idnits 2.17.1 draft-ietf-rtcweb-rtp-usage-24.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (May 29, 2015) is 3254 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: A later version (-13) exists of draft-ietf-avtcore-multi-media-rtp-session-07 == Outdated reference: A later version (-18) exists of draft-ietf-avtcore-rtp-circuit-breakers-10 == Outdated reference: A later version (-11) exists of draft-ietf-avtcore-rtp-multi-stream-07 == Outdated reference: A later version (-12) exists of draft-ietf-avtcore-rtp-multi-stream-optimisation-05 == Outdated reference: A later version (-54) exists of draft-ietf-mmusic-sdp-bundle-negotiation-19 == Outdated reference: A later version (-11) exists of draft-ietf-rtcweb-audio-08 == Outdated reference: A later version (-10) exists of draft-ietf-rtcweb-fec-01 == Outdated reference: A later version (-12) exists of draft-ietf-rtcweb-security-08 == Outdated reference: A later version (-20) exists of draft-ietf-rtcweb-security-arch-11 == Outdated reference: A later version (-06) exists of draft-ietf-rtcweb-video-05 ** Obsolete normative reference: RFC 4566 (Obsoleted by RFC 8866) ** Obsolete normative reference: RFC 5285 (Obsoleted by RFC 8285) == Outdated reference: A later version (-12) exists of draft-ietf-avtcore-multiplex-guidelines-03 == Outdated reference: A later version (-10) exists of draft-ietf-avtcore-rtp-topologies-update-07 == Outdated reference: A later version (-08) exists of draft-ietf-avtext-rtp-grouping-taxonomy-06 == Outdated reference: A later version (-17) exists of draft-ietf-mmusic-msid-10 == Outdated reference: A later version (-19) exists of draft-ietf-rtcweb-overview-13 == Outdated reference: A later version (-18) exists of draft-ietf-tsvwg-rtcweb-qos-03 -- Obsolete informational reference (is this intentional?): RFC 5245 (Obsoleted by RFC 8445, RFC 8839) Summary: 2 errors (**), 0 flaws (~~), 17 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 RTCWEB Working Group C. Perkins 3 Internet-Draft University of Glasgow 4 Intended status: Standards Track M. Westerlund 5 Expires: November 30, 2015 Ericsson 6 J. Ott 7 Aalto University 8 May 29, 2015 10 Web Real-Time Communication (WebRTC): Media Transport and Use of RTP 11 draft-ietf-rtcweb-rtp-usage-24 13 Abstract 15 The Web Real-Time Communication (WebRTC) framework provides support 16 for direct interactive rich communication using audio, video, text, 17 collaboration, games, etc. between two peers' web-browsers. This 18 memo describes the media transport aspects of the WebRTC framework. 19 It specifies how the Real-time Transport Protocol (RTP) is used in 20 the WebRTC context, and gives requirements for which RTP features, 21 profiles, and extensions need to be supported. 23 Status of This Memo 25 This Internet-Draft is submitted in full conformance with the 26 provisions of BCP 78 and BCP 79. 28 Internet-Drafts are working documents of the Internet Engineering 29 Task Force (IETF). Note that other groups may also distribute 30 working documents as Internet-Drafts. The list of current Internet- 31 Drafts is at http://datatracker.ietf.org/drafts/current/. 33 Internet-Drafts are draft documents valid for a maximum of six months 34 and may be updated, replaced, or obsoleted by other documents at any 35 time. It is inappropriate to use Internet-Drafts as reference 36 material or to cite them other than as "work in progress." 38 This Internet-Draft will expire on November 30, 2015. 40 Copyright Notice 42 Copyright (c) 2015 IETF Trust and the persons identified as the 43 document authors. All rights reserved. 45 This document is subject to BCP 78 and the IETF Trust's Legal 46 Provisions Relating to IETF Documents 47 (http://trustee.ietf.org/license-info) in effect on the date of 48 publication of this document. Please review these documents 49 carefully, as they describe your rights and restrictions with respect 50 to this document. Code Components extracted from this document must 51 include Simplified BSD License text as described in Section 4.e of 52 the Trust Legal Provisions and are provided without warranty as 53 described in the Simplified BSD License. 55 Table of Contents 57 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 58 2. Rationale . . . . . . . . . . . . . . . . . . . . . . . . . . 4 59 3. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 4 60 4. WebRTC Use of RTP: Core Protocols . . . . . . . . . . . . . . 5 61 4.1. RTP and RTCP . . . . . . . . . . . . . . . . . . . . . . 5 62 4.2. Choice of the RTP Profile . . . . . . . . . . . . . . . . 7 63 4.3. Choice of RTP Payload Formats . . . . . . . . . . . . . . 8 64 4.4. Use of RTP Sessions . . . . . . . . . . . . . . . . . . . 9 65 4.5. RTP and RTCP Multiplexing . . . . . . . . . . . . . . . . 10 66 4.6. Reduced Size RTCP . . . . . . . . . . . . . . . . . . . . 11 67 4.7. Symmetric RTP/RTCP . . . . . . . . . . . . . . . . . . . 11 68 4.8. Choice of RTP Synchronisation Source (SSRC) . . . . . . . 11 69 4.9. Generation of the RTCP Canonical Name (CNAME) . . . . . . 12 70 4.10. Handling of Leap Seconds . . . . . . . . . . . . . . . . 13 71 5. WebRTC Use of RTP: Extensions . . . . . . . . . . . . . . . . 13 72 5.1. Conferencing Extensions and Topologies . . . . . . . . . 13 73 5.1.1. Full Intra Request (FIR) . . . . . . . . . . . . . . 15 74 5.1.2. Picture Loss Indication (PLI) . . . . . . . . . . . . 15 75 5.1.3. Slice Loss Indication (SLI) . . . . . . . . . . . . . 15 76 5.1.4. Reference Picture Selection Indication (RPSI) . . . . 16 77 5.1.5. Temporal-Spatial Trade-off Request (TSTR) . . . . . . 16 78 5.1.6. Temporary Maximum Media Stream Bit Rate Request 79 (TMMBR) . . . . . . . . . . . . . . . . . . . . . . . 16 80 5.2. Header Extensions . . . . . . . . . . . . . . . . . . . . 16 81 5.2.1. Rapid Synchronisation . . . . . . . . . . . . . . . . 17 82 5.2.2. Client-to-Mixer Audio Level . . . . . . . . . . . . . 17 83 5.2.3. Mixer-to-Client Audio Level . . . . . . . . . . . . . 18 84 5.2.4. Media Stream Identification . . . . . . . . . . . . . 18 85 5.2.5. Coordination of Video Orientation . . . . . . . . . . 18 86 6. WebRTC Use of RTP: Improving Transport Robustness . . . . . . 18 87 6.1. Negative Acknowledgements and RTP Retransmission . . . . 19 88 6.2. Forward Error Correction (FEC) . . . . . . . . . . . . . 20 89 7. WebRTC Use of RTP: Rate Control and Media Adaptation . . . . 20 90 7.1. Boundary Conditions and Circuit Breakers . . . . . . . . 21 91 7.2. Congestion Control Interoperability and Legacy Systems . 21 92 8. WebRTC Use of RTP: Performance Monitoring . . . . . . . . . . 22 93 9. WebRTC Use of RTP: Future Extensions . . . . . . . . . . . . 22 94 10. Signalling Considerations . . . . . . . . . . . . . . . . . . 23 95 11. WebRTC API Considerations . . . . . . . . . . . . . . . . . . 24 96 12. RTP Implementation Considerations . . . . . . . . . . . . . . 27 97 12.1. Configuration and Use of RTP Sessions . . . . . . . . . 27 98 12.1.1. Use of Multiple Media Sources Within an RTP Session 27 99 12.1.2. Use of Multiple RTP Sessions . . . . . . . . . . . . 28 100 12.1.3. Differentiated Treatment of RTP Packet Streams . . . 33 101 12.2. Media Source, RTP Packet Streams, and Participant 102 Identification . . . . . . . . . . . . . . . . . . . . . 35 103 12.2.1. Media Source Identification . . . . . . . . . . . . 35 104 12.2.2. SSRC Collision Detection . . . . . . . . . . . . . . 36 105 12.2.3. Media Synchronisation Context . . . . . . . . . . . 37 106 13. Security Considerations . . . . . . . . . . . . . . . . . . . 37 107 14. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 39 108 15. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 39 109 16. References . . . . . . . . . . . . . . . . . . . . . . . . . 39 110 16.1. Normative References . . . . . . . . . . . . . . . . . . 39 111 16.2. Informative References . . . . . . . . . . . . . . . . . 42 112 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 44 114 1. Introduction 116 The Real-time Transport Protocol (RTP) [RFC3550] provides a framework 117 for delivery of audio and video teleconferencing data and other real- 118 time media applications. Previous work has defined the RTP protocol, 119 along with numerous profiles, payload formats, and other extensions. 120 When combined with appropriate signalling, these form the basis for 121 many teleconferencing systems. 123 The Web Real-Time communication (WebRTC) framework provides the 124 protocol building blocks to support direct, interactive, real-time 125 communication using audio, video, collaboration, games, etc., between 126 two peers' web-browsers. This memo describes how the RTP framework 127 is to be used in the WebRTC context. It proposes a baseline set of 128 RTP features that are to be implemented by all WebRTC Endpoints, 129 along with suggested extensions for enhanced functionality. 131 This memo specifies a protocol intended for use within the WebRTC 132 framework, but is not restricted to that context. An overview of the 133 WebRTC framework is given in [I-D.ietf-rtcweb-overview]. 135 The structure of this memo is as follows. Section 2 outlines our 136 rationale in preparing this memo and choosing these RTP features. 137 Section 3 defines terminology. Requirements for core RTP protocols 138 are described in Section 4 and suggested RTP extensions are described 139 in Section 5. Section 6 outlines mechanisms that can increase 140 robustness to network problems, while Section 7 describes congestion 141 control and rate adaptation mechanisms. The discussion of mandated 142 RTP mechanisms concludes in Section 8 with a review of performance 143 monitoring and network management tools. Section 9 gives some 144 guidelines for future incorporation of other RTP and RTP Control 145 Protocol (RTCP) extensions into this framework. Section 10 describes 146 requirements placed on the signalling channel. Section 11 discusses 147 the relationship between features of the RTP framework and the WebRTC 148 application programming interface (API), and Section 12 discusses RTP 149 implementation considerations. The memo concludes with security 150 considerations (Section 13) and IANA considerations (Section 14). 152 2. Rationale 154 The RTP framework comprises the RTP data transfer protocol, the RTP 155 control protocol, and numerous RTP payload formats, profiles, and 156 extensions. This range of add-ons has allowed RTP to meet various 157 needs that were not envisaged by the original protocol designers, and 158 to support many new media encodings, but raises the question of what 159 extensions are to be supported by new implementations. The 160 development of the WebRTC framework provides an opportunity to review 161 the available RTP features and extensions, and to define a common 162 baseline RTP feature set for all WebRTC Endpoints. This builds on 163 the past 20 years development of RTP to mandate the use of extensions 164 that have shown widespread utility, while still remaining compatible 165 with the wide installed base of RTP implementations where possible. 167 RTP and RTCP extensions that are not discussed in this document can 168 be implemented by WebRTC Endpoints if they are beneficial for new use 169 cases. However, they are not necessary to address the WebRTC use 170 cases and requirements identified in [RFC7478]. 172 While the baseline set of RTP features and extensions defined in this 173 memo is targeted at the requirements of the WebRTC framework, it is 174 expected to be broadly useful for other conferencing-related uses of 175 RTP. In particular, it is likely that this set of RTP features and 176 extensions will be appropriate for other desktop or mobile video 177 conferencing systems, or for room-based high-quality telepresence 178 applications. 180 3. Terminology 182 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 183 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 184 document are to be interpreted as described in [RFC2119]. The RFC 185 2119 interpretation of these key words applies only when written in 186 ALL CAPS. Lower- or mixed-case uses of these key words are not to be 187 interpreted as carrying special significance in this memo. 189 We define the following additional terms: 191 WebRTC MediaStream: The MediaStream concept defined by the W3C in 192 the WebRTC API [W3C.WD-mediacapture-streams-20130903]. 194 Transport-layer Flow: A uni-directional flow of transport packets 195 that are identified by having a particular 5-tuple of source IP 196 address, source port, destination IP address, destination port, 197 and transport protocol used. 199 Bi-directional Transport-layer Flow: A bi-directional transport- 200 layer flow is a transport-layer flow that is symmetric. That is, 201 the transport-layer flow in the reverse direction has a 5-tuple 202 where the source and destination address and ports are swapped 203 compared to the forward path transport-layer flow, and the 204 transport protocol is the same. 206 This document uses the terminology from 207 [I-D.ietf-avtext-rtp-grouping-taxonomy] and 208 [I-D.ietf-rtcweb-overview]. Other terms are used according to their 209 definitions from the RTP Specification [RFC3550]. Especially note 210 the following frequently used terms: RTP Packet Stream, RTP Session, 211 and End-point. 213 4. WebRTC Use of RTP: Core Protocols 215 The following sections describe the core features of RTP and RTCP 216 that need to be implemented, along with the mandated RTP profiles. 217 Also described are the core extensions providing essential features 218 that all WebRTC Endpoints need to implement to function effectively 219 on today's networks. 221 4.1. RTP and RTCP 223 The Real-time Transport Protocol (RTP) [RFC3550] is REQUIRED to be 224 implemented as the media transport protocol for WebRTC. RTP itself 225 comprises two parts: the RTP data transfer protocol, and the RTP 226 control protocol (RTCP). RTCP is a fundamental and integral part of 227 RTP, and MUST be implemented and used in all WebRTC Endpoints. 229 The following RTP and RTCP features are sometimes omitted in limited 230 functionality implementations of RTP, but are REQUIRED in all WebRTC 231 Endpoints: 233 o Support for use of multiple simultaneous SSRC values in a single 234 RTP session, including support for RTP end-points that send many 235 SSRC values simultaneously, following [RFC3550] and 236 [I-D.ietf-avtcore-rtp-multi-stream]. The RTCP optimisations for 237 multi-SSRC sessions defined in 238 [I-D.ietf-avtcore-rtp-multi-stream-optimisation] MAY be supported; 239 if supported the usage MUST be signalled. 