hakenberg 6 Internet Draft RTP Retransmission Payload Format December 2002 used in the original stream, multiple dynamic payload types will be mapped to this retransmission payload format. See Section 8 for the specification of how the mapping between original and retransmission payload types is done. As the retransmission packet timestamp carries the original media timestamp, the timestamp clockrate used by the retransmission payload type is the same as the one used by the associated original payload type. It is thus possible to retransmit RTP packets whose payload types have different timestamp clockrates in the same retransmission stream if the original payload types have different clock rates, but this is usually not the case. If the original RTP header carried any profile-specific payload header, the retransmission packet MUST include this payload header. If the original RTP header carried an RTP header extension, the retransmission packet SHOULD carry the same header extension. The retransmission payload carries a payload header followed by the original RTP packet payload. The length of payload header is 2 octets. The payload header contains only one field, OSN, which MUST be set to the sequence number of the associated original RTP packet. If the original RTP packet contained RTP padding, that padding must be removed before constructing the retransmission packet. If padding of the retransmission packet is needed, padding is performed as with any RTP packets and the padding bit is set. All other fields of the RTP header MUST have the same value as in the associated original RTP packet 5. Association of a retransmission stream with its original stream 5.1 Retransmission session sharing In the case of session-multiplexing, a retransmission session MUST map to exactly one original session, i.e. the same retransmission session cannot be used for different original sessions. If retransmission session sharing were allowed, a receiver joining the retransmission session would also receive the retransmissions belonging to all other original sessions which the receiver may have not joined. For example, a receiver wishing to receive only audio would receive retransmitted video packets if an audio and video session would share the same retransmission session. 5.2 CNAME use A sender MUST use the same CNAME for an original stream and its associated retransmission stream. Rey/Leon/Miyazaki/Varsa/Hakenberg 7 Internet Draft RTP Retransmission Payload Format December 2002 5.3 Association at the receiver A receiver receiving multiple original and retransmission streams needs to associate each retransmission stream with its original stream. The association is done differently depending on whether session-multiplexing or SSRC-multiplexing is used. If session-multiplexing is used, the receiver associates the two streams having the same SSRC in the two sessions. Note that the payload type field cannot be used to do this coupling as several media streams may have the same payload type value. The two sessions are themselves associated out-of-band. See the SDP section to see how the grouping of the two sessions is done with SDP. If SSRC-multiplexing is used, the receiver should first of all look for two streams that have the same CNAME in the session. In some cases, the CNAME may not be enough to determine the association as multiple original streams in the same session may share the same CNAME. For example, there can be in the same video session multiple video streams mapping to different SSRCs and still using the same CNAME and possibly the same PT values. Each (or some of) these streams may have an associated retransmission stream. In order to find out the association between original and retransmission streams having the same CNAME, the receiver SHOULD behave as follows. The association can generally be resolved when the receiver receives a retransmission packet matching a retransmission request which had been sent earlier. Upon reception of a retransmission whose original sequence number had been previously requested, the receiver can derive that the SSRC of the retransmission packet is associated to the sender SSRC from which the packet was requested. In order to avoid ambiguity, the receiver MUST NOT have two outstanding requests for the same packet sequence number in two different original streams before the association is resolved. Note that since the initial packet timestamps are random, the probability of having two outstanding requests for the same packet sequence number would be very small. If the receiver discovers that two senders are using the same SSRC or if it receives an RTCP BYE packet, it MUST stop requesting retransmissions for that SSRC. Upon reception of original RTP packets with a new SSRC, the receiver MUST perform the SSRC association again as described in this section. 