idnits 2.17.1 draft-ietf-rtcweb-rtp-usage-18.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 (October 21, 2014) is 3473 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(See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: A later version (-13) exists of draft-ietf-avtcore-multi-media-rtp-session-06 == Outdated reference: A later version (-18) exists of draft-ietf-avtcore-rtp-circuit-breakers-06 == Outdated reference: A later version (-12) exists of draft-ietf-avtcore-rtp-multi-stream-optimisation-00 == Outdated reference: A later version (-11) exists of draft-ietf-avtcore-rtp-multi-stream-05 == Outdated reference: A later version (-20) exists of draft-ietf-rtcweb-security-arch-10 == Outdated reference: A later version (-12) exists of draft-ietf-rtcweb-security-07 ** Obsolete normative reference: RFC 4566 (Obsoleted by RFC 8866) ** Obsolete normative reference: RFC 5285 (Obsoleted by RFC 8285) == Outdated reference: A later version (-12) exists of draft-ietf-avtcore-multiplex-guidelines-03 == Outdated reference: A later version (-10) exists of draft-ietf-avtcore-rtp-topologies-update-04 == Outdated reference: A later version (-08) exists of draft-ietf-avtext-rtp-grouping-taxonomy-02 == Outdated reference: A later version (-17) exists of draft-ietf-mmusic-msid-07 == Outdated reference: A later version (-54) exists of draft-ietf-mmusic-sdp-bundle-negotiation-12 == Outdated reference: A later version (-14) exists of draft-ietf-payload-rtp-howto-13 == Outdated reference: A later version (-09) exists of draft-ietf-rmcat-cc-requirements-06 == Outdated reference: A later version (-11) exists of draft-ietf-rtcweb-audio-06 == Outdated reference: A later version (-19) exists of draft-ietf-rtcweb-overview-12 == Outdated reference: A later version (-16) exists of draft-ietf-rtcweb-use-cases-and-requirements-14 == Outdated reference: A later version (-18) exists of draft-ietf-tsvwg-rtcweb-qos-02 -- Obsolete informational reference (is this intentional?): RFC 5245 (Obsoleted by RFC 8445, RFC 8839) Summary: 2 errors (**), 0 flaws (~~), 18 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 RTCWEB Working Group C. S. Perkins 3 Internet-Draft University of Glasgow 4 Intended status: Standards Track M. Westerlund 5 Expires: April 24, 2015 Ericsson 6 J. Ott 7 Aalto University 8 October 21, 2014 10 Web Real-Time Communication (WebRTC): Media Transport and Use of RTP 11 draft-ietf-rtcweb-rtp-usage-18 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 April 24, 2015. 40 Copyright Notice 42 Copyright (c) 2014 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 . . . . . . . . . . . . . . . . . . . . 10 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) . . . . 15 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 . . . . . . . . . . . . . 17 84 6. WebRTC Use of RTP: Improving Transport Robustness . . . . . . 18 85 6.1. Negative Acknowledgements and RTP Retransmission . . . . 18 86 6.2. Forward Error Correction (FEC) . . . . . . . . . . . . . 19 87 7. WebRTC Use of RTP: Rate Control and Media Adaptation . . . . 19 88 7.1. Boundary Conditions and Circuit Breakers . . . . . . . . 20 89 7.2. Congestion Control Interoperability and Legacy Systems . 21 90 8. WebRTC Use of RTP: Performance Monitoring . . . . . . . . . . 22 91 9. WebRTC Use of RTP: Future Extensions . . . . . . . . . . . . 22 92 10. Signalling Considerations . . . . . . . . . . . . . . . . . . 22 93 11. WebRTC API Considerations . . . . . . . . . . . . . . . . . . 24 94 12. RTP Implementation Considerations . . . . . . . . . . . . . . 26 95 12.1. Configuration and Use of RTP Sessions . . . . . . . . . 26 96 12.1.1. Use of Multiple Media Sources Within an RTP Session 26 97 12.1.2. Use of Multiple RTP Sessions . . . . . . . . . . . . 28 98 12.1.3. Differentiated Treatment of RTP Packet Streams . . . 32 99 12.2. Media Source, RTP Packet Streams, and Participant 100 Identification . . . . . . . . . . . . . . . . . . . . . 34 101 12.2.1. Media Source Identification . . . . . . . . . . . . 34 102 12.2.2. SSRC Collision Detection . . . . . . . . . . . . . . 34 103 12.2.3. Media Synchronisation Context . . . . . . . . . . . 36 104 13. Security Considerations . . . . . . . . . . . . . . . . . . . 36 105 14. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 38 106 15. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 38 107 16. References . . . . . . . . . . . . . . . . . . . . . . . . . 38 108 16.1. Normative References . . . . . . . . . . . . . . . . . . 38 109 16.2. Informative References . . . . . . . . . . . . . . . . . 41 110 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 43 112 1. Introduction 114 The Real-time Transport Protocol (RTP) [RFC3550] provides a framework 115 for delivery of audio and video teleconferencing data and other real- 116 time media applications. Previous work has defined the RTP protocol, 117 along with numerous profiles, payload formats, and other extensions. 118 When combined with appropriate signalling, these form the basis for 119 many teleconferencing systems. 121 The Web Real-Time communication (WebRTC) framework provides the 122 protocol building blocks to support direct, interactive, real-time 123 communication using audio, video, collaboration, games, etc., between 124 two peers' web-browsers. This memo describes how the RTP framework 125 is to be used in the WebRTC context. It proposes a baseline set of 126 RTP features that are to be implemented by all WebRTC Endpoints, 127 along with suggested extensions for enhanced functionality. 129 This memo specifies a protocol intended for use within the WebRTC 130 framework, but is not restricted to that context. An overview of the 131 WebRTC framework is given in [I-D.ietf-rtcweb-overview]. 133 The structure of this memo is as follows. Section 2 outlines our 134 rationale in preparing this memo and choosing these RTP features. 135 Section 3 defines terminology. Requirements for core RTP protocols 136 are described in Section 4 and suggested RTP extensions are described 137 in Section 5. Section 6 outlines mechanisms that can increase 138 robustness to network problems, while Section 7 describes congestion 139 control and rate adaptation mechanisms. The discussion of mandated 140 RTP mechanisms concludes in Section 8 with a review of performance 141 monitoring and network management tools. Section 9 gives some 142 guidelines for future incorporation of other RTP and RTP Control 143 Protocol (RTCP) extensions into this framework. Section 10 describes 144 requirements placed on the signalling channel. Section 11 discusses 145 the relationship between features of the RTP framework and the WebRTC 146 application programming interface (API), and Section 12 discusses RTP 147 implementation considerations. The memo concludes with security 148 considerations (Section 13) and IANA considerations (Section 14). 150 2. Rationale 152 The RTP framework comprises the RTP data transfer protocol, the RTP 153 control protocol, and numerous RTP payload formats, profiles, and 154 extensions. This range of add-ons has allowed RTP to meet various 155 needs that were not envisaged by the original protocol designers, and 156 to support many new media encodings, but raises the question of what 157 extensions are to be supported by new implementations. The 158 development of the WebRTC framework provides an opportunity to review 159 the available RTP features and extensions, and to define a common 160 baseline RTP feature set for all WebRTC Endpoints. This builds on 161 the past 20 years development of RTP to mandate the use of extensions 162 that have shown widespread utility, while still remaining compatible 163 with the wide installed base of RTP implementations where possible. 165 RTP and RTCP extensions that are not discussed in this document can 166 be implemented by WebRTC Endpoints if they are beneficial for new use 167 cases. However, they are not necessary to address the WebRTC use 168 cases and requirements identified in 169 [I-D.ietf-rtcweb-use-cases-and-requirements]. 171 While the baseline set of RTP features and extensions defined in this 172 memo is targeted at the requirements of the WebRTC framework, it is 173 expected to be broadly useful for other conferencing-related uses of 174 RTP. In particular, it is likely that this set of RTP features and 175 extensions will be appropriate for other desktop or mobile video 176 conferencing systems, or for room-based high-quality telepresence 177 applications. 