241 o Random choice of SSRC on joining a session; collision detection 242 and resolution for SSRC values (see also Section 4.8). 244 o Support for reception of RTP data packets containing CSRC lists, 245 as generated by RTP mixers, and RTCP packets relating to CSRCs. 247 o Sending correct synchronisation information in the RTCP Sender 248 Reports, to allow receivers to implement lip-synchronisation; see 249 Section 5.2.1 regarding support for the rapid RTP synchronisation 250 extensions. 252 o Support for multiple synchronisation contexts. Participants that 253 send multiple simultaneous RTP packet streams SHOULD do so as part 254 of a single synchronisation context, using a single RTCP CNAME for 255 all streams and allowing receivers to play the streams out in a 256 synchronised manner. For compatibility with potential future 257 versions of this specification, or for interoperability with non- 258 WebRTC devices through a gateway, receivers MUST support multiple 259 synchronisation contexts, indicated by the use of multiple RTCP 260 CNAMEs in an RTP session. This specification requires the usage 261 of a single CNAME when sending RTP Packet Streams in some 262 circumstances, see Section 4.9. 264 o Support for sending and receiving RTCP SR, RR, SDES, and BYE 265 packet types, with OPTIONAL support for other RTCP packet types 266 unless mandated by other parts of this specification. Note that 267 additional RTCP Packet types are used by the RTP/SAVPF Profile 268 (Section 4.2) and the other RTCP extensions (Section 5). WebRTC 269 endpoints that implement the SDP bundle negotiation extension will 270 use the SDP grouping framework 'mid' attribute to identify media 271 streams. Such endpoints MUST implement the RTCP SDES MID item 272 described in [I-D.ietf-mmusic-sdp-bundle-negotiation]. 274 o Support for multiple end-points in a single RTP session, and for 275 scaling the RTCP transmission interval according to the number of 276 participants in the session; support for randomised RTCP 277 transmission intervals to avoid synchronisation of RTCP reports; 278 support for RTCP timer reconsideration (Section 6.3.6 of 279 [RFC3550]) and reverse reconsideration (Section 6.3.4 of 280 [RFC3550]). 282 o Support for configuring the RTCP bandwidth as a fraction of the 283 media bandwidth, and for configuring the fraction of the RTCP 284 bandwidth allocated to senders, e.g., using the SDP "b=" line 285 [RFC4566][RFC3556]. 287 o Support for the reduced minimum RTCP reporting interval described 288 in Section 6.2 of [RFC3550]. When using the reduced minimum RTCP 289 reporting interval, the fixed (non-reduced) minimum interval MUST 290 be used when calculating the participant timeout interval (see 291 Sections 6.2 and 6.3.5 of [RFC3550]). The delay before sending 292 the initial compound RTCP packet can be set to zero (see 293 Section 6.2 of [RFC3550] as updated by 294 [I-D.ietf-avtcore-rtp-multi-stream]). 296 o Support for discontinuous transmission. RTP allows endpoints to 297 pause and resume transmission at any time. When resuming, the RTP 298 sequence number will increase by one, as usual, while the increase 299 in the RTP timestamp value will depend on the duration of the 300 pause. Discontinuous transmission is most commonly used with some 301 audio payload formats, but is not audio specific, and can be used 302 with any RTP payload format. 304 o Ignore unknown RTCP packet types and RTP header extensions. This 305 to ensure robust handling of future extensions, middlebox 306 behaviours, etc., that can result in not signalled RTCP packet 307 types or RTP header extensions being received. If a compound RTCP 308 packet is received that contains a mixture of known and unknown 309 RTCP packet types, the known packets types need to be processed as 310 usual, with only the unknown packet types being discarded. 312 It is known that a significant number of legacy RTP implementations, 313 especially those targeted at VoIP-only systems, do not support all of 314 the above features, and in some cases do not support RTCP at all. 315 Implementers are advised to consider the requirements for graceful 316 degradation when interoperating with legacy implementations. 318 Other implementation considerations are discussed in Section 12. 320 4.2. Choice of the RTP Profile 322 The complete specification of RTP for a particular application domain 323 requires the choice of an RTP Profile. For WebRTC use, the Extended 324 Secure RTP Profile for RTCP-Based Feedback (RTP/SAVPF) [RFC5124], as 325 extended by [RFC7007], MUST be implemented. The RTP/SAVPF profile is 326 the combination of basic RTP/AVP profile [RFC3551], the RTP profile 327 for RTCP-based feedback (RTP/AVPF) [RFC4585], and the secure RTP 328 profile (RTP/SAVP) [RFC3711]. 330 The RTCP-based feedback extensions [RFC4585] are needed for the 331 improved RTCP timer model. This allows more flexible transmission of 332 RTCP packets in response to events, rather than strictly according to 333 bandwidth, and is vital for being able to report congestion signals 334 as well as media events. These extensions also allow saving RTCP 335 bandwidth, and an end-point will commonly only use the full RTCP 336 bandwidth allocation if there are many events that require feedback. 338 The timer rules are also needed to make use of the RTP conferencing 339 extensions discussed in Section 5.1. 341 Note: The enhanced RTCP timer model defined in the RTP/AVPF 342 profile is backwards compatible with legacy systems that implement 343 only the RTP/AVP or RTP/SAVP profile, given some constraints on 344 parameter configuration such as the RTCP bandwidth value and "trr- 345 int" (the most important factor for interworking with RTP/(S)AVP 346 end-points via a gateway is to set the trr-int parameter to a 347 value representing 4 seconds, see Section 6.1 in 348 [I-D.ietf-avtcore-rtp-multi-stream]). 350 The secure RTP (SRTP) profile extensions [RFC3711] are needed to 351 provide media encryption, integrity protection, replay protection and 352 a limited form of source authentication. WebRTC Endpoints MUST NOT 353 send packets using the basic RTP/AVP profile or the RTP/AVPF profile; 354 they MUST employ the full RTP/SAVPF profile to protect all RTP and 355 RTCP packets that are generated (i.e., implementations MUST use SRTP 356 and SRTCP). The RTP/SAVPF profile MUST be configured using the 357 cipher suites, DTLS-SRTP protection profiles, keying mechanisms, and 358 other parameters described in [I-D.ietf-rtcweb-security-arch]. 360 4.3. Choice of RTP Payload Formats 362 Mandatory to implement audio codecs and RTP payload formats for 363 WebRTC endpoints are defined in [I-D.ietf-rtcweb-audio]. Mandatory 364 to implement video codecs and RTP payload formats for WebRTC 365 endpoints are defined in [I-D.ietf-rtcweb-video]. WebRTC endpoints 366 MAY additionally implement any other codec for which an RTP payload 367 format and associated signalling has been defined. 369 WebRTC Endpoints cannot assume that the other participants in an RTP 370 session understand any RTP payload format, no matter how common. The 371 mapping between RTP payload type numbers and specific configurations 372 of particular RTP payload formats MUST be agreed before those payload 373 types/formats can be used. In an SDP context, this can be done using 374 the "a=rtpmap:" and "a=fmtp:" attributes associated with an "m=" 375 line, along with any other SDP attributes needed to configure the RTP 376 payload format. 378 End-points can signal support for multiple RTP payload formats, or 379 multiple configurations of a single RTP payload format, as long as 380 each unique RTP payload format configuration uses a different RTP 381 payload type number. As outlined in Section 4.8, the RTP payload 382 type number is sometimes used to associate an RTP packet stream with 383 a signalling context. This association is possible provided unique 384 RTP payload type numbers are used in each context. For example, an 385 RTP packet stream can be associated with an SDP "m=" line by 386 comparing the RTP payload type numbers used by the RTP packet stream 387 with payload types signalled in the "a=rtpmap:" lines in the media 388 sections of the SDP. This leads to the following considerations: 390 If RTP packet streams are being associated with signalling 391 contexts based on the RTP payload type, then the assignment of RTP 392 payload type numbers MUST be unique across signalling contexts. 394 If the same RTP payload format configuration is used in multiple 395 contexts, then a different RTP payload type number has to be 396 assigned in each context to ensure uniqueness. 398 If the RTP payload type number is not being used to associate RTP 399 packet streams with a signalling context, then the same RTP 400 payload type number can be used to indicate the exact same RTP 401 payload format configuration in multiple contexts. 403 A single RTP payload type number MUST NOT be assigned to different 404 RTP payload formats, or different configurations of the same RTP 405 payload format, within a single RTP session (note that the "m=" lines 406 in an SDP bundle group [I-D.ietf-mmusic-sdp-bundle-negotiation] form 407 a single RTP session). 409 An end-point that has signalled support for multiple RTP payload 410 formats MUST be able to accept data in any of those payload formats 411 at any time, unless it has previously signalled limitations on its 412 decoding capability. This requirement is constrained if several 413 types of media (e.g., audio and video) are sent in the same RTP 414 session. In such a case, a source (SSRC) is restricted to switching 415 only between the RTP payload formats signalled for the type of media 416 that is being sent by that source; see Section 4.4. To support rapid 417 rate adaptation by changing codec, RTP does not require advance 418 signalling for changes between RTP payload formats used by a single 419 SSRC that were signalled during session set-up. 421 If performing changes between two RTP payload types that use 422 different RTP clock rates, an RTP sender MUST follow the 423 recommendations in Section 4.1 of [RFC7160]. RTP receivers MUST 424 follow the recommendations in Section 4.3 of [RFC7160] in order to 425 support sources that switch between clock rates in an RTP session 426 (these recommendations for receivers are backwards compatible with 427 the case where senders use only a single clock rate). 429 4.4. Use of RTP Sessions 431 An association amongst a set of end-points communicating using RTP is 432 known as an RTP session [RFC3550]. An end-point can be involved in 433 several RTP sessions at the same time. In a multimedia session, each 434 type of media has typically been carried in a separate RTP session 435 (e.g., using one RTP session for the audio, and a separate RTP 436 session using a different transport-layer flow for the video). 437 WebRTC Endpoints are REQUIRED to implement support for multimedia 438 sessions in this way, separating each RTP session using different 439 transport-layer flows for compatibility with legacy systems (this is 440 sometimes called session multiplexing). 442 In modern day networks, however, with the widespread use of network 443 address/port translators (NAT/NAPT) and firewalls, it is desirable to 444 reduce the number of transport-layer flows used by RTP applications. 445 This can be done by sending all the RTP packet streams in a single 446 RTP session, which will comprise a single transport-layer flow (this 447 will prevent the use of some quality-of-service mechanisms, as 448 discussed in Section 12.1.3). Implementations are therefore also 449 REQUIRED to support transport of all RTP packet streams, independent 450 of media type, in a single RTP session using a single transport layer 451 flow, according to [I-D.ietf-avtcore-multi-media-rtp-session] (this 452 is sometimes called SSRC multiplexing). If multiple types of media 453 are to be used in a single RTP session, all participants in that RTP 454 session MUST agree to this usage. In an SDP context, 455 [I-D.ietf-mmusic-sdp-bundle-negotiation] can be used to signal such a 456 bundle of RTP packet streams forming a single RTP session. 458 Further discussion about the suitability of different RTP session 459 structures and multiplexing methods to different scenarios can be 460 found in [I-D.ietf-avtcore-multiplex-guidelines]. 462 4.5. RTP and RTCP Multiplexing 464 Historically, RTP and RTCP have been run on separate transport layer 465 flows (e.g., two UDP ports for each RTP session, one port for RTP and 466 one port for RTCP). With the increased use of Network Address/Port 467 Translation (NAT/NAPT) this has become problematic, since maintaining 468 multiple NAT bindings can be costly. It also complicates firewall 469 administration, since multiple ports need to be opened to allow RTP 470 traffic. To reduce these costs and session set-up times, 471 implementations are REQUIRED to support multiplexing RTP data packets 472 and RTCP control packets on a single transport-layer flow [RFC5761]. 473 Such RTP and RTCP multiplexing MUST be negotiated in the signalling 474 channel before it is used. If SDP is used for signalling, this 475 negotiation MUST use the attributes defined in [RFC5761]. For 476 backwards compatibility, implementations are also REQUIRED to support 477 RTP and RTCP sent on separate transport-layer flows. 479 Note that the use of RTP and RTCP multiplexed onto a single 480 transport-layer flow ensures that there is occasional traffic sent on 481 that port, even if there is no active media traffic. This can be 482 useful to keep NAT bindings alive [RFC6263]. 484 4.6. Reduced Size RTCP 486 RTCP packets are usually sent as compound RTCP packets, and [RFC3550] 487 requires that those compound packets start with an Sender Report (SR) 488 or Receiver Report (RR) packet. When using frequent RTCP feedback 489 messages under the RTP/AVPF Profile [RFC4585] these statistics are 490 not needed in every packet, and unnecessarily increase the mean RTCP 491 packet size. This can limit the frequency at which RTCP packets can 492 be sent within the RTCP bandwidth share. 