6. Use with the extended RTP profile for RTCP-based feedback This section gives general hints for the usage of this payload format with the extended RTP profile for RTCP-based feedback [1], denoted AVPF. Rey/Leon/Miyazaki/Varsa/Hakenberg 8 Internet Draft RTP Retransmission Payload Format December 2002 6.1 RTCP Receiver reports If the original RTP stream and the retransmission stream are sent to separate RTP sessions, the receiver will then send report blocks for the original stream and the retransmission stream in separate RTCP receiver reports (RR) packets belonging to separate RTP sessions. RTCP packets reporting on the original stream are sent in the original RTP session while RTCP packets reporting on the retransmission stream are sent in the retransmission session. The RTCP bandwidth for these two sessions may be chosen independently (for example through RTCP bandwidth modifiers RFC ZZZZ [4]). If the original RTP stream and the retransmission stream are sent in the same session (SSRC multiplexing), the receiver sends report blocks for the original and the retransmission streams in the same RTCP RR packet. 6.2 Retransmission requests The NACK message format defined in the AVPF profile SHOULD be used by receivers to send retransmission requests. Whether a receiver chooses to request a packet or not is an implementation issue. An actual receiver implementation should take into account such factors as the tolerable application delay, the network environment and the media type. The receiver should generally assess whether the retransmitted packet would still be useful at the time it is received. The timestamp of the missing packet can be estimated from the timestamps of packets preceding and/or following the sequence number gap caused by the missing packet in the original stream. In most cases, some form of linear estimate of the timestamp is good enough. Furthermore, a receiver should compute an estimate of the round-trip time (RTT) to the sender. This can be done, for example, by measuring the retransmission delay to receive a retransmission packet after a NACK message has been sent for that packet. This estimate may also be obtained from past observations, RTCP report round-trip time if available or any other means. The receiver should not send a retransmission request as soon as it detects a missing sequence number but should add some extra delay to compensate for packet reordering. This extra delay may, for example, be based on past observations of the experienced packet reordering. To increase the robustness to the loss of a NACK message or of a retransmission packet, a receiver may send a new NACK message. This is referred to as multiple retransmissions. Before sending a new NACK message for a missing packet, the receiver should rely on a timer to be reasonably sure that the previous retransmission attempt has failed in order not to cause unnecessary retransmissions. Rey/Leon/Miyazaki/Varsa/Hakenberg 9 Internet Draft RTP Retransmission Payload Format December 2002 NACK packets MUST be sent only for the original RTP stream. If a receiver wanted to perform multiple-retransmissions by sending a NACK in the retransmission stream, it would not be able to know the original sequence number and a timestamp estimation of the packet it requests. 6.3 Timing rules The RTCP NACK packet may be sent in a regular full compound RTCP packet or in an early RTCP packet, as per AVPF [1]. Sending a NACK in an early packet allows to react more quickly to a given packet loss. However, in that case if a new packet loss occurs right after the early RTCP packet was sent, the receiver will then have to wait for the next regular RTCP compound packet after the early packet. Sending NACK packets only in regular RTCP compound decreases the maximum delay between detecting an original packet loss and being able to send a NACK message for that packet. Implementers should consider the possible implications of this fact for the application being used. Furthermore, receivers may make use of the minimum interval between regular RTCP compound packets. This can be used, for example, to keep reception reporting down to a given minimum, while still allowing receivers to react to periods requiring more frequent feedback, e.g. times of higher packet loss rate. In this way, receivers will try to keep the amount of sent RTCP packets as low as specified by the minimum interval, but are still able to report packet losses quickly enough. Note that although RTCP packets may be suppressed because they do not contain NACK packets, the reserved RTCP bandwidth is the same as if they were sent. See AVPF [1] for details. 