179 3. Terminology 181 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 182 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 183 document are to be interpreted as described in [RFC2119]. The RFC 184 2119 interpretation of these key words applies only when written in 185 ALL CAPS. Lower- or mixed-case uses of these key words are not to be 186 interpreted as carrying special significance in this memo. 188 We define the following additional terms: 190 WebRTC MediaStream: The MediaStream concept defined by the W3C in 191 the WebRTC API [W3C.WD-mediacapture-streams-20130903]. 193 Transport-layer Flow: A uni-directional flow of transport packets 194 that are identified by having a particular 5-tuple of source IP 195 address, source port, destination IP address, destination port, 196 and transport protocol used. 198 Bi-directional Transport-layer Flow: A bi-directional transport- 199 layer flow is a transport-layer flow that is symmetric. That is, 200 the transport-layer flow in the reverse direction has a 5-tuple 201 where the source and destination address and ports are swapped 202 compared to the forward path transport-layer flow, and the 203 transport protocol is the same. 205 This document uses the terminology from 206 [I-D.ietf-avtext-rtp-grouping-taxonomy] and 207 [I-D.ietf-rtcweb-overview]. Other terms are used according to their 208 definitions from the RTP Specification [RFC3550]. Especially note 209 the following frequently used terms: RTP Packet Stream, RTP Session, 210 and End-point. 212 4. WebRTC Use of RTP: Core Protocols 214 The following sections describe the core features of RTP and RTCP 215 that need to be implemented, along with the mandated RTP profiles. 216 Also described are the core extensions providing essential features 217 that all WebRTC Endpoints need to implement to function effectively 218 on today's networks. 220 4.1. RTP and RTCP 222 The Real-time Transport Protocol (RTP) [RFC3550] is REQUIRED to be 223 implemented as the media transport protocol for WebRTC. RTP itself 224 comprises two parts: the RTP data transfer protocol, and the RTP 225 control protocol (RTCP). RTCP is a fundamental and integral part of 226 RTP, and MUST be implemented and used in all WebRTC Endpoints. 228 The following RTP and RTCP features are sometimes omitted in limited 229 functionality implementations of RTP, but are REQUIRED in all WebRTC 230 Endpoints: 232 o Support for use of multiple simultaneous SSRC values in a single 233 RTP session, including support for RTP end-points that send many 234 SSRC values simultaneously, following [RFC3550] and 235 [I-D.ietf-avtcore-rtp-multi-stream]. The RTCP optimisations for 236 multi-SSRC sessions defined in 237 [I-D.ietf-avtcore-rtp-multi-stream-optimisation] MAY be supported; 238 if supported the usage MUST be signalled. 240 o Random choice of SSRC on joining a session; collision detection 241 and resolution for SSRC values (see also Section 4.8). 243 o Support for reception of RTP data packets containing CSRC lists, 244 as generated by RTP mixers, and RTCP packets relating to CSRCs. 246 o Sending correct synchronisation information in the RTCP Sender 247 Reports, to allow receivers to implement lip-synchronisation; see 248 Section 5.2.1 regarding support for the rapid RTP synchronisation 249 extensions. 251 o Support for multiple synchronisation contexts. Participants that 252 send multiple simultaneous RTP packet streams SHOULD do so as part 253 of a single synchronisation context, using a single RTCP CNAME for 254 all streams and allowing receivers to play the streams out in a 255 synchronised manner. For compatibility with potential future 256 versions of this specification, or for interoperability with non- 257 WebRTC devices through a gateway, receivers MUST support multiple 258 synchronisation contexts, indicated by the use of multiple RTCP 259 CNAMEs in an RTP session. This specification requires the usage 260 of a single CNAME when sending RTP Packet Streams in some 261 circumstances, see Section 4.9. 263 o Support for sending and receiving RTCP SR, RR, SDES, and BYE 264 packet types, with OPTIONAL support for other RTCP packet types 265 unless mandated by other parts of this specification. Note that 266 additional RTCP Packet types are used by the RTP/SAVPF Profile 267 (Section 4.2) and the other RTCP extensions (Section 5). 