494 To avoid this problem, [RFC5506] specifies how to reduce the mean 495 RTCP message size and allow for more frequent feedback. Frequent 496 feedback, in turn, is essential to make real-time applications 497 quickly aware of changing network conditions, and to allow them to 498 adapt their transmission and encoding behaviour. Implementations 499 MUST support sending and receiving non-compound RTCP feedback packets 500 [RFC5506]. Use of non-compound RTCP packets MUST be negotiated using 501 the signalling channel. If SDP is used for signalling, this 502 negotiation MUST use the attributes defined in [RFC5506]. For 503 backwards compatibility, implementations are also REQUIRED to support 504 the use of compound RTCP feedback packets if the remote end-point 505 does not agree to the use of non-compound RTCP in the signalling 506 exchange. 508 4.7. Symmetric RTP/RTCP 510 To ease traversal of NAT and firewall devices, implementations are 511 REQUIRED to implement and use Symmetric RTP [RFC4961]. The reason 512 for using symmetric RTP is primarily to avoid issues with NATs and 513 Firewalls by ensuring that the send and receive RTP packet streams, 514 as well as RTCP, are actually bi-directional transport-layer flows. 515 This will keep alive the NAT and firewall pinholes, and help indicate 516 consent that the receive direction is a transport-layer flow the 517 intended recipient actually wants. In addition, it saves resources, 518 specifically ports at the end-points, but also in the network as NAT 519 mappings or firewall state is not unnecessary bloated. The amount of 520 per flow QoS state kept in the network is also reduced. 522 4.8. Choice of RTP Synchronisation Source (SSRC) 524 Implementations are REQUIRED to support signalled RTP synchronisation 525 source (SSRC) identifiers. If SDP is used, this MUST be done using 526 the "a=ssrc:" SDP attribute defined in Section 4.1 and Section 5 of 527 [RFC5576] and the "previous-ssrc" source attribute defined in 528 Section 6.2 of [RFC5576]; other per-SSRC attributes defined in 529 [RFC5576] MAY be supported. 531 While support for signalled SSRC identifiers is mandated, their use 532 in an RTP session is OPTIONAL. Implementations MUST be prepared to 533 accept RTP and RTCP packets using SSRCs that have not been explicitly 534 signalled ahead of time. Implementations MUST support random SSRC 535 assignment, and MUST support SSRC collision detection and resolution, 536 according to [RFC3550]. When using signalled SSRC values, collision 537 detection MUST be performed as described in Section 5 of [RFC5576]. 539 It is often desirable to associate an RTP packet stream with a non- 540 RTP context. For users of the WebRTC API a mapping between SSRCs and 541 MediaStreamTracks are provided per Section 11. For gateways or other 542 usages it is possible to associate an RTP packet stream with an "m=" 543 line in a session description formatted using SDP. If SSRCs are 544 signalled this is straightforward (in SDP the "a=ssrc:" line will be 545 at the media level, allowing a direct association with an "m=" line). 546 If SSRCs are not signalled, the RTP payload type numbers used in an 547 RTP packet stream are often sufficient to associate that packet 548 stream with a signalling context (e.g., if RTP payload type numbers 549 are assigned as described in Section 4.3 of this memo, the RTP 550 payload types used by an RTP packet stream can be compared with 551 values in SDP "a=rtpmap:" lines, which are at the media level in SDP, 552 and so map to an "m=" line). 554 4.9. Generation of the RTCP Canonical Name (CNAME) 556 The RTCP Canonical Name (CNAME) provides a persistent transport-level 557 identifier for an RTP end-point. While the Synchronisation Source 558 (SSRC) identifier for an RTP end-point can change if a collision is 559 detected, or when the RTP application is restarted, its RTCP CNAME is 560 meant to stay unchanged for the duration of a RTCPeerConnection 561 [W3C.WD-webrtc-20130910], so that RTP end-points can be uniquely 562 identified and associated with their RTP packet streams within a set 563 of related RTP sessions. 565 Each RTP end-point MUST have at least one RTCP CNAME, and that RTCP 566 CNAME MUST be unique within the RTCPeerConnection. RTCP CNAMEs 567 identify a particular synchronisation context, i.e., all SSRCs 568 associated with a single RTCP CNAME share a common reference clock. 569 If an end-point has SSRCs that are associated with several 570 unsynchronised reference clocks, and hence different synchronisation 571 contexts, it will need to use multiple RTCP CNAMEs, one for each 572 synchronisation context. 574 Taking the discussion in Section 11 into account, a WebRTC Endpoint 575 MUST NOT use more than one RTCP CNAME in the RTP sessions belonging 576 to single RTCPeerConnection (that is, an RTCPeerConnection forms a 577 synchronisation context). RTP middleboxes MAY generate RTP packet 578 streams associated with more than one RTCP CNAME, to allow them to 579 avoid having to resynchronize media from multiple different end- 580 points part of a multi-party RTP session. 582 The RTP specification [RFC3550] includes guidelines for choosing a 583 unique RTP CNAME, but these are not sufficient in the presence of NAT 584 devices. In addition, long-term persistent identifiers can be 585 problematic from a privacy viewpoint (Section 13). Accordingly, a 586 WebRTC Endpoint MUST generate a new, unique, short-term persistent 587 RTCP CNAME for each RTCPeerConnection, following [RFC7022], with a 588 single exception; if explicitly requested at creation an 589 RTCPeerConnection MAY use the same CNAME as as an existing 590 RTCPeerConnection within their common same-origin context. 592 An WebRTC Endpoint MUST support reception of any CNAME that matches 593 the syntax limitations specified by the RTP specification [RFC3550] 594 and cannot assume that any CNAME will be chosen according to the form 595 suggested above. 597 4.10. Handling of Leap Seconds 599 The guidelines regarding handling of leap seconds to limit their 600 impact on RTP media play-out and synchronization given in [RFC7164] 601 SHOULD be followed. 603 5. WebRTC Use of RTP: Extensions 605 There are a number of RTP extensions that are either needed to obtain 606 full functionality, or extremely useful to improve on the baseline 607 performance, in the WebRTC context. One set of these extensions is 608 related to conferencing, while others are more generic in nature. 609 The following subsections describe the various RTP extensions 610 mandated or suggested for use within WebRTC. 612 5.1. Conferencing Extensions and Topologies 614 RTP is a protocol that inherently supports group communication. 615 Groups can be implemented by having each endpoint send its RTP packet 616 streams to an RTP middlebox that redistributes the traffic, by using 617 a mesh of unicast RTP packet streams between endpoints, or by using 618 an IP multicast group to distribute the RTP packet streams. These 619 topologies can be implemented in a number of ways as discussed in 620 [I-D.ietf-avtcore-rtp-topologies-update]. 622 While the use of IP multicast groups is popular in IPTV systems, the 623 topologies based on RTP middleboxes are dominant in interactive video 624 conferencing environments. Topologies based on a mesh of unicast 625 transport-layer flows to create a common RTP session have not seen 626 widespread deployment to date. Accordingly, WebRTC Endpoints are not 627 expected to support topologies based on IP multicast groups or to 628 support mesh-based topologies, such as a point-to-multipoint mesh 629 configured as a single RTP session (Topo-Mesh in the terminology of 630 [I-D.ietf-avtcore-rtp-topologies-update]). However, a point-to- 631 multipoint mesh constructed using several RTP sessions, implemented 632 in WebRTC using independent RTCPeerConnections 633 [W3C.WD-webrtc-20130910], can be expected to be used in WebRTC, and 634 needs to be supported. 636 WebRTC Endpoints implemented according to this memo are expected to 637 support all the topologies described in 638 [I-D.ietf-avtcore-rtp-topologies-update] where the RTP endpoints send 639 and receive unicast RTP packet streams to and from some peer device, 640 provided that peer can participate in performing congestion control 641 on the RTP packet streams. The peer device could be another RTP 642 endpoint, or it could be an RTP middlebox that redistributes the RTP 643 packet streams to other RTP endpoints. This limitation means that 644 some of the RTP middlebox-based topologies are not suitable for use 645 in WebRTC. Specifically: 647 o Video switching MCUs (Topo-Video-switch-MCU) SHOULD NOT be used, 648 since they make the use of RTCP for congestion control and quality 649 of service reports problematic (see Section 3.8 of 650 [I-D.ietf-avtcore-rtp-topologies-update]). 652 o The Relay-Transport Translator (Topo-PtM-Trn-Translator) topology 653 SHOULD NOT be used because its safe use requires a congestion 654 control algorithm or RTP circuit breaker that handles point to 655 multipoint, which has not yet been standardised. 657 The following topology can be used, however it has some issues worth 658 noting: 660 o Content modifying MCUs with RTCP termination (Topo-RTCP- 661 terminating-MCU) MAY be used. Note that in this RTP Topology, RTP 662 loop detection and identification of active senders is the 663 responsibility of the WebRTC application; since the clients are 664 isolated from each other at the RTP layer, RTP cannot assist with 665 these functions (see section 3.9 of 666 [I-D.ietf-avtcore-rtp-topologies-update]). 668 The RTP extensions described in Section 5.1.1 to Section 5.1.6 are 669 designed to be used with centralised conferencing, where an RTP 670 middlebox (e.g., a conference bridge) receives a participant's RTP 671 packet streams and distributes them to the other participants. These 672 extensions are not necessary for interoperability; an RTP end-point 673 that does not implement these extensions will work correctly, but 674 might offer poor performance. Support for the listed extensions will 675 greatly improve the quality of experience and, to provide a 676 reasonable baseline quality, some of these extensions are mandatory 677 to be supported by WebRTC Endpoints. 679 The RTCP conferencing extensions are defined in Extended RTP Profile 680 for Real-time Transport Control Protocol (RTCP)-Based Feedback (RTP/ 681 AVPF) [RFC4585] and the memo on Codec Control Messages (CCM) in RTP/ 682 AVPF [RFC5104]; they are fully usable by the Secure variant of this 683 profile (RTP/SAVPF) [RFC5124]. 685 5.1.1. Full Intra Request (FIR) 687 The Full Intra Request message is defined in Sections 3.5.1 and 4.3.1 688 of the Codec Control Messages [RFC5104]. It is used to make the 689 mixer request a new Intra picture from a participant in the session. 690 This is used when switching between sources to ensure that the 691 receivers can decode the video or other predictive media encoding 692 with long prediction chains. WebRTC Endpoints that are sending media 693 MUST understand and react to FIR feedback messages they receive, 694 since this greatly improves the user experience when using 695 centralised mixer-based conferencing. Support for sending FIR 696 messages is OPTIONAL. 698 5.1.2. Picture Loss Indication (PLI) 700 The Picture Loss Indication message is defined in Section 6.3.1 of 701 the RTP/AVPF profile [RFC4585]. It is used by a receiver to tell the 702 sending encoder that it lost the decoder context and would like to 703 have it repaired somehow. This is semantically different from the 704 Full Intra Request above as there could be multiple ways to fulfil 705 the request. WebRTC Endpoints that are sending media MUST understand 706 and react to PLI feedback messages as a loss tolerance mechanism. 707 Receivers MAY send PLI messages. 709 5.1.3. Slice Loss Indication (SLI) 711 The Slice Loss Indication message is defined in Section 6.3.2 of the 712 RTP/AVPF profile [RFC4585]. It is used by a receiver to tell the 713 encoder that it has detected the loss or corruption of one or more 714 consecutive macro blocks, and would like to have these repaired 715 somehow. It is RECOMMENDED that receivers generate SLI feedback 716 messages if slices are lost when using a codec that supports the 717 concept of macro blocks. A sender that receives an SLI feedback 718 message SHOULD attempt to repair the lost slice(s). 720 5.1.4. Reference Picture Selection Indication (RPSI) 722 Reference Picture Selection Indication (RPSI) messages are defined in 723 Section 6.3.3 of the RTP/AVPF profile [RFC4585]. Some video encoding 724 standards allow the use of older reference pictures than the most 725 recent one for predictive coding. If such a codec is in use, and if 726 the encoder has learnt that encoder-decoder synchronisation has been 727 lost, then a known as correct reference picture can be used as a base 728 for future coding. The RPSI message allows this to be signalled. 729 Receivers that detect that encoder-decoder synchronisation has been 730 lost SHOULD generate an RPSI feedback message if codec being used 731 supports reference picture selection. A RTP packet stream sender 732 that receives such an RPSI message SHOULD act on that messages to 733 change the reference picture, if it is possible to do so within the 734 available bandwidth constraints, and with the codec being used. 736 5.1.5. Temporal-Spatial Trade-off Request (TSTR) 738 The temporal-spatial trade-off request and notification are defined 739 in Sections 3.5.2 and 4.3.2 of [RFC5104]. This request can be used 740 to ask the video encoder to change the trade-off it makes between 741 temporal and spatial resolution, for example to prefer high spatial 742 image quality but low frame rate. Support for TSTR requests and 743 notifications is OPTIONAL. 