7. Congestion control RTP retransmission poses a risk of increasing network congestion. In a best-effort environment, packet loss is caused by congestion. Reacting to loss by retransmission of older data without decreasing the rate of the original stream would thus further increase congestion. Implementations SHOULD follow the recommendations below in order to use retransmission. The RTP profile under which the retransmission scheme is used defines an appropriate congestion control mechanism in different environments. Following the rules under the profile, an RTP application can determine its acceptable bitrate and packet rate in order to be fair to other TCP or RTP flows. If an RTP application uses retransmission, the acceptable packet rate and bitrate includes both the original and retransmitted data. This guarantees that an application using retransmission achieves the same fairness as one that does not. Such a rule would translate in practice into the following actions: Rey/Leon/Miyazaki/Varsa/Hakenberg 10 Internet Draft RTP Retransmission Payload Format December 2002 If enhanced service is used, it should be made sure that the total bitrate and packet rate do not exceed that of the requested service. It should be further monitored that the requested services are actually delivered. In a best-effort environment, the sender SHOULD NOT send retransmission packets without reducing the packet rate and bitrate of the original stream (for example by encoding the data at a lower rate). In addition, the sender MAY selectively retransmit only the packets that it deems important and ignore NACK messages for other packets in order to limit the bitrate. These congestion control mechanisms should keep the packet loss rate within acceptable parameters. Packet loss is considered acceptable if a TCP flow across the same network path and experiencing the same network conditions would achieve an average throughput, measured on a reasonable timescale, that is not less than the RTP flow is achieving. If the packet loss rate exceed acceptable parameters, this would mean that congestion is not kept under control and retransmission should then not be used. It may further be necessary to adapt the transmission rate (or the number of layers subscribed for a layered multicast session), or to arrange for the receiver to leave the session if the loss rate is unacceptably high. 8. SDP usage 8.1 Introduction This section specifies how to describe the use of retransmission with the Session Description Protocol (SDP), RFC 2327 [5]. As specified in this document, the retransmission stream may be conveyed in a separate RTP session, i.e. through session- multiplexing, or in the same RTP session as the original stream through SSRC-multiplexing. The following attributes and parameters are introduced in this document: "rtx", "rtx-time" and "apt". The binding used for the retransmission stream to the payload type number is indicated by an rtpmap attribute. The MIME subtype name used in the binding is "rtx", as specified in Section 11. An OPTIONAL payload format-specific parameter indicates the maximum time a server will try to retransmit a packet. The syntax is as follows: a=fmtp : rtx-time= where, indicates the dynamic payload type number assigned to the retransmission payload format in an rtpmap attribute. Rey/Leon/Miyazaki/Varsa/Hakenberg 11 Internet Draft RTP Retransmission Payload Format December 2002 indicates the time in milliseconds, measured from the time a packet was first sent until the time the server will stop trying to retransmit the packet. The absence of the rtx-time parameter for a retransmission stream means that the maximum retransmission time is not defined, but MAY be negotiated by other means. Additionally, a new SDP payload-format-specific parameter "apt" MUST be used to map the RTX payload type to the associated original stream payload type as seen in the SDP description examples below. If multiple payload types are used in the original stream, then multiple "apt" parameters MUST be included to map each original stream payload type to a different RTX payload type. The syntax of this parameter is as follows: a=fmtp : apt= where, indicates the dynamic payload type number assigned to the retransmission payload format. indicates the original stream payload type to which this retransmission stream payload type is associated. Some SDP description examples are presented in the following subsections. 8.2 Mapping MIME Parameters into SDP The information carried in the MIME media type specification has a specific mapping to fields in SDP [5], which is commonly used to describe RTP sessions. When SDP is used to specify retransmissions for an RTP stream, the mapping is done as follows: - The MIME types ("video"), ("audio") and ("text") go in the SDP "m=" as the media name. - The MIME subtype ("rtx") goes in SDP "a=rtpmap" as the encoding name. The RTP clock rate in "a=rtpmap" MUST be that of the retransmission payload type. See Section 4 for details on this. - The AVPF profile-specific parameters "ack" and "nack" go in SDP "a=rtcp-fb". Several SDP "a=rtcp-fb" are used for several types of feedback. See the AVPF profile [1] for details. - The retransmission payload format-specific parameters "apt" and "rtx-time" go in the SDP "a=fmtp" as a semicolon separated list of parameter=value pairs. - Any remaining parameters go in the SDP "a=fmtp" attribute by copying them directly from the MIME media type string as a semicolon separated list of parameter=value pairs. Rey/Leon/Miyazaki/Varsa/Hakenberg 12 Internet Draft RTP Retransmission Payload Format December 2002 In the following sections some example SDP descriptions are presented. 