269 o Support for multiple end-points in a single RTP session, and for 270 scaling the RTCP transmission interval according to the number of 271 participants in the session; support for randomised RTCP 272 transmission intervals to avoid synchronisation of RTCP reports; 273 support for RTCP timer reconsideration (Section 6.3.6 of 274 [RFC3550]) and reverse reconsideration (Section 6.3.4 of 275 [RFC3550]). 277 o Support for configuring the RTCP bandwidth as a fraction of the 278 media bandwidth, and for configuring the fraction of the RTCP 279 bandwidth allocated to senders, e.g., using the SDP "b=" line 280 [RFC4566][RFC3556]. 282 o Support for the reduced minimum RTCP reporting interval described 283 in Section 6.2 of [RFC3550] is REQUIRED. When using the reduced 284 minimum RTCP reporting interval, the fixed (non-reduced) minimum 285 interval MUST be used when calculating the participant timeout 286 interval (see Sections 6.2 and 6.3.5 of [RFC3550]). The delay 287 before sending the initial compound RTCP packet can be set to zero 288 (see Section 6.2 of [RFC3550] as updated by 289 [I-D.ietf-avtcore-rtp-multi-stream]). 291 o Support for discontinuous transmission. RTP allows endpoints to 292 pause and resume transmission at any time. When resuming, the RTP 293 sequence number will increase by one, as usual, while the increase 294 in the RTP timestamp value will depend on the duration of the 295 pause. Discontinuous transmission is most commonly used with some 296 audio payload formats, but is not audio specific, and can be used 297 with any RTP payload format. 299 o Ignore unknown RTCP packet types and RTP header extensions. This 300 to ensure robust handling of future extensions, middlebox 301 behaviours, etc., that can result in not signalled RTCP packet 302 types or RTP header extensions being received. If a compound RTCP 303 packet is received that contains a mixture of known and unknown 304 RTCP packet types, the known packets types need to be processed as 305 usual, with only the unknown packet types being discarded. 307 It is known that a significant number of legacy RTP implementations, 308 especially those targeted at VoIP-only systems, do not support all of 309 the above features, and in some cases do not support RTCP at all. 310 Implementers are advised to consider the requirements for graceful 311 degradation when interoperating with legacy implementations. 313 Other implementation considerations are discussed in Section 12. 315 4.2. Choice of the RTP Profile 317 The complete specification of RTP for a particular application domain 318 requires the choice of an RTP Profile. For WebRTC use, the Extended 319 Secure RTP Profile for RTCP-Based Feedback (RTP/SAVPF) [RFC5124], as 320 extended by [RFC7007], MUST be implemented. The RTP/SAVPF profile is 321 the combination of basic RTP/AVP profile [RFC3551], the RTP profile 322 for RTCP-based feedback (RTP/AVPF) [RFC4585], and the secure RTP 323 profile (RTP/SAVP) [RFC3711]. 325 The RTCP-based feedback extensions [RFC4585] are needed for the 326 improved RTCP timer model. This allows more flexible transmission of 327 RTCP packets in response to events, rather than strictly according to 328 bandwidth, and is vital for being able to report congestion signals 329 as well as media events. These extensions also allow saving RTCP 330 bandwidth, and an end-point will commonly only use the full RTCP 331 bandwidth allocation if there are many events that require feedback. 332 The timer rules are also needed to make use of the RTP conferencing 333 extensions discussed in Section 5.1. 335 Note: The enhanced RTCP timer model defined in the RTP/AVPF 336 profile is backwards compatible with legacy systems that implement 337 only the RTP/AVP or RTP/SAVP profile, given some constraints on 338 parameter configuration such as the RTCP bandwidth value and "trr- 339 int" (the most important factor for interworking with RTP/(S)AVP 340 end-points via a gateway is to set the trr-int parameter to a 341 value representing 4 seconds, see Section 6.1 in 342 [I-D.ietf-avtcore-rtp-multi-stream]). 344 The secure RTP (SRTP) profile extensions [RFC3711] are needed to 345 provide media encryption, integrity protection, replay protection and 346 a limited form of source authentication. WebRTC Endpoints MUST NOT 347 send packets using the basic RTP/AVP profile or the RTP/AVPF profile; 348 they MUST employ the full RTP/SAVPF profile to protect all RTP and 349 RTCP packets that are generated (i.e., implementations MUST use SRTP 350 and SRTCP). The RTP/SAVPF profile MUST be configured using the 351 cipher suites, DTLS-SRTP protection profiles, keying mechanisms, and 352 other parameters described in [I-D.ietf-rtcweb-security-arch]. 