745 5.1.6. Temporary Maximum Media Stream Bit Rate Request (TMMBR) 747 The TMMBR feedback message is defined in Sections 3.5.4 and 4.2.1 of 748 the Codec Control Messages [RFC5104]. This request and its 749 notification message are used by a media receiver to inform the 750 sending party that there is a current limitation on the amount of 751 bandwidth available to this receiver. There can be various reasons 752 for this: for example, an RTP mixer can use this message to limit the 753 media rate of the sender being forwarded by the mixer (without doing 754 media transcoding) to fit the bottlenecks existing towards the other 755 session participants. WebRTC Endpoints that are sending media are 756 REQUIRED to implement support for TMMBR messages, and MUST follow 757 bandwidth limitations set by a TMMBR message received for their SSRC. 758 The sending of TMMBR requests is OPTIONAL. 760 5.2. Header Extensions 762 The RTP specification [RFC3550] provides the capability to include 763 RTP header extensions containing in-band data, but the format and 764 semantics of the extensions are poorly specified. The use of header 765 extensions is OPTIONAL in WebRTC, but if they are used, they MUST be 766 formatted and signalled following the general mechanism for RTP 767 header extensions defined in [RFC5285], since this gives well-defined 768 semantics to RTP header extensions. 770 As noted in [RFC5285], the requirement from the RTP specification 771 that header extensions are "designed so that the header extension may 772 be ignored" [RFC3550] stands. To be specific, header extensions MUST 773 only be used for data that can safely be ignored by the recipient 774 without affecting interoperability, and MUST NOT be used when the 775 presence of the extension has changed the form or nature of the rest 776 of the packet in a way that is not compatible with the way the stream 777 is signalled (e.g., as defined by the payload type). Valid examples 778 of RTP header extensions might include metadata that is additional to 779 the usual RTP information, but that can safely be ignored without 780 compromising interoperability. 782 5.2.1. Rapid Synchronisation 784 Many RTP sessions require synchronisation between audio, video, and 785 other content. This synchronisation is performed by receivers, using 786 information contained in RTCP SR packets, as described in the RTP 787 specification [RFC3550]. This basic mechanism can be slow, however, 788 so it is RECOMMENDED that the rapid RTP synchronisation extensions 789 described in [RFC6051] be implemented in addition to RTCP SR-based 790 synchronisation. 792 This header extension uses the [RFC5285] generic header extension 793 framework, and so needs to be negotiated before it can be used. 795 5.2.2. Client-to-Mixer Audio Level 797 The Client to Mixer Audio Level extension [RFC6464] is an RTP header 798 extension used by an endpoint to inform a mixer about the level of 799 audio activity in the packet to which the header is attached. This 800 enables an RTP middlebox to make mixing or selection decisions 801 without decoding or detailed inspection of the payload, reducing the 802 complexity in some types of mixers. It can also save decoding 803 resources in receivers, which can choose to decode only the most 804 relevant RTP packet streams based on audio activity levels. 806 The Client-to-Mixer Audio Level [RFC6464] header extension MUST be 807 implemented. It is REQUIRED that implementations are capable of 808 encrypting the header extension according to [RFC6904] since the 809 information contained in these header extensions can be considered 810 sensitive. The use of this encryption is RECOMMENDED, however usage 811 of the encryption can be explicitly disabled through API or 812 signalling. 814 This header extension uses the [RFC5285] generic header extension 815 framework, and so needs to be negotiated before it can be used. 817 5.2.3. Mixer-to-Client Audio Level 819 The Mixer to Client Audio Level header extension [RFC6465] provides 820 an endpoint with the audio level of the different sources mixed into 821 a common source stream by a RTP mixer. This enables a user interface 822 to indicate the relative activity level of each session participant, 823 rather than just being included or not based on the CSRC field. This 824 is a pure optimisation of non critical functions, and is hence 825 OPTIONAL to implement. If this header extension is implemented, it 826 is REQUIRED that implementations are capable of encrypting the header 827 extension according to [RFC6904] since the information contained in 828 these header extensions can be considered sensitive. It is further 829 RECOMMENDED that this encryption is used, unless the encryption has 830 been explicitly disabled through API or signalling. 832 This header extension uses the [RFC5285] generic header extension 833 framework, and so needs to be negotiated before it can be used. 835 5.2.4. Media Stream Identification 837 WebRTC endpoints that implement the SDP bundle negotiation extension 838 will use the SDP grouping framework 'mid' attribute to identify media 839 streams. Such endpoints MUST implement the RTP MID header extension 840 described in [I-D.ietf-mmusic-sdp-bundle-negotiation]. 842 This header extension uses the [RFC5285] generic header extension 843 framework, and so needs to be negotiated before it can be used. 845 5.2.5. Coordination of Video Orientation 847 WebRTC endpoints that send or receive video MUST implement the 848 coordination of video orientation (CVO) RTP header extension as 849 described in Section 4 of [I-D.ietf-rtcweb-video]. 851 This header extension uses the [RFC5285] generic header extension 852 framework, and so needs to be negotiated before it can be used. 854 6. WebRTC Use of RTP: Improving Transport Robustness 856 There are tools that can make RTP packet streams robust against 857 packet loss and reduce the impact of loss on media quality. However, 858 they generally add some overhead compared to a non-robust stream. 859 The overhead needs to be considered, and the aggregate bit-rate MUST 860 be rate controlled to avoid causing network congestion (see 861 Section 7). As a result, improving robustness might require a lower 862 base encoding quality, but has the potential to deliver that quality 863 with fewer errors. The mechanisms described in the following sub- 864 sections can be used to improve tolerance to packet loss. 866 6.1. Negative Acknowledgements and RTP Retransmission 868 As a consequence of supporting the RTP/SAVPF profile, implementations 869 can send negative acknowledgements (NACKs) for RTP data packets 870 [RFC4585]. This feedback can be used to inform a sender of the loss 871 of particular RTP packets, subject to the capacity limitations of the 872 RTCP feedback channel. A sender can use this information to optimise 873 the user experience by adapting the media encoding to compensate for 874 known lost packets. 876 RTP packet stream senders are REQUIRED to understand the Generic NACK 877 message defined in Section 6.2.1 of [RFC4585], but MAY choose to 878 ignore some or all of this feedback (following Section 4.2 of 879 [RFC4585]). Receivers MAY send NACKs for missing RTP packets. 880 Guidelines on when to send NACKs are provided in [RFC4585]. It is 881 not expected that a receiver will send a NACK for every lost RTP 882 packet, rather it needs to consider the cost of sending NACK 883 feedback, and the importance of the lost packet, to make an informed 884 decision on whether it is worth telling the sender about a packet 885 loss event. 887 The RTP Retransmission Payload Format [RFC4588] offers the ability to 888 retransmit lost packets based on NACK feedback. Retransmission needs 889 to be used with care in interactive real-time applications to ensure 890 that the retransmitted packet arrives in time to be useful, but can 891 be effective in environments with relatively low network RTT (an RTP 892 sender can estimate the RTT to the receivers using the information in 893 RTCP SR and RR packets, as described at the end of Section 6.4.1 of 894 [RFC3550]). The use of retransmissions can also increase the forward 895 RTP bandwidth, and can potentially caused increased packet loss if 896 the original packet loss was caused by network congestion. Note, 897 however, that retransmission of an important lost packet to repair 898 decoder state can have lower cost than sending a full intra frame. 899 It is not appropriate to blindly retransmit RTP packets in response 900 to a NACK. The importance of lost packets and the likelihood of them 901 arriving in time to be useful needs to be considered before RTP 902 retransmission is used. 904 Receivers are REQUIRED to implement support for RTP retransmission 905 packets [RFC4588] sent using SSRC multiplexing, and MAY also support 906 RTP retransmission packets sent using session multiplexing. Senders 907 MAY send RTP retransmission packets in response to NACKs if support 908 for the RTP retransmission payload format has been negotiated, and if 909 the sender believes it is useful to send a retransmission of the 910 packet(s) referenced in the NACK. Senders do not need to retransmit 911 every NACKed packet. 913 6.2. Forward Error Correction (FEC) 915 The use of Forward Error Correction (FEC) can provide an effective 916 protection against some degree of packet loss, at the cost of steady 917 bandwidth overhead. There are several FEC schemes that are defined 918 for use with RTP. Some of these schemes are specific to a particular 919 RTP payload format, others operate across RTP packets and can be used 920 with any payload format. It needs to be noted that using redundant 921 encoding or FEC will lead to increased play out delay, which needs to 922 be considered when choosing FEC schemes and their parameters. 924 WebRTC endpoints MUST follow the recommendations for FEC use given in 925 [I-D.ietf-rtcweb-fec]. WebRTC endpoints MAY support other types of 926 FEC, but these MUST be negotiated before they are used. 928 7. WebRTC Use of RTP: Rate Control and Media Adaptation 930 WebRTC will be used in heterogeneous network environments using a 931 variety set of link technologies, including both wired and wireless 932 links, to interconnect potentially large groups of users around the 933 world. As a result, the network paths between users can have widely 934 varying one-way delays, available bit-rates, load levels, and traffic 935 mixtures. Individual end-points can send one or more RTP packet 936 streams to each participant, and there can be several participants. 937 Each of these RTP packet streams can contain different types of 938 media, and the type of media, bit rate, and number of RTP packet 939 streams as well as transport-layer flows can be highly asymmetric. 940 Non-RTP traffic can share the network paths with RTP transport-layer 941 flows. Since the network environment is not predictable or stable, 942 WebRTC Endpoints MUST ensure that the RTP traffic they generate can 943 adapt to match changes in the available network capacity. 945 The quality of experience for users of WebRTC is very dependent on 946 effective adaptation of the media to the limitations of the network. 947 End-points have to be designed so they do not transmit significantly 948 more data than the network path can support, except for very short 949 time periods, otherwise high levels of network packet loss or delay 950 spikes will occur, causing media quality degradation. The limiting 951 factor on the capacity of the network path might be the link 952 bandwidth, or it might be competition with other traffic on the link 953 (this can be non-WebRTC traffic, traffic due to other WebRTC flows, 954 or even competition with other WebRTC flows in the same session). 956 An effective media congestion control algorithm is therefore an 957 essential part of the WebRTC framework. However, at the time of this 958 writing, there is no standard congestion control algorithm that can 959 be used for interactive media applications such as WebRTC's flows. 960 Some requirements for congestion control algorithms for 961 RTCPeerConnections are discussed in [I-D.ietf-rmcat-cc-requirements]. 962 If a standardized congestion control algorithm that satisfies these 963 requirements is developed in the future, this memo will need to be be 964 updated to mandate its use. 966 7.1. Boundary Conditions and Circuit Breakers 968 WebRTC Endpoints MUST implement the RTP circuit breaker algorithm 969 that is described in [I-D.ietf-avtcore-rtp-circuit-breakers]. The 970 RTP circuit breaker is designed to enable applications to recognise 971 and react to situations of extreme network congestion. However, 972 since the RTP circuit breaker might not be triggered until congestion 973 becomes extreme, it cannot be considered a substitute for congestion 974 control, and applications MUST also implement congestion control to 975 allow them to adapt to changes in network capacity. Any future RTP 976 congestion control algorithms are expected to operate within the 977 envelope allowed by the circuit breaker. 979 The session establishment signalling will also necessarily establish 980 boundaries to which the media bit-rate will conform. The choice of 981 media codecs provides upper- and lower-bounds on the supported bit- 982 rates that the application can utilise to provide useful quality, and 983 the packetisation choices that exist. In addition, the signalling 984 channel can establish maximum media bit-rate boundaries using, for 985 example, the SDP "b=AS:" or "b=CT:" lines and the RTP/AVPF Temporary 986 Maximum Media Stream Bit Rate (TMMBR) Requests (see Section 5.1.6 of 987 this memo). Signalled bandwidth limitations, such as SDP "b=AS:" or 988 "b=CT:" lines received from the peer, MUST be followed when sending 989 RTP packet streams. A WebRTC Endpoint receiving media SHOULD signal 990 its bandwidth limitations. These limitations have to be based on 991 known bandwidth limitations, for example the capacity of the edge 992 links. 994 7.2. Congestion Control Interoperability and Legacy Systems 996 All endpoints that wish to interwork with WebRTC MUST implement RTCP 997 and provide congestion feedback via the defined RTCP reporting 998 mechanisms. 1000 When interworking with legacy implementations that support RTCP using 1001 the RTP/AVP profile [RFC3551], congestion feedback is provided in 1002 RTCP RR packets every few seconds. Implementations that have to 1003 interwork with such end-points MUST ensure that they keep within the 1004 RTP circuit breaker [I-D.ietf-avtcore-rtp-circuit-breakers] 1005 constraints to limit the congestion they can cause. 