8.3 SDP description with session-multiplexing In the case of session-multiplexing, the SDP description contains one media specification "m" line per RTP session. The SDP MUST provide the grouping of the original and associated retransmission sessions' "m" lines, using the Flow Identification (FID) semantics defined in RFC YYYY [6]. The following example specifies two original, AMR and MPEG-4, streams on ports 49170 and 49174 and their corresponding retransmission streams on ports 49172 and 49176, respectively: v=0 o=mascha 2980675221 2980675778 IN IP4 at.home.ru c=IN IP4 125.25.5.1 a=group:FID 1 2 a=group:FID 3 4 m=audio 49170 RTP/AVPF 96 a=rtpmap:96 AMR/8000 a=fmtp:96 octet-align=1 a=rtcp-fb:96 nack a=mid:1 m=audio 49172 RTP/AVPF 97 a=rtpmap:97 rtx/8000 a=fmtp:97 apt=96;rtx-time=3000 a=mid:2 m=video 49174 RTP/AVPF 98 a=rtpmap:98 MP4V-ES/90000 a=rtcp-fb:98 nack a=fmtp:98 profile-level-id=8;config=01010000012000884006682C2090A21F a=mid:3 m=video 49176 RTP/AVPF 99 a=rtpmap:99 rtx/90000 a=fmtp:99 apt=98;rtx-time=3000 a=mid:4 A special case of the SDP description is a description that contains only one original session "m" line and one retransmission session "m" line, the grouping is then obvious and FID semantics MAY be omitted in this special case only. This is illustrated in the following example, which is an SDP description for a single original MPEG-4 stream and its corresponding retransmission session: v=0 o=mascha 2980675221 2980675778 IN IP4 at.home.ru c=IN IP4 125.25.5.1 m=video 49170 RTP/AVPF 96 Rey/Leon/Miyazaki/Varsa/Hakenberg 13 Internet Draft RTP Retransmission Payload Format December 2002 a=rtpmap:96 MP4V-ES/90000 a=rtcp-fb:96 nack a=fmtp:96 profile-level-id=8;config=01010000012000884006682C2090A21F m=video 49172 RTP/AVPF 97 a=rtpmap:97 rtx/90000 a=fmtp:97 apt=96;rtx-time=3000 8.4 SDP description with SSRC-multiplexing The following is an example of an SDP description for an RTP video session using SSRC-multiplexing with similar parameters as in the single-session example above: v=0 o=mascha 2980675221 2980675778 IN IP4 at.home.ru c=IN IP4 125.25.5.1 m=video 49170 RTP/AVPF 96 97 a=rtpmap:96 MP4V-ES/90000 a=rtcp-fb:96 nack a=fmtp:96 profile-level-id=8;config=01010000012000884006682C2090A21F a=rtpmap:97 rtx/90000 a=fmtp:97 apt=96;rtx-time=3000 9. RTSP considerations The Real-time Streaming Protocol (RTSP), RFC 2326 [7] is an application-level protocol for control over the delivery of data with real-time properties. This section looks at the issues involved in controlling RTP sessions that use retransmissions. 9.1 RTSP control with SSRC-multiplexing In the case of SSRC-multiplexing, the "m" line includes both original and retransmission payload types and has a single RTSP "control" attribute. The receiver uses the "m" line to request SETUP and TEARDOWN of the whole media session. The RTP profile contained in the transport header MUST be the AVPF profile or another suitable profile allowing extended feedback. In order to control the sending of the session original media stream, the receiver sends as usual PLAY and PAUSE requests to the sender for the session. The RTP-info header that is used to set RTP- specific parameters in the PLAY response MUST be set according to the RTP information of the original stream. When the receiver starts receiving the original stream, it can then request retransmission through RTCP NACKs without additional RTSP signalling. 9.2 RTSP control with session-multiplexing Rey/Leon/Miyazaki/Varsa/Hakenberg 14 Internet Draft RTP Retransmission Payload Format December 2002 In the case of session-multiplexing, each SDP "m" line has an RTSP "control" attribute. Hence, when retransmission is used, both the original session and the retransmission have their own "control" attributes. The receiver can associate the original session and the retransmission session through the FID semantics as specified in Section 8. The original and the retransmission streams are set up and torn down separately through their respective media "control" attribute. The RTP profile contained in the transport header MUST be the AVPF profile or another suitable profile allowing extended feedback for both the original and the retransmission session. The RTSP presentation SHOULD support aggregate control and SHOULD contain a session level RTSP URL. The receiver SHOULD use aggregate control for an original session and its associated retransmission session. Otherwise, there would need to be two different 'session- id' values, i.e. different values for the original and retransmission sessions, and the sender would not know how to associate them. The session-level "control" attribute is then used as usual to control the playing of the original stream. When the receiver starts receiving the original stream, it can then request retransmissions through RTCP without additional RTSP signalling. 