354 4.3. Choice of RTP Payload Formats 356 The set of mandatory to implement codecs and RTP payload formats for 357 WebRTC is not specified in this memo, instead they are defined in 358 separate specifications, such as [I-D.ietf-rtcweb-audio]. 359 Implementations can support any codec for which an RTP payload format 360 and associated signalling is defined. Implementation cannot assume 361 that the other participants in an RTP session understand any RTP 362 payload format, no matter how common; the mapping between RTP payload 363 type numbers and specific configurations of particular RTP payload 364 formats MUST be agreed before those payload types/formats can be 365 used. In an SDP context, this can be done using the "a=rtpmap:" and 366 "a=fmtp:" attributes associated with an "m=" line, along with any 367 other SDP attributes needed to configure the RTP payload format. 369 End-points can signal support for multiple RTP payload formats, or 370 multiple configurations of a single RTP payload format, as long as 371 each unique RTP payload format configuration uses a different RTP 372 payload type number. As outlined in Section 4.8, the RTP payload 373 type number is sometimes used to associate an RTP packet stream with 374 a signalling context. This association is possible provided unique 375 RTP payload type numbers are used in each context. For example, an 376 RTP packet stream can be associated with an SDP "m=" line by 377 comparing the RTP payload type numbers used by the RTP packet stream 378 with payload types signalled in the "a=rtpmap:" lines in the media 379 sections of the SDP. This leads to the following considerations: 381 If RTP packet streams are being associated with signalling 382 contexts based on the RTP payload type, then the assignment of RTP 383 payload type numbers MUST be unique across signalling contexts. 385 If the same RTP payload format configuration is used in multiple 386 contexts, then a different RTP payload type number has to be 387 assigned in each context to ensure uniqueness. 389 If the RTP payload type number is not being used to associate RTP 390 packet streams with a signalling context, then the same RTP 391 payload type number can be used to indicate the exact same RTP 392 payload format configuration in multiple contexts. 394 A single RTP payload type number MUST NOT be assigned to different 395 RTP payload formats, or different configurations of the same RTP 396 payload format, within a single RTP session (note that the "m=" lines 397 in an SDP bundle group [I-D.ietf-mmusic-sdp-bundle-negotiation] form 398 a single RTP session). 400 An end-point that has signalled support for multiple RTP payload 401 formats MUST be able to accept data in any of those payload formats 402 at any time, unless it has previously signalled limitations on its 403 decoding capability. This requirement is constrained if several 404 types of media (e.g., audio and video) are sent in the same RTP 405 session. In such a case, a source (SSRC) is restricted to switching 406 only between the RTP payload formats signalled for the type of media 407 that is being sent by that source; see Section 4.4. To support rapid 408 rate adaptation by changing codec, RTP does not require advance 409 signalling for changes between RTP payload formats used by a single 410 SSRC that were signalled during session set-up. 412 If performing changes between two RTP payload types that use 413 different RTP clock rates, an RTP sender MUST follow the 414 recommendations in Section 4.1 of [RFC7160]. RTP receivers MUST 415 follow the recommendations in Section 4.3 of [RFC7160] in order to 416 support sources that switch between clock rates in an RTP session 417 (these recommendations for receivers are backwards compatible with 418 the case where senders use only a single clock rate). 420 4.4. Use of RTP Sessions 422 An association amongst a set of end-points communicating using RTP is 423 known as an RTP session [RFC3550]. An end-point can be involved in 424 several RTP sessions at the same time. In a multimedia session, each 425 type of media has typically been carried in a separate RTP session 426 (e.g., using one RTP session for the audio, and a separate RTP 427 session using a different transport-layer flow for the video). 