1007 If a legacy end-point supports RTP/AVPF, this enables negotiation of 1008 important parameters for frequent reporting, such as the "trr-int" 1009 parameter, and the possibility that the end-point supports some 1010 useful feedback format for congestion control purpose such as TMMBR 1011 [RFC5104]. Implementations that have to interwork with such end- 1012 points MUST ensure that they stay within the RTP circuit breaker 1013 [I-D.ietf-avtcore-rtp-circuit-breakers] constraints to limit the 1014 congestion they can cause, but might find that they can achieve 1015 better congestion response depending on the amount of feedback that 1016 is available. 1018 With proprietary congestion control algorithms issues can arise when 1019 different algorithms and implementations interact in a communication 1020 session. If the different implementations have made different 1021 choices in regards to the type of adaptation, for example one sender 1022 based, and one receiver based, then one could end up in situation 1023 where one direction is dual controlled, when the other direction is 1024 not controlled. This memo cannot mandate behaviour for proprietary 1025 congestion control algorithms, but implementations that use such 1026 algorithms ought to be aware of this issue, and try to ensure that 1027 effective congestion control is negotiated for media flowing in both 1028 directions. If the IETF were to standardise both sender- and 1029 receiver-based congestion control algorithms for WebRTC traffic in 1030 the future, the issues of interoperability, control, and ensuring 1031 that both directions of media flow are congestion controlled would 1032 also need to be considered. 1034 8. WebRTC Use of RTP: Performance Monitoring 1036 As described in Section 4.1, implementations are REQUIRED to generate 1037 RTCP Sender Report (SR) and Reception Report (RR) packets relating to 1038 the RTP packet streams they send and receive. These RTCP reports can 1039 be used for performance monitoring purposes, since they include basic 1040 packet loss and jitter statistics. 1042 A large number of additional performance metrics are supported by the 1043 RTCP Extended Reports (XR) framework [RFC3611][RFC6792]. At the time 1044 of this writing, it is not clear what extended metrics are suitable 1045 for use in WebRTC, so there is no requirement that implementations 1046 generate RTCP XR packets. However, implementations that can use 1047 detailed performance monitoring data MAY generate RTCP XR packets as 1048 appropriate; the use of such packets SHOULD be signalled in advance. 1050 9. WebRTC Use of RTP: Future Extensions 1052 It is possible that the core set of RTP protocols and RTP extensions 1053 specified in this memo will prove insufficient for the future needs 1054 of WebRTC. In this case, future updates to this memo MUST be made 1055 following the Guidelines for Writers of RTP Payload Format 1056 Specifications [RFC2736], How to Write an RTP Payload Format 1057 [I-D.ietf-payload-rtp-howto] and Guidelines for Extending the RTP 1058 Control Protocol [RFC5968], and SHOULD take into account any future 1059 guidelines for extending RTP and related protocols that have been 1060 developed. 1062 Authors of future extensions are urged to consider the wide range of 1063 environments in which RTP is used when recommending extensions, since 1064 extensions that are applicable in some scenarios can be problematic 1065 in others. Where possible, the WebRTC framework will adopt RTP 1066 extensions that are of general utility, to enable easy implementation 1067 of a gateway to other applications using RTP, rather than adopt 1068 mechanisms that are narrowly targeted at specific WebRTC use cases. 1070 10. Signalling Considerations 1072 RTP is built with the assumption that an external signalling channel 1073 exists, and can be used to configure RTP sessions and their features. 1074 The basic configuration of an RTP session consists of the following 1075 parameters: 1077 RTP Profile: The name of the RTP profile to be used in session. The 1078 RTP/AVP [RFC3551] and RTP/AVPF [RFC4585] profiles can interoperate 1079 on basic level, as can their secure variants RTP/SAVP [RFC3711] 1080 and RTP/SAVPF [RFC5124]. The secure variants of the profiles do 1081 not directly interoperate with the non-secure variants, due to the 1082 presence of additional header fields for authentication in SRTP 1083 packets and cryptographic transformation of the payload. WebRTC 1084 requires the use of the RTP/SAVPF profile, and this MUST be 1085 signalled. Interworking functions might transform this into the 1086 RTP/SAVP profile for a legacy use case, by indicating to the 1087 WebRTC Endpoint that the RTP/SAVPF is used and configuring a trr- 1088 int value of 4 seconds. 1090 Transport Information: Source and destination IP address(s) and 1091 ports for RTP and RTCP MUST be signalled for each RTP session. In 1092 WebRTC these transport addresses will be provided by ICE [RFC5245] 1093 that signals candidates and arrives at nominated candidate address 1094 pairs. If RTP and RTCP multiplexing [RFC5761] is to be used, such 1095 that a single port, i.e. transport-layer flow, is used for RTP and 1096 RTCP flows, this MUST be signalled (see Section 4.5). 1098 RTP Payload Types, media formats, and format parameters: The mapping 1099 between media type names (and hence the RTP payload formats to be 1100 used), and the RTP payload type numbers MUST be signalled. Each 1101 media type MAY also have a number of media type parameters that 1102 MUST also be signalled to configure the codec and RTP payload 1103 format (the "a=fmtp:" line from SDP). Section 4.3 of this memo 1104 discusses requirements for uniqueness of payload types. 1106 RTP Extensions: The use of any additional RTP header extensions and 1107 RTCP packet types, including any necessary parameters, MUST be 1108 signalled. This signalling is to ensure that a WebRTC Endpoint's 1109 behaviour, especially when sending, of any extensions is 1110 predictable and consistent. For robustness, and for compatibility 1111 with non-WebRTC systems that might be connected to a WebRTC 1112 session via a gateway, implementations are REQUIRED to ignore 1113 unknown RTCP packets and RTP header extensions (see also 1114 Section 4.1). 1116 RTCP Bandwidth: Support for exchanging RTCP Bandwidth values to the 1117 end-points will be necessary. This SHALL be done as described in 1118 "Session Description Protocol (SDP) Bandwidth Modifiers for RTP 1119 Control Protocol (RTCP) Bandwidth" [RFC3556] if using SDP, or 1120 something semantically equivalent. This also ensures that the 1121 end-points have a common view of the RTCP bandwidth. A common 1122 RTCP bandwidth is important as a too different view of the 1123 bandwidths can lead to failure to interoperate. 1125 These parameters are often expressed in SDP messages conveyed within 1126 an offer/answer exchange. RTP does not depend on SDP or on the 1127 offer/answer model, but does require all the necessary parameters to 1128 be agreed upon, and provided to the RTP implementation. Note that in 1129 WebRTC it will depend on the signalling model and API how these 1130 parameters need to be configured but they will be need to either be 1131 set in the API or explicitly signalled between the peers. 1133 11. WebRTC API Considerations 1135 The WebRTC API [W3C.WD-webrtc-20130910] and the Media Capture and 1136 Streams API [W3C.WD-mediacapture-streams-20130903] defines and uses 1137 the concept of a MediaStream that consists of zero or more 1138 MediaStreamTracks. A MediaStreamTrack is an individual stream of 1139 media from any type of media source like a microphone or a camera, 1140 but also conceptual sources, like a audio mix or a video composition, 1141 are possible. The MediaStreamTracks within a MediaStream need to be 1142 possible to play out synchronised. 1144 A MediaStreamTrack's realisation in RTP in the context of an 1145 RTCPeerConnection consists of a source packet stream identified with 1146 an SSRC within an RTP session part of the RTCPeerConnection. The 1147 MediaStreamTrack can also result in additional packet streams, and 1148 thus SSRCs, in the same RTP session. These can be dependent packet 1149 streams from scalable encoding of the source stream associated with 1150 the MediaStreamTrack, if such a media encoder is used. They can also 1151 be redundancy packet streams, these are created when applying Forward 1152 Error Correction (Section 6.2) or RTP retransmission (Section 6.1) to 1153 the source packet stream. 1155 It is important to note that the same media source can be feeding 1156 multiple MediaStreamTracks. As different sets of constraints or 1157 other parameters can be applied to the MediaStreamTrack, each 1158 MediaStreamTrack instance added to a RTCPeerConnection SHALL result 1159 in an independent source packet stream, with its own set of 1160 associated packet streams, and thus different SSRC(s). It will 1161 depend on applied constraints and parameters if the source stream and 1162 the encoding configuration will be identical between different 1163 MediaStreamTracks sharing the same media source. If the encoding 1164 parameters and constraints are the same, an implementation could 1165 choose to use only one encoded stream to create the different RTP 1166 packet streams. Note that such optimisations would need to take into 1167 account that the constraints for one of the MediaStreamTracks can at 1168 any moment change, meaning that the encoding configurations might no 1169 longer be identical and two different encoder instances would then be 1170 needed. 1172 The same MediaStreamTrack can also be included in multiple 1173 MediaStreams, thus multiple sets of MediaStreams can implicitly need 1174 to use the same synchronisation base. To ensure that this works in 1175 all cases, and does not force an end-point to to disrupt the media by 1176 changing synchronisation base and CNAME during delivery of any 1177 ongoing packet streams, all MediaStreamTracks and their associated 1178 SSRCs originating from the same end-point need to be sent using the 1179 same CNAME within one RTCPeerConnection. This is motivating the 1180 discussion in Section 4.9 to only use a single CNAME. 1182 The requirement on using the same CNAME for all SSRCs that 1183 originate from the same end-point, does not require a middlebox 1184 that forwards traffic from multiple end-points to only use a 1185 single CNAME. 1187 Different CNAMEs normally need to be used for different 1188 RTCPeerConnection instances, as specified in Section 4.9. Having two 1189 communication sessions with the same CNAME could enable tracking of a 1190 user or device across different services (see Section 4.4.1 of 1191 [I-D.ietf-rtcweb-security] for details). A web application can 1192 request that the CNAMEs used in different RTCPeerConnections (within 1193 a same-orign context) be the same, this allows for synchronization of 1194 the endpoint's RTP packet streams across the different 1195 RTCPeerConnections. 1197 Note: this doesn't result in a tracking issue, since the creation 1198 of matching CNAMEs depends on existing tracking within a single 1199 origin. 1201 The above will currently force a WebRTC Endpoint that receives a 1202 MediaStreamTrack on one RTCPeerConnection and adds it as an outgoing 1203 on any RTCPeerConnection to perform resynchronisation of the stream. 1204 Since the sending party needs to change the CNAME to the one it uses, 1205 this implies it has to use a local system clock as timebase for the 1206 synchronisation. Thus, the relative relation between the timebase of 1207 the incoming stream and the system sending out needs to be defined. 1208 This relation also needs monitoring for clock drift and likely 1209 adjustments of the synchronisation. The sending entity is also 1210 responsible for congestion control for its sent streams. In cases of 1211 packet loss the loss of incoming data also needs to be handled. This 1212 leads to the observation that the method that is least likely to 1213 cause issues or interruptions in the outgoing source packet stream is 1214 a model of full decoding, including repair etc., followed by encoding 1215 of the media again into the outgoing packet stream. Optimisations of 1216 this method is clearly possible and implementation specific. 1218 A WebRTC Endpoint MUST support receiving multiple MediaStreamTracks, 1219 where each of different MediaStreamTracks (and their sets of 1220 associated packet streams) uses different CNAMEs. However, 1221 MediaStreamTracks that are received with different CNAMEs have no 1222 defined synchronisation. 1224 Note: The motivation for supporting reception of multiple CNAMEs 1225 is to allow for forward compatibility with any future changes that 1226 enable more efficient stream handling when end-points relay/ 1227 forward streams. It also ensures that end-points can interoperate 1228 with certain types of multi-stream middleboxes or end-points that 1229 are not WebRTC. 1231 The binding between the WebRTC MediaStreams, MediaStreamTracks and 1232 the SSRC is done as specified in "Cross Session Stream Identification 1233 in the Session Description Protocol" [I-D.ietf-mmusic-msid]. This 1234 document [I-D.ietf-mmusic-msid] also defines, in section 4.1, how to 1235 map unknown source packet stream SSRCs to MediaStreamTracks and 1236 MediaStreams. This later is relevant to handle some cases of legacy 1237 interop. Commonly the RTP Payload Type of any incoming packets will 1238 reveal if the packet stream is a source stream or a redundancy or 1239 dependent packet stream. The association to the correct source 1240 packet stream depends on the payload format in use for the packet 1241 stream. 1243 Finally this specification puts a requirement on the WebRTC API to 1244 realize a method for determining the CSRC list (Section 4.1) as well 1245 as the Mixer-to-Client audio levels (Section 5.2.3) (when supported) 1246 and the basic requirements for this is further discussed in 1247 Section 12.2.1. 1249 12. RTP Implementation Considerations 1251 The following discussion provides some guidance on the implementation 1252 of the RTP features described in this memo. The focus is on a WebRTC 1253 Endpoint implementation perspective, and while some mention is made 1254 of the behaviour of middleboxes, that is not the focus of this memo. 1256 12.1. Configuration and Use of RTP Sessions 1258 A WebRTC Endpoint will be a simultaneous participant in one or more 1259 RTP sessions. Each RTP session can convey multiple media sources, 1260 and can include media data from multiple end-points. In the 1261 following, some ways in which WebRTC Endpoints can configure and use 1262 RTP sessions is outlined. 1264 12.1.1. Use of Multiple Media Sources Within an RTP Session 1266 RTP is a group communication protocol, and every RTP session can 1267 potentially contain multiple RTP packet streams. There are several 1268 reasons why this might be desirable: 1270 Multiple media types: Outside of WebRTC, it is common to use one RTP 1271 session for each type of media sources (e.g., one RTP session for 1272 audio sources and one for video sources, each sent over different 1273 transport layer flows). However, to reduce the number of UDP 1274 ports used, the default in WebRTC is to send all types of media in 1275 a single RTP session, as described in Section 4.4, using RTP and 1276 RTCP multiplexing (Section 4.5) to further reduce the number of 1277 UDP ports needed. This RTP session then uses only one bi- 1278 directional transport-layer flow, but will contain multiple RTP 1279 packet streams, each containing a different type of media. A 1280 common example might be an end-point with a camera and microphone 1281 that sends two RTP packet streams, one video and one audio, into a 1282 single RTP session. 1284 Multiple Capture Devices: A WebRTC Endpoint might have multiple 1285 cameras, microphones, or other media capture devices, and so might 1286 want to generate several RTP packet streams of the same media 1287 type. Alternatively, it might want to send media from a single 1288 capture device in several different formats or quality settings at 1289 once. Both can result in a single end-point sending multiple RTP 1290 packet streams of the same media type into a single RTP session at 1291 the same time. 1293 Associated Repair Data: An end-point might send a RTP packet stream 1294 that is somehow associated with another stream. For example, it 1295 might send an RTP packet stream that contains FEC or 1296 retransmission data relating to another stream. Some RTP payload 1297 formats send this sort of associated repair data as part of the 1298 source packet stream, while others send it as a separate packet 1299 stream. 1301 Layered or Multiple Description Coding: An end-point can use a 1302 layered media codec, for example H.264 SVC, or a multiple 1303 description codec, that generates multiple RTP packet streams, 1304 each with a distinct RTP SSRC, within a single RTP session. 1306 RTP Mixers, Translators, and Other Middleboxes: An RTP session, in 1307 the WebRTC context, is a point-to-point association between an 1308 end-point and some other peer device, where those devices share a 1309 common SSRC space. The peer device might be another WebRTC 1310 Endpoint, or it might be an RTP mixer, translator, or some other 1311 form of media processing middlebox. In the latter cases, the 1312 middlebox might send mixed or relayed RTP streams from several 1313 participants, that the WebRTC Endpoint will need to render. Thus, 1314 even though a WebRTC Endpoint might only be a member of a single 1315 RTP session, the peer device might be extending that RTP session 1316 to incorporate other end-points. WebRTC is a group communication 1317 environment and end-points need to be capable of receiving, 1318 decoding, and playing out multiple RTP packet streams at once, 1319 even in a single RTP session. 1321 12.1.2. Use of Multiple RTP Sessions 1323 In addition to sending and receiving multiple RTP packet streams 1324 within a single RTP session, a WebRTC Endpoint might participate in 1325 multiple RTP sessions. There are several reasons why a WebRTC 1326 Endpoint might choose to do this: 1328 To interoperate with legacy devices: The common practice in the non- 1329 WebRTC world is to send different types of media in separate RTP 1330 sessions, for example using one RTP session for audio and another 1331 RTP session, on a separate transport layer flow, for video. All 1332 WebRTC Endpoints need to support the option of sending different 1333 types of media on different RTP sessions, so they can interwork 1334 with such legacy devices. This is discussed further in 1335 Section 4.4. 1337 To provide enhanced quality of service: Some network-based quality 1338 of service mechanisms operate on the granularity of transport 1339 layer flows. If it is desired to use these mechanisms to provide 1340 differentiated quality of service for some RTP packet streams, 1341 then those RTP packet streams need to be sent in a separate RTP 1342 session using a different transport-layer flow, and with 1343 appropriate quality of service marking. This is discussed further 1344 in Section 12.1.3. 1346 To separate media with different purposes: An end-point might want 1347 to send RTP packet streams that have different purposes on 1348 different RTP sessions, to make it easy for the peer device to 1349 distinguish them. For example, some centralised multiparty 1350 conferencing systems display the active speaker in high 1351 resolution, but show low resolution "thumbnails" of other 1352 participants. Such systems might configure the end-points to send 1353 simulcast high- and low-resolution versions of their video using 1354 separate RTP sessions, to simplify the operation of the RTP 1355 middlebox. In the WebRTC context this is currently possible by 1356 establishing multiple WebRTC MediaStreamTracks that have the same 1357 media source in one (or more) RTCPeerConnection. Each 1358 MediaStreamTrack is then configured to deliver a particular media 1359 quality and thus media bit-rate, and will produce an independently 1360 encoded version with the codec parameters agreed specifically in 1361 the context of that RTCPeerConnection. The RTP middlebox can 1362 distinguish packets corresponding to the low- and high-resolution 1363 streams by inspecting their SSRC, RTP payload type, or some other 1364 information contained in RTP payload, RTP header extension or RTCP 1365 packets, but it can be easier to distinguish the RTP packet 1366 streams if they arrive on separate RTP sessions on separate 1367 transport-layer flows. 1369 To directly connect with multiple peers: A multi-party conference 1370 does not need to use an RTP middlebox. Rather, a multi-unicast 1371 mesh can be created, comprising several distinct RTP sessions, 1372 with each participant sending RTP traffic over a separate RTP 1373 session (that is, using an independent RTCPeerConnection object) 1374 to every other participant, as shown in Figure 1. This topology 1375 has the benefit of not requiring an RTP middlebox node that is 1376 trusted to access and manipulate the media data. The downside is 1377 that it increases the used bandwidth at each sender by requiring 1378 one copy of the RTP packet streams for each participant that are 1379 part of the same session beyond the sender itself. 1381 +---+ +---+ 1382 | A |<--->| B | 1383 +---+ +---+ 1384 ^ ^ 1385 \ / 1386 \ / 1387 v v 1388 +---+ 1389 | C | 1390 +---+ 1392 Figure 1: Multi-unicast using several RTP sessions 1394 The multi-unicast topology could also be implemented as a single 1395 RTP session, spanning multiple peer-to-peer transport layer 1396 connections, or as several pairwise RTP sessions, one between each 1397 pair of peers. To maintain a coherent mapping between the 1398 relation between RTP sessions and RTCPeerConnection objects it is 1399 recommend that this is implemented as several individual RTP 1400 sessions. The only downside is that end-point A will not learn of 1401 the quality of any transmission happening between B and C, since 1402 it will not see RTCP reports for the RTP session between B and C, 1403 whereas it would it all three participants were part of a single 1404 RTP session. Experience with the Mbone tools (experimental RTP- 1405 based multicast conferencing tools from the late 1990s) has showed 1406 that RTCP reception quality reports for third parties can be 1407 presented to users in a way that helps them understand asymmetric 1408 network problems, and the approach of using separate RTP sessions 1409 prevents this. However, an advantage of using separate RTP 1410 sessions is that it enables using different media bit-rates and 1411 RTP session configurations between the different peers, thus not 1412 forcing B to endure the same quality reductions if there are 1413 limitations in the transport from A to C as C will. It is 1414 believed that these advantages outweigh the limitations in 1415 debugging power. 1417 To indirectly connect with multiple peers: A common scenario in 1418 multi-party conferencing is to create indirect connections to 1419 multiple peers, using an RTP mixer, translator, or some other type 1420 of RTP middlebox. Figure 2 outlines a simple topology that might 1421 be used in a four-person centralised conference. The middlebox 1422 acts to optimise the transmission of RTP packet streams from 1423 certain perspectives, either by only sending some of the received 1424 RTP packet stream to any given receiver, or by providing a 1425 combined RTP packet stream out of a set of contributing streams. 1427 +---+ +-------------+ +---+ 1428 | A |<---->| |<---->| B | 1429 +---+ | RTP mixer, | +---+ 1430 | translator, | 1431 | or other | 1432 +---+ | middlebox | +---+ 1433 | C |<---->| |<---->| D | 1434 +---+ +-------------+ +---+ 1436 Figure 2: RTP mixer with only unicast paths 1438 There are various methods of implementation for the middlebox. If 1439 implemented as a standard RTP mixer or translator, a single RTP 1440 session will extend across the middlebox and encompass all the 1441 end-points in one multi-party session. Other types of middlebox 1442 might use separate RTP sessions between each end-point and the 1443 middlebox. A common aspect is that these RTP middleboxes can use 1444 a number of tools to control the media encoding provided by a 1445 WebRTC Endpoint. This includes functions like requesting the 1446 breaking of the encoding chain and have the encoder produce a so 1447 called Intra frame. Another is limiting the bit-rate of a given 1448 stream to better suit the mixer view of the multiple down-streams. 1449 Others are controlling the most suitable frame-rate, picture 1450 resolution, the trade-off between frame-rate and spatial quality. 1451 The middlebox has the responsibility to correctly perform 1452 congestion control, source identification, manage synchronisation 1453 while providing the application with suitable media optimisations. 1454 The middlebox also has to be a trusted node when it comes to 1455 security, since it manipulates either the RTP header or the media 1456 itself (or both) received from one end-point, before sending it on 1457 towards the end-point(s), thus they need to be able to decrypt and 1458 then re-encrypt the RTP packet stream before sending it out. 1460 RTP Mixers can create a situation where an end-point experiences a 1461 situation in-between a session with only two end-points and 1462 multiple RTP sessions. Mixers are expected to not forward RTCP 1463 reports regarding RTP packet streams across themselves. This is 1464 due to the difference in the RTP packet streams provided to the 1465 different end-points. The original media source lacks information 1466 about a mixer's manipulations prior to sending it the different 1467 receivers. This scenario also results in that an end-point's 1468 feedback or requests goes to the mixer. When the mixer can't act 1469 on this by itself, it is forced to go to the original media source 1470 to fulfil the receivers request. This will not necessarily be 1471 explicitly visible any RTP and RTCP traffic, but the interactions 1472 and the time to complete them will indicate such dependencies. 1474 Providing source authentication in multi-party scenarios is a 1475 challenge. In the mixer-based topologies, end-points source 1476 authentication is based on, firstly, verifying that media comes 1477 from the mixer by cryptographic verification and, secondly, trust 1478 in the mixer to correctly identify any source towards the end- 1479 point. In RTP sessions where multiple end-points are directly 1480 visible to an end-point, all end-points will have knowledge about 1481 each others' master keys, and can thus inject packets claimed to 1482 come from another end-point in the session. Any node performing 1483 relay can perform non-cryptographic mitigation by preventing 1484 forwarding of packets that have SSRC fields that came from other 1485 end-points before. For cryptographic verification of the source, 1486 SRTP would require additional security mechanisms, for example 1487 TESLA for SRTP [RFC4383], that are not part of the base WebRTC 1488 standards. 1490 To forward media between multiple peers: It is sometimes desirable 1491 for an end-point that receives an RTP packet stream to be able to 1492 forward that RTP packet stream to a third party. The are some 1493 obvious security and privacy implications in supporting this, but 1494 also potential uses. This is supported in the W3C API by taking 1495 the received and decoded media and using it as media source that 1496 is re-encoding and transmitted as a new stream. 1498 At the RTP layer, media forwarding acts as a back-to-back RTP 1499 receiver and RTP sender. The receiving side terminates the RTP 1500 session and decodes the media, while the sender side re-encodes 1501 and transmits the media using an entirely separate RTP session. 1502 The original sender will only see a single receiver of the media, 1503 and will not be able to tell that forwarding is happening based on 1504 RTP-layer information since the RTP session that is used to send 1505 the forwarded media is not connected to the RTP session on which 1506 the media was received by the node doing the forwarding. 1508 The end-point that is performing the forwarding is responsible for 1509 producing an RTP packet stream suitable for onwards transmission. 1510 The outgoing RTP session that is used to send the forwarded media 1511 is entirely separate to the RTP session on which the media was 1512 received. This will require media transcoding for congestion 1513 control purpose to produce a suitable bit-rate for the outgoing 1514 RTP session, reducing media quality and forcing the forwarding 1515 end-point to spend the resource on the transcoding. The media 1516 transcoding does result in a separation of the two different legs 1517 removing almost all dependencies, and allowing the forwarding end- 1518 point to optimise its media transcoding operation. The cost is 1519 greatly increased computational complexity on the forwarding node. 1520 Receivers of the forwarded stream will see the forwarding device 1521 as the sender of the stream, and will not be able to tell from the 1522 RTP layer that they are receiving a forwarded stream rather than 1523 an entirely new RTP packet stream generated by the forwarding 1524 device. 1526 12.1.3. Differentiated Treatment of RTP Packet Streams 1528 There are use cases for differentiated treatment of RTP packet 1529 streams. Such differentiation can happen at several places in the 1530 system. First of all is the prioritization within the end-point 1531 sending the media, which controls, both which RTP packet streams that 1532 will be sent, and their allocation of bit-rate out of the current 1533 available aggregate as determined by the congestion control. 1535 It is expected that the WebRTC API [W3C.WD-webrtc-20130910] will 1536 allow the application to indicate relative priorities for different 1537 MediaStreamTracks. These priorities can then be used to influence 1538 the local RTP processing, especially when it comes to congestion 1539 control response in how to divide the available bandwidth between the 1540 RTP packet streams. Any changes in relative priority will also need 1541 to be considered for RTP packet streams that are associated with the 1542 main RTP packet streams, such as redundant streams for RTP 1543 retransmission and FEC. The importance of such redundant RTP packet 1544 streams is dependent on the media type and codec used, in regards to 1545 how robust that codec is to packet loss. However, a default policy 1546 might to be to use the same priority for redundant RTP packet stream 1547 as for the source RTP packet stream. 1549 Secondly, the network can prioritize transport-layer flows and sub- 1550 flows, including RTP packet streams. Typically, differential 1551 treatment includes two steps, the first being identifying whether an 1552 IP packet belongs to a class that has to be treated differently, the 1553 second consisting of the actual mechanism to prioritize packets. 1554 Three common methods for classifying IP packets are: 1556 DiffServ: The end-point marks a packet with a DiffServ code point to 1557 indicate to the network that the packet belongs to a particular 1558 class. 1560 Flow based: Packets that need to be given a particular treatment are 1561 identified using a combination of IP and port address. 1563 Deep Packet Inspection: A network classifier (DPI) inspects the 1564 packet and tries to determine if the packet represents a 1565 particular application and type that is to be prioritized. 1567 Flow-based differentiation will provide the same treatment to all 1568 packets within a transport-layer flow, i.e., relative prioritization 1569 is not possible. Moreover, if the resources are limited it might not 1570 be possible to provide differential treatment compared to best-effort 1571 for all the RTP packet streams used in a WebRTC session. The use of 1572 flow-based differentiation needs to be coordinated between the WebRTC 1573 system and the network(s). The WebRTC endpoint needs to know that 1574 flow-based differentiation might be used to provide the separation of 1575 the RTP packet streams onto different UDP flows to enable a more 1576 granular usage of flow based differentiation. The used flows, their 1577 5-tuples and prioritization will need to be communicated to the 1578 network so that it can identify the flows correctly to enable 1579 prioritization. No specific protocol support for this is specified. 1581 DiffServ assumes that either the end-point or a classifier can mark 1582 the packets with an appropriate DSCP so that the packets are treated 1583 according to that marking. If the end-point is to mark the traffic 1584 two requirements arise in the WebRTC context: 1) The WebRTC Endpoint 1585 has to know which DSCP to use and that it can use them on some set of 1586 RTP packet streams. 2) The information needs to be propagated to the 1587 operating system when transmitting the packet. Details of this 1588 process are outside the scope of this memo and are further discussed 1589 in "DSCP and other packet markings for RTCWeb QoS" 1590 [I-D.ietf-tsvwg-rtcweb-qos]. 1592 Deep Packet Inspectors will, despite the SRTP media encryption, still 1593 be fairly capable at classifying the RTP streams. The reason is that 1594 SRTP leaves the first 12 bytes of the RTP header unencrypted. This 1595 enables easy RTP stream identification using the SSRC and provides 1596 the classifier with useful information that can be correlated to 1597 determine for example the stream's media type. Using packet sizes, 1598 reception times, packet inter-spacing, RTP timestamp increments and 1599 sequence numbers, fairly reliable classifications are achieved. 1601 For packet based marking schemes it might be possible to mark 1602 individual RTP packets differently based on the relative priority of 1603 the RTP payload. For example video codecs that have I, P, and B 1604 pictures could prioritise any payloads carrying only B frames less, 1605 as these are less damaging to loose. However, depending on the QoS 1606 mechanism and what markings that are applied, this can result in not 1607 only different packet drop probabilities but also packet reordering, 1608 see [I-D.ietf-tsvwg-rtcweb-qos] for further discussion. As a default 1609 policy all RTP packets related to a RTP packet stream ought to be 1610 provided with the same prioritization; per-packet prioritization is 1611 outside the scope of this memo, but might be specified elsewhere in 1612 future. 1614 It is also important to consider how RTCP packets associated with a 1615 particular RTP packet stream need to be marked. RTCP compound 1616 packets with Sender Reports (SR), ought to be marked with the same 1617 priority as the RTP packet stream itself, so the RTCP-based round- 1618 trip time (RTT) measurements are done using the same transport-layer 1619 flow priority as the RTP packet stream experiences. RTCP compound 1620 packets containing RR packet ought to be sent with the priority used 1621 by the majority of the RTP packet streams reported on. RTCP packets 1622 containing time-critical feedback packets can use higher priority to 1623 improve the timeliness and likelihood of delivery of such feedback. 1625 12.2. Media Source, RTP Packet Streams, and Participant Identification 1627 12.2.1. Media Source Identification 1629 Each RTP packet stream is identified by a unique synchronisation 1630 source (SSRC) identifier. The SSRC identifier is carried in each of 1631 the RTP packets comprising a RTP packet stream, and is also used to 1632 identify that stream in the corresponding RTCP reports. The SSRC is 1633 chosen as discussed in Section 4.8. The first stage in 1634 demultiplexing RTP and RTCP packets received on a single transport 1635 layer flow at a WebRTC Endpoint is to separate the RTP packet streams 1636 based on their SSRC value; once that is done, additional 1637 demultiplexing steps can determine how and where to render the media. 1639 RTP allows a mixer, or other RTP-layer middlebox, to combine encoded 1640 streams from multiple media sources to form a new encoded stream from 1641 a new media source (the mixer). The RTP packets in that new RTP 1642 packet stream can include a Contributing Source (CSRC) list, 1643 indicating which original SSRCs contributed to the combined source 1644 stream. As described in Section 4.1, implementations need to support 1645 reception of RTP data packets containing a CSRC list and RTCP packets 1646 that relate to sources present in the CSRC list. The CSRC list can 1647 change on a packet-by-packet basis, depending on the mixing operation 1648 being performed. Knowledge of what media sources contributed to a 1649 particular RTP packet can be important if the user interface 1650 indicates which participants are active in the session. Changes in 1651 the CSRC list included in packets needs to be exposed to the WebRTC 1652 application using some API, if the application is to be able to track 1653 changes in session participation. It is desirable to map CSRC values 1654 back into WebRTC MediaStream identities as they cross this API, to 1655 avoid exposing the SSRC/CSRC name space to WebRTC applications. 1657 If the mixer-to-client audio level extension [RFC6465] is being used 1658 in the session (see Section 5.2.3), the information in the CSRC list 1659 is augmented by audio level information for each contributing source. 1660 It is desirable to expose this information to the WebRTC application 1661 using some API, after mapping the CSRC values to WebRTC MediaStream 1662 identities, so it can be exposed in the user interface. 1664 12.2.2. SSRC Collision Detection 1666 The RTP standard requires RTP implementations to have support for 1667 detecting and handling SSRC collisions, i.e., resolve the conflict 1668 when two different end-points use the same SSRC value (see section 1669 8.2 of [RFC3550]). This requirement also applies to WebRTC 1670 Endpoints. There are several scenarios where SSRC collisions can 1671 occur: 1673 o In a point-to-point session where each SSRC is associated with 1674 either of the two end-points and where the main media carrying 1675 SSRC identifier will be announced in the signalling channel, a 1676 collision is less likely to occur due to the information about 1677 used SSRCs. If SDP is used, this information is provided by 1678 Source-Specific SDP Attributes [RFC5576]. Still, collisions can 1679 occur if both end-points start using a new SSRC identifier prior 1680 to having signalled it to the peer and received acknowledgement on 1681 the signalling message. The Source-Specific SDP Attributes 1682 [RFC5576] contains a mechanism to signal how the end-point 1683 resolved the SSRC collision. 1685 o SSRC values that have not been signalled could also appear in an 1686 RTP session. This is more likely than it appears, since some RTP 1687 functions use extra SSRCs to provide their functionality. For 1688 example, retransmission data might be transmitted using a separate 1689 RTP packet stream that requires its own SSRC, separate to the SSRC 1690 of the source RTP packet stream [RFC4588]. In those cases, an 1691 end-point can create a new SSRC that strictly doesn't need to be 1692 announced over the signalling channel to function correctly on 1693 both RTP and RTCPeerConnection level. 1695 o Multiple end-points in a multiparty conference can create new 1696 sources and signal those towards the RTP middlebox. In cases 1697 where the SSRC/CSRC are propagated between the different end- 1698 points from the RTP middlebox collisions can occur. 1700 o An RTP middlebox could connect an end-point's RTCPeerConnection to 1701 another RTCPeerConnection from the same end-point, thus forming a 1702 loop where the end-point will receive its own traffic. While it 1703 is clearly considered a bug, it is important that the end-point is 1704 able to recognise and handle the case when it occurs. This case 1705 becomes even more problematic when media mixers, and so on, are 1706 involved, where the stream received is a different stream but 1707 still contains this client's input. 1709 These SSRC/CSRC collisions can only be handled on RTP level as long 1710 as the same RTP session is extended across multiple 1711 RTCPeerConnections by a RTP middlebox. To resolve the more generic 1712 case where multiple RTCPeerConnections are interconnected, 1713 identification of the media source(s) part of a MediaStreamTrack 1714 being propagated across multiple interconnected RTCPeerConnection 1715 needs to be preserved across these interconnections. 1717 12.2.3. Media Synchronisation Context 1719 When an end-point sends media from more than one media source, it 1720 needs to consider if (and which of) these media sources are to be 1721 synchronized. In RTP/RTCP, synchronisation is provided by having a 1722 set of RTP packet streams be indicated as coming from the same 1723 synchronisation context and logical end-point by using the same RTCP 1724 CNAME identifier. 1726 The next provision is that the internal clocks of all media sources, 1727 i.e., what drives the RTP timestamp, can be correlated to a system 1728 clock that is provided in RTCP Sender Reports encoded in an NTP 1729 format. By correlating all RTP timestamps to a common system clock 1730 for all sources, the timing relation of the different RTP packet 1731 streams, also across multiple RTP sessions can be derived at the 1732 receiver and, if desired, the streams can be synchronized. The 1733 requirement is for the media sender to provide the correlation 1734 information; it is up to the receiver to use it or not. 1736 13. Security Considerations 1738 The overall security architecture for WebRTC is described in 1739 [I-D.ietf-rtcweb-security-arch], and security considerations for the 1740 WebRTC framework are described in [I-D.ietf-rtcweb-security]. These 1741 considerations also apply to this memo. 1743 The security considerations of the RTP specification, the RTP/SAVPF 1744 profile, and the various RTP/RTCP extensions and RTP payload formats 1745 that form the complete protocol suite described in this memo apply. 1746 It is not believed there are any new security considerations 1747 resulting from the combination of these various protocol extensions. 1749 The Extended Secure RTP Profile for Real-time Transport Control 1750 Protocol (RTCP)-Based Feedback [RFC5124] (RTP/SAVPF) provides 1751 handling of fundamental issues by offering confidentiality, integrity 1752 and partial source authentication. A mandatory to implement media 1753 security solution is created by combing this secured RTP profile and 1754 DTLS-SRTP keying [RFC5764] as defined by Section 5.5 of 1755 [I-D.ietf-rtcweb-security-arch]. 1757 RTCP packets convey a Canonical Name (CNAME) identifier that is used 1758 to associate RTP packet streams that need to be synchronised across 1759 related RTP sessions. Inappropriate choice of CNAME values can be a 1760 privacy concern, since long-term persistent CNAME identifiers can be 1761 used to track users across multiple WebRTC calls. Section 4.9 of 1762 this memo provides guidelines for generation of untraceable CNAME 1763 values that alleviate this risk. 1765 Some potential denial of service attacks exist if the RTCP reporting 1766 interval is configured to an inappropriate value. This could be done 1767 by configuring the RTCP bandwidth fraction to an excessively large or 1768 small value using the SDP "b=RR:" or "b=RS:" lines [RFC3556], or some 1769 similar mechanism, or by choosing an excessively large or small value 1770 for the RTP/AVPF minimal receiver report interval (if using SDP, this 1771 is the "a=rtcp-fb:... trr-int" parameter) [RFC4585]. The risks are 1772 as follows: 1774 1. the RTCP bandwidth could be configured to make the regular 1775 reporting interval so large that effective congestion control 1776 cannot be maintained, potentially leading to denial of service 1777 due to congestion caused by the media traffic; 1779 2. the RTCP interval could be configured to a very small value, 1780 causing endpoints to generate high rate RTCP traffic, potentially 1781 leading to denial of service due to the non-congestion controlled 1782 RTCP traffic; and 1784 3. RTCP parameters could be configured differently for each 1785 endpoint, with some of the endpoints using a large reporting 1786 interval and some using a smaller interval, leading to denial of 1787 service due to premature participant timeouts due to mismatched 1788 timeout periods which are based on the reporting interval (this 1789 is a particular concern if endpoints use a small but non-zero 1790 value for the RTP/AVPF minimal receiver report interval (trr-int) 1791 [RFC4585], as discussed in Section 6.1 of 1792 [I-D.ietf-avtcore-rtp-multi-stream]). 1794 Premature participant timeout can be avoided by using the fixed (non- 1795 reduced) minimum interval when calculating the participant timeout 1796 (see Section 4.1 of this memo and Section 6.1 of 1797 [I-D.ietf-avtcore-rtp-multi-stream]). To address the other concerns, 1798 endpoints SHOULD ignore parameters that configure the RTCP reporting 1799 interval to be significantly longer than the default five second 1800 interval specified in [RFC3550] (unless the media data rate is so low 1801 that the longer reporting interval roughly corresponds to 5% of the 1802 media data rate), or that configure the RTCP reporting interval small 1803 enough that the RTCP bandwidth would exceed the media bandwidth. 1805 The guidelines in [RFC6562] apply when using variable bit rate (VBR) 1806 audio codecs such as Opus (see Section 4.3 for discussion of mandated 1807 audio codecs). The guidelines in [RFC6562] also apply, but are of 1808 lesser importance, when using the client-to-mixer audio level header 1809 extensions (Section 5.2.2) or the mixer-to-client audio level header 1810 extensions (Section 5.2.3). The use of the encryption of the header 1811 extensions are RECOMMENDED, unless there are known reasons, like RTP 1812 middleboxes performing voice activity based source selection or third 1813 party monitoring that will greatly benefit from the information, and 1814 this has been expressed using API or signalling. If further evidence 1815 are produced to show that information leakage is significant from 1816 audio level indications, then use of encryption needs to be mandated 1817 at that time. 1819 14. IANA Considerations 1821 This memo makes no request of IANA. 1823 Note to RFC Editor: this section is to be removed on publication as 1824 an RFC. 1826 15. Acknowledgements 1828 The authors would like to thank Bernard Aboba, Harald Alvestrand, 1829 Cary Bran, Ben Campbell, Alissa Cooper, Charles Eckel, Alex 1830 Eleftheriadis, Christian Groves, Cullen Jennings, Olle Johansson, 1831 Suhas Nandakumar, Dan Romascanu, Jim Spring, Martin Thomson, and the 1832 other members of the IETF RTCWEB working group for their valuable 1833 feedback. 1835 16. References 1837 16.1. Normative References 1839 [I-D.ietf-avtcore-multi-media-rtp-session] 1840 Westerlund, M., Perkins, C., and J. Lennox, "Sending 1841 Multiple Types of Media in a Single RTP Session", draft- 1842 ietf-avtcore-multi-media-rtp-session-07 (work in 1843 progress), March 2015. 1845 [I-D.ietf-avtcore-rtp-circuit-breakers] 1846 Perkins, C. and V. Singh, "Multimedia Congestion Control: 1847 Circuit Breakers for Unicast RTP Sessions", draft-ietf- 1848 avtcore-rtp-circuit-breakers-10 (work in progress), March 1849 2015. 1851 [I-D.ietf-avtcore-rtp-multi-stream] 1852 Lennox, J., Westerlund, M., Wu, W., and C. Perkins, 1853 "Sending Multiple Media Streams in a Single RTP Session", 1854 draft-ietf-avtcore-rtp-multi-stream-07 (work in progress), 1855 March 2015. 1857 [I-D.ietf-avtcore-rtp-multi-stream-optimisation] 1858 Lennox, J., Westerlund, M., Wu, W., and C. Perkins, 1859 "Sending Multiple Media Streams in a Single RTP Session: 1860 Grouping RTCP Reception Statistics and Other Feedback", 1861 draft-ietf-avtcore-rtp-multi-stream-optimisation-05 (work 1862 in progress), February 2015. 1864 [I-D.ietf-mmusic-sdp-bundle-negotiation] 1865 Holmberg, C., Alvestrand, H., and C. Jennings, 1866 "Negotiating Media Multiplexing Using the Session 1867 Description Protocol (SDP)", draft-ietf-mmusic-sdp-bundle- 1868 negotiation-19 (work in progress), March 2015. 1870 [I-D.ietf-rtcweb-audio] 1871 Valin, J. and C. Bran, "WebRTC Audio Codec and Processing 1872 Requirements", draft-ietf-rtcweb-audio-08 (work in 1873 progress), April 2015. 1875 [I-D.ietf-rtcweb-fec] 1876 Uberti, J., "WebRTC Forward Error Correction 1877 Requirements", draft-ietf-rtcweb-fec-01 (work in 1878 progress), March 2015. 1880 [I-D.ietf-rtcweb-security] 1881 Rescorla, E., "Security Considerations for WebRTC", draft- 1882 ietf-rtcweb-security-08 (work in progress), February 2015. 1884 [I-D.ietf-rtcweb-security-arch] 1885 Rescorla, E., "WebRTC Security Architecture", draft-ietf- 1886 rtcweb-security-arch-11 (work in progress), March 2015. 1888 [I-D.ietf-rtcweb-video] 1889 Roach, A., "WebRTC Video Processing and Codec 1890 Requirements", draft-ietf-rtcweb-video-05 (work in 1891 progress), March 2015. 1893 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1894 Requirement Levels", BCP 14, RFC 2119, March 1997. 1896 [RFC2736] Handley, M. and C. Perkins, "Guidelines for Writers of RTP 1897 Payload Format Specifications", BCP 36, RFC 2736, December 1898 1999. 1900 [RFC3550] Schulzrinne, H., Casner, S., Frederick, R., and V. 1901 Jacobson, "RTP: A Transport Protocol for Real-Time 1902 Applications", STD 64, RFC 3550, July 2003. 1904 [RFC3551] Schulzrinne, H. and S. Casner, "RTP Profile for Audio and 1905 Video Conferences with Minimal Control", STD 65, RFC 3551, 1906 July 2003. 1908 [RFC3556] Casner, S., "Session Description Protocol (SDP) Bandwidth 1909 Modifiers for RTP Control Protocol (RTCP) Bandwidth", RFC 1910 3556, July 2003. 1912 [RFC3711] Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K. 1913 Norrman, "The Secure Real-time Transport Protocol (SRTP)", 1914 RFC 3711, March 2004. 1916 [RFC4566] Handley, M., Jacobson, V., and C. Perkins, "SDP: Session 1917 Description Protocol", RFC 4566, July 2006. 1919 [RFC4585] Ott, J., Wenger, S., Sato, N., Burmeister, C., and J. Rey, 1920 "Extended RTP Profile for Real-time Transport Control 1921 Protocol (RTCP)-Based Feedback (RTP/AVPF)", RFC 4585, July 1922 2006. 1924 [RFC4588] Rey, J., Leon, D., Miyazaki, A., Varsa, V., and R. 1925 Hakenberg, "RTP Retransmission Payload Format", RFC 4588, 1926 July 2006. 1928 [RFC4961] Wing, D., "Symmetric RTP / RTP Control Protocol (RTCP)", 1929 BCP 131, RFC 4961, July 2007. 1931 [RFC5104] Wenger, S., Chandra, U., Westerlund, M., and B. Burman, 1932 "Codec Control Messages in the RTP Audio-Visual Profile 1933 with Feedback (AVPF)", RFC 5104, February 2008. 1935 [RFC5124] Ott, J. and E. Carrara, "Extended Secure RTP Profile for 1936 Real-time Transport Control Protocol (RTCP)-Based Feedback 1937 (RTP/SAVPF)", RFC 5124, February 2008. 1939 [RFC5285] Singer, D. and H. Desineni, "A General Mechanism for RTP 1940 Header Extensions", RFC 5285, July 2008. 1942 [RFC5506] Johansson, I. and M. Westerlund, "Support for Reduced-Size 1943 Real-Time Transport Control Protocol (RTCP): Opportunities 1944 and Consequences", RFC 5506, April 2009. 1946 [RFC5761] Perkins, C. and M. Westerlund, "Multiplexing RTP Data and 1947 Control Packets on a Single Port", RFC 5761, April 2010. 1949 [RFC5764] McGrew, D. and E. Rescorla, "Datagram Transport Layer 1950 Security (DTLS) Extension to Establish Keys for the Secure 1951 Real-time Transport Protocol (SRTP)", RFC 5764, May 2010. 1953 [RFC6051] Perkins, C. and T. Schierl, "Rapid Synchronisation of RTP 1954 Flows", RFC 6051, November 2010. 1956 [RFC6464] Lennox, J., Ivov, E., and E. Marocco, "A Real-time 1957 Transport Protocol (RTP) Header Extension for Client-to- 1958 Mixer Audio Level Indication", RFC 6464, December 2011. 1960 [RFC6465] Ivov, E., Marocco, E., and J. Lennox, "A Real-time 1961 Transport Protocol (RTP) Header Extension for Mixer-to- 1962 Client Audio Level Indication", RFC 6465, December 2011. 1964 [RFC6562] Perkins, C. and JM. Valin, "Guidelines for the Use of 1965 Variable Bit Rate Audio with Secure RTP", RFC 6562, March 1966 2012. 1968 [RFC6904] Lennox, J., "Encryption of Header Extensions in the Secure 1969 Real-time Transport Protocol (SRTP)", RFC 6904, April 1970 2013. 1972 [RFC7007] Terriberry, T., "Update to Remove DVI4 from the 1973 Recommended Codecs for the RTP Profile for Audio and Video 1974 Conferences with Minimal Control (RTP/AVP)", RFC 7007, 1975 August 2013. 1977 [RFC7022] Begen, A., Perkins, C., Wing, D., and E. Rescorla, 1978 "Guidelines for Choosing RTP Control Protocol (RTCP) 1979 Canonical Names (CNAMEs)", RFC 7022, September 2013. 1981 [RFC7160] Petit-Huguenin, M. and G. Zorn, "Support for Multiple 1982 Clock Rates in an RTP Session", RFC 7160, April 2014. 1984 [RFC7164] Gross, K. and R. Brandenburg, "RTP and Leap Seconds", RFC 1985 7164, March 2014. 1987 16.2. Informative References 1989 [I-D.ietf-avtcore-multiplex-guidelines] 1990 Westerlund, M., Perkins, C., and H. Alvestrand, 1991 "Guidelines for using the Multiplexing Features of RTP to 1992 Support Multiple Media Streams", draft-ietf-avtcore- 1993 multiplex-guidelines-03 (work in progress), October 2014. 1995 [I-D.ietf-avtcore-rtp-topologies-update] 1996 Westerlund, M. and S. Wenger, "RTP Topologies", draft- 1997 ietf-avtcore-rtp-topologies-update-07 (work in progress), 1998 April 2015. 2000 [I-D.ietf-avtext-rtp-grouping-taxonomy] 2001 Lennox, J., Gross, K., Nandakumar, S., Salgueiro, G., and 2002 B. Burman, "A Taxonomy of Grouping Semantics and 2003 Mechanisms for Real-Time Transport Protocol (RTP) 2004 Sources", draft-ietf-avtext-rtp-grouping-taxonomy-06 (work 2005 in progress), March 2015. 2007 [I-D.ietf-mmusic-msid] 2008 Alvestrand, H., "WebRTC MediaStream Identification in the 2009 Session Description Protocol", draft-ietf-mmusic-msid-10 2010 (work in progress), April 2015. 2012 [I-D.ietf-payload-rtp-howto] 2013 Westerlund, M., "How to Write an RTP Payload Format", 2014 draft-ietf-payload-rtp-howto-14 (work in progress), May 2015 2015. 2017 [I-D.ietf-rmcat-cc-requirements] 2018 Jesup, R. and Z. Sarker, "Congestion Control Requirements 2019 for Interactive Real-Time Media", draft-ietf-rmcat-cc- 2020 requirements-09 (work in progress), December 2014. 2022 [I-D.ietf-rtcweb-overview] 2023 Alvestrand, H., "Overview: Real Time Protocols for 2024 Browser-based Applications", draft-ietf-rtcweb-overview-13 2025 (work in progress), November 2014. 2027 [I-D.ietf-tsvwg-rtcweb-qos] 2028 Dhesikan, S., Jennings, C., Druta, D., Jones, P., and J. 2029 Polk, "DSCP and other packet markings for RTCWeb QoS", 2030 draft-ietf-tsvwg-rtcweb-qos-03 (work in progress), 2031 November 2014. 2033 [RFC3611] Friedman, T., Caceres, R., and A. Clark, "RTP Control 2034 Protocol Extended Reports (RTCP XR)", RFC 3611, November 2035 2003. 2037 [RFC4383] Baugher, M. and E. Carrara, "The Use of Timed Efficient 2038 Stream Loss-Tolerant Authentication (TESLA) in the Secure 2039 Real-time Transport Protocol (SRTP)", RFC 4383, February 2040 2006. 2042 [RFC5245] Rosenberg, J., "Interactive Connectivity Establishment 2043 (ICE): A Protocol for Network Address Translator (NAT) 2044 Traversal for Offer/Answer Protocols", RFC 5245, April 2045 2010. 2047 [RFC5576] Lennox, J., Ott, J., and T. Schierl, "Source-Specific 2048 Media Attributes in the Session Description Protocol 2049 (SDP)", RFC 5576, June 2009. 2051 [RFC5968] Ott, J. and C. Perkins, "Guidelines for Extending the RTP 2052 Control Protocol (RTCP)", RFC 5968, September 2010. 2054 [RFC6263] Marjou, X. and A. Sollaud, "Application Mechanism for 2055 Keeping Alive the NAT Mappings Associated with RTP / RTP 2056 Control Protocol (RTCP) Flows", RFC 6263, June 2011. 2058 [RFC6792] Wu, Q., Hunt, G., and P. Arden, "Guidelines for Use of the 2059 RTP Monitoring Framework", RFC 6792, November 2012. 2061 [RFC7478] Holmberg, C., Hakansson, S., and G. Eriksson, "Web Real- 2062 Time Communication Use Cases and Requirements", RFC 7478, 2063 March 2015. 2065 [W3C.WD-mediacapture-streams-20130903] 2066 Burnett, D., Bergkvist, A., Jennings, C., and A. 2067 Narayanan, "Media Capture and Streams", World Wide Web 2068 Consortium WD WD-mediacapture-streams-20130903, September 2069 2013, . 2072 [W3C.WD-webrtc-20130910] 2073 Bergkvist, A., Burnett, D., Jennings, C., and A. 2074 Narayanan, "WebRTC 1.0: Real-time Communication Between 2075 Browsers", World Wide Web Consortium WD WD-webrtc- 2076 20130910, September 2013, 2077 . 2079 Authors' Addresses 2081 Colin Perkins 2082 University of Glasgow 2083 School of Computing Science 2084 Glasgow G12 8QQ 2085 United Kingdom 2087 Email: csp@csperkins.org 2088 URI: http://csperkins.org/ 2089 Magnus Westerlund 2090 Ericsson 2091 Farogatan 6 2092 SE-164 80 Kista 2093 Sweden 2095 Phone: +46 10 714 82 87 2096 Email: magnus.westerlund@ericsson.com 2098 Joerg Ott 2099 Aalto University 2100 School of Electrical Engineering 2101 Espoo 02150 2102 Finland 2104 Email: jorg.ott@aalto.fi