9.3 Retransmission in pause state Because of the nature of retransmissions, the sending of retransmission packets SHOULD NOT be controlled through RTSP PLAY and PAUSE requests. The PLAY and PAUSE requests should not affect the retransmission stream. Retransmission packets are sent upon receiver requests in the original RTCP stream, regardless of the state. 9.4 Cache control Retransmission streams SHOULD NOT be cached. In the case of session-multiplexing, the "Cache-Control" header SHOULD be set to "no-cache" for the retransmission stream. In the case of SSRC-multiplexing, RTSP cannot specify independent caching for the retransmission stream, because there is a single "m" line in SDP. Therefore, the implementer should take this fact into account when deciding whether to cache an SSRC-multiplexed session or not. 10. Implementation examples This document mandates only the sender and receiver behaviours that are necessary for interoperability. In addition, certain algorithms, Rey/Leon/Miyazaki/Varsa/Hakenberg 15 Internet Draft RTP Retransmission Payload Format December 2002 such as rate control or buffer management when targeted at specific environments, may enhance the retransmission efficiency. This section gives an overview of different implementation options allowed within this specification. The first example is a server-driven retransmission implementation. With this implementation, it is possible to retransmit lost RTP packets, detect efficiently the loss of retransmissions and perform multiple retransmissions, if needed. Most of the necessary processing is done at the server. The second example shows a receiver-driven implementation. It illustrates how a receiver may increase the retransmission efficiency. This implementation also increases the sender scalability by reducing the load required at the sender. The third example shows how retransmissions may be used in (small) multicast groups in conjunction with layered encoding. It illustrates that retransmissions and layered encoding may be complementary techniques. 10.1 A sender-driven retransmission example This section gives an implementation example of multiple retransmissions. The sender transmits the original data in RTP packets using the MPEG-4 video RTP payload format. It is assumed that Generic NACK feedback messages are used, as per [1]. An SDP description example with SSRC-multiplexing is given below: v=0 o=mascha 2980675221 2980675778 IN IP4 at.home.ru c=IN IP4 125.25.5.1 m=video 49170 RTP/AVPF 96 97 a=rtpmap:96 MP4V-ES/90000 a=rtcp-fb:96 nack a=rtpmap:97 rtx/90000 a=fmtp:97 apt=96;rtx-time=3000 The format-specific parameter "rtx-time" would indicate that the server will buffer the sent packets in a retransmission buffer for 3.0 seconds, after which the packets are deleted from the retransmission buffer and will never be sent again. In this implementation example, the required RTP receiver processing to handle retransmission is very limited. The receiver detects packet loss from the gaps observed in the received sequence numbers. It signals lost packets to the sender through RTCP NACK messages as defined in the AVPF profile [1]. The receiver should take into account the signalled sender retransmission buffer length in order to dimension its own reception buffer. It should also derive from the Rey/Leon/Miyazaki/Varsa/Hakenberg 16 Internet Draft RTP Retransmission Payload Format December 2002 buffer length the maximum number of times retransmission of a packet can be requested. The sender should retransmit the packets selectively, i.e. it should choose whether to retransmit a requested packet depending on the packet importance, the observed QoS and congestion state of the network connection to the receiver. Obviously, the sender processing increases with the number of receivers as state information and processing load must be allocated to each receiver. 10.2 A receiver-driven retransmission example The receiver may have more accurate information than the sender about the current network QoS such as available bandwidth, packet loss rate, delay and jitter. In addition, other receiver-specific parameters like buffer level, estimated importance of the lost packet and application level QoS may be used by the receiver to make a more efficient use of RTP retransmission through selective requests. Furthermore, a receiver may acknowledge the received packets. This can be done by sending ACK messages, as per [1]. Upon receiving an ACK, the sender may delete all the acknowledged packets from its retransmission buffer. Note that this would also require only limited increase in the required RTCP bandwidth as long as ACK packets are sent seldom enough. With the receiver-driven retransmission implementation, processing load and buffer requirements at the sender are decreased, allowing greater sender scalability. Note that choosing between the sender-driven implementation and the receiver-driven implementation does not imply any changes in the SDP description, except for the need to signal the use of ACK RTCP packets, by means of an additional SDP "a=rtcp-fb" line, as follows: v=0 o=mascha 2980675221 2980675778 IN IP4 at.home.ru c=IN IP4 125.25.5.1 m=video 49170 RTP/AVPF 96 97 a=rtpmap:96 MP4V-ES/90000 a=rtcp-fb:96 nack a=rtcp-fb:96 ack a=rtpmap:97 rtx/90000 a=fmtp:97 apt=96;rtx-time=3000 10.3 Retransmissions with Layered Transmissions This section shows how to combine retransmissions with layered encoding. Note that the retransmission framework is not intended as a complete solution to reliable multicast. Refer to RFC 2887 [11], for an overview of the problems related with reliable multicast transmission. Packets of different importance are sent in different RTP sessions. The retransmission streams corresponding to the different layers can Rey/Leon/Miyazaki/Varsa/Hakenberg 17 Internet Draft RTP Retransmission Payload Format December 2002 themselves be seen as different retransmission layers. The relative importance of the different retransmission streams should reflect the relative importance of the different original streams. A retransmission stream may be sent in the same RTP session as its corresponding original layer through SSRC multiplexing or in a different RTP session through session multiplexing. An SDP description example for SSRC-multiplexing is given below: c=IN IP4 224.2.1.1/127/3 m=video 8000 RTP/AVPF 98 99 a=rtpmap:98 MP4V-ES/90000 a=rtcp-fb:98 nack a=rtpmap:99 rtx/90000 a=fmtp:99 apt=98;rtx-time=3000 The server and the receiver may implement the retransmission methods illustrated in the previous examples. In addition, they may choose to request and retransmit a lost packet depending on the layer it belongs to. 11. IANA considerations 11.1 Registration of audio/rtx MIME type: audio MIME subtype: rtx Required parameters: rate: the RTP timestamp clockrate is equal to the RTP timestamp clockrate of the media that is retransmitted. apt: associated payload type. The value of this parameter is the payload type of the associated original stream. Optional parameters: rtx-time: indicates the time in milliseconds, measured from the time a packet was first sent until the time the server will stop trying to retransmit the packet Encoding considerations: this type is only defined for transfer via RTP. Security considerations: see Section 12 of RFC XXXX Interoperability considerations: none Rey/Leon/Miyazaki/Varsa/Hakenberg 18 Internet Draft RTP Retransmission Payload Format December 2002 Published specification: RFC XXXX Applications which use this media type: multimedia streaming applications Additional information: none Person & email address to contact for further information: rey@panasonic.de david.leon@nokia.com avt@ietf.org Intended usage: COMMON Author/Change controller: Jose Rey David Leon IETF AVT WG 11.2 Registration of video/rtx MIME type: video MIME subtype: rtx Required parameters: rate: the RTP timestamp clockrate is equal to the RTP timestamp clockrate of the media that is retransmitted. apt: associated payload type. The value of this parameter is the payload type of the associated original stream. Optional parameters: rtx-time: indicates the time in milliseconds, measured from the time a packet was first sent until the time the server will stop trying to retransmit the packet Encoding considerations: this type is only defined for transfer via RTP. Security considerations: see Section 12 of RFC XXXX Interoperability considerations: none Published specification: RFC XXXX Applications which use this media type: multimedia streaming applications Additional information: none Rey/Leon/Miyazaki/Varsa/Hakenberg 19 Internet Draft RTP Retransmission Payload Format December 2002 Person & email address to contact for further information: rey@panasonic.de david.leon@nokia.com avt@ietf.org Intended usage: COMMON Author/Change controller: Jose Rey David Leon IETF AVT WG 11.3 Registration of text/rtx MIME type: text MIME subtype: rtx Required parameters: rate: the RTP timestamp clockrate is equal to the RTP timestamp clockrate of the media that is retransmitted. apt: associated payload type. The value of this parameter is the payload type of the associated original stream. Optional parameters: rtx-time: indicates the time in milliseconds, measured from the time a packet was first sent until the time the server will stop trying to retransmit the packet Encoding considerations: this type is only defined for transfer via RTP. Security considerations: see Section 12 of RFC XXXX Interoperability considerations: none Published specification: RFC XXXX Applications which use this media type: multimedia streaming applications Additional information: none Person & email address to contact for further information: rey@panasonic.de david.leon@nokia.com avt@ietf.org Intended usage: COMMON Rey/Leon/Miyazaki/Varsa/Hakenberg 20 Internet Draft RTP Retransmission Payload Format December 2002 Author/Change controller: Jose Rey David Leon IETF AVT WG 11.4 Registration of application/rtx MIME type: application MIME subtype: rtx Required parameters: rate: the RTP timestamp clockrate is equal to the RTP timestamp clockrate of the media that is retransmitted. apt: associated payload type. The value of this parameter is the payload type of the associated original stream. Optional parameters: rtx-time: indicates the time in milliseconds, measured from the time a packet was first sent until the time the server will stop trying to retransmit the packet Encoding considerations: this type is only defined for transfer via RTP. Security considerations: see Section 12 of RFC XXXX Interoperability considerations: none Published specification: RFC XXXX Applications which use this media type: multimedia streaming applications Additional information: none Person & email address to contact for further information: rey@panasonic.de david.leon@nokia.com avt@ietf.org Intended usage: COMMON Author/Change controller: Jose Rey David Leon IETF AVT WG Rey/Leon/Miyazaki/Varsa/Hakenberg 21 Internet Draft RTP Retransmission Payload Format December 2002 12. Security considerations Applications utilising encryption SHOULD encrypt both the original and the retransmission stream. Old keys will likely need to be cached so that when the keys change for the original stream, the old key is used until it is determined that the key has changed on the retransmission packets as well. The use of the same key for the retransmitted stream and the original stream may lead to security problems, e.g. two-time pads. This sharing has to be evaluated towards the chosen security protocol and security algorithms, e.g. the Secure Real-Time Transport Protocol (SRTP) RFC UUUU [8] establishes requirements for avoiding the two-time pad. RTP recommends that the initial RTP timestamp SHOULD be random to secure the stream against known plain text attacks. This payload format does not follow this recommendation as the initial timestamp will be the media timestamp of the first retransmitted packet. However, since the initial timestamp of the original stream is itself random, if the original stream is encrypted, the first retransmitted packet timestamp would also be random to an attacker. Therefore, security would not be compromised. Congestion control considerations with the use of retransmission are dealt with in Section 7 of this document. Any other security considerations of the profile under which the retransmission scheme is used should be applied. 13. Acknowledgements We would like to express our gratitude to Carsten Burmeister for his participation in the development of this document. Our thanks also go to Koichi Hata, Colin Perkins, Stephen Casner, Magnus Westerlund, Go Hori and Rahul Agarwal for their helpful comments. 14. References 14.1 Normative References 1 J. Ott, S. Wenger, N. Sato, C. Burmeister, J. Rey, "Extended RTP profile for RTCP-based feedback", RFC VVVV, September 2002. 2 S. Bradner, "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997 3 H. Schulzrinne, S. Casner, R. Frederick and V. Jacobson, "RTP: A Transport Protocol for Real-Time Applications", RFC WWWW, May 2002. Rey/Leon/Miyazaki/Varsa/Hakenberg 22 Internet Draft RTP Retransmission Payload Format December 2002 4 S. Casner, "SDP bandwidth modifiers for RTCP bandwidth", RFC ZZZZ, May 2002. 5 M. Handley, V. Jacobson, "SDP: Session Description Protocol", RFC 2327, April 1998. 6 G. Camarillo,J. Holler, G. AP. Eriksson, "Grouping of media lines in SDP", RFC YYYY, February 2002. 7 H. Schulzrinne, A. Rao, R. Lanphier, "Real Time Streaming Protocol (RTSP)", RFC 2326, April 1998. 8 M. Baugher, D. A. McGrew, D. Oran, R. Blom, E. Carrara, M. Naslund, K. Norrman, "The Secure Real-Time Transport Protocol", RFC UUUU, June 2002. 14.2 Non-normative References 9 C. Perkins, O. Hodson, "Options for Repair of Streaming Media", RFC 2354, June 1998. 10 J. Hellstrom, "RTP for conversational text", RFC 2793, May 2000 11 M. Handley, et al., "The Reliable Multicast Design Space for Bulk Data Transfer", RFC 2887, August 2000. Author's Addresses Jose Rey rey@panasonic.de Panasonic European Laboratories GmbH Monzastr. 4c D-63225 Langen, Germany Phone: +49-6103-766-134 Fax: +49-6103-766-166 David Leon david.leon@nokia.com Nokia Research Center 6000 Connection Drive Irving, TX. USA Phone: 1-972-374-1860 Akihiro Miyazaki akihiro@isl.mei.co.jp Core Software Development Center Corporate Software Development Division Matsushita Electric Industrial Co., Ltd. 1006 Kadoma, Kadoma City, Osaka 571-8501, Japan Phone: +81-6-6900-9192 Fax: +81-6-6900-9193 Viktor Varsa viktor.varsa@nokia.com Nokia Research Center Rey/Leon/Miyazaki/Varsa/Hakenberg 23 Internet Draft RTP Retransmission Payload Format December 2002 6000 Connection Drive Irving, TX. USA Phone: 1-972-374-1861 Rolf Hakenberg hakenberg@panasonic.de Panasonic European Laboratories GmbH Monzastr. 4c D-63225 Langen, Germany Phone: +49-6103-766-162 Fax: +49-6103-766-166 Rey/Leon/Miyazaki/Varsa/Hakenberg 24