428 WebRTC Endpoints are REQUIRED to implement support for multimedia 429 sessions in this way, separating each RTP session using different 430 transport-layer flows for compatibility with legacy systems. 432 In modern day networks, however, with the widespread use of network 433 address/port translators (NAT/NAPT) and firewalls, it is desirable to 434 reduce the number of transport-layer flows used by RTP applications. 435 This can be done by sending all the RTP packet streams in a single 436 RTP session, which will comprise a single transport-layer flow (this 437 will prevent the use of some quality-of-service mechanisms, as 438 discussed in Section 12.1.3). Implementations are therefore also 439 REQUIRED to support transport of all RTP packet streams, independent 440 of media type, in a single RTP session using a single transport layer 441 flow, according to [I-D.ietf-avtcore-multi-media-rtp-session]. If 442 multiple types of media are to be used in a single RTP session, all 443 participants in that RTP session MUST agree to this usage. In an SDP 444 context, [I-D.ietf-mmusic-sdp-bundle-negotiation] can be used to 445 signal such a bundle of RTP packet streams forming a single RTP 446 session. 448 Further discussion about the suitability of different RTP session 449 structures and multiplexing methods to different scenarios can be 450 found in [I-D.ietf-avtcore-multiplex-guidelines]. 452 4.5. RTP and RTCP Multiplexing 454 Historically, RTP and RTCP have been run on separate transport layer 455 flows (e.g., two UDP ports for each RTP session, one port for RTP and 456 one port for RTCP). With the increased use of Network Address/Port 457 Translation (NAT/NAPT) this has become problematic, since maintaining 458 multiple NAT bindings can be costly. It also complicates firewall 459 administration, since multiple ports need to be opened to allow RTP 460 traffic. To reduce these costs and session set-up times, 461 implementations are REQUIRED to support multiplexing RTP data packets 462 and RTCP control packets on a single transport-layer flow [RFC5761]. 463 Such RTP and RTCP multiplexing MUST be negotiated in the signalling 464 channel before it is used. If SDP is used for signalling, this 465 negotiation MUST use the attributes defined in [RFC5761]. For 466 backwards compatibility, implementations are also REQUIRED to support 467 RTP and RTCP sent on separate transport-layer flows. 469 Note that the use of RTP and RTCP multiplexed onto a single 470 transport-layer flow ensures that there is occasional traffic sent on 471 that port, even if there is no active media traffic. This can be 472 useful to keep NAT bindings alive [RFC6263]. 474 4.6. Reduced Size RTCP 475 RTCP packets are usually sent as compound RTCP packets, and [RFC3550] 476 requires that those compound packets start with an Sender Report (SR) 477 or Receiver Report (RR) packet. When using frequent RTCP feedback 478 messages under the RTP/AVPF Profile [RFC4585] these statistics are 479 not needed in every packet, and unnecessarily increase the mean RTCP 480 packet size. This can limit the frequency at which RTCP packets can 481 be sent within the RTCP bandwidth share. 483 To avoid this problem, [RFC5506] specifies how to reduce the mean 484 RTCP message size and allow for more frequent feedback. Frequent 485 feedback, in turn, is essential to make real-time applications 486 quickly aware of changing network conditions, and to allow them to 487 adapt their transmission and encoding behaviour. Implementations 488 MUST support sending and receiving non-compound RTCP feedback packets 489 [RFC5506]. Use of non-compound RTCP packets MUST be negotiated using 490 the signalling channel. If SDP is used for signalling, this 491 negotiation MUST use the attributes defined in [RFC5506]. For 492 backwards compatibility, implementations are also REQUIRED to support 493 the use of compound RTCP feedback packets if the remote end-point 494 does not agree to the use of non-compound RTCP in the signalling 495 exchange. 497 4.7. Symmetric RTP/RTCP 499 To ease traversal of NAT and firewall devices, implementations are 500 REQUIRED to implement and use Symmetric RTP [RFC4961]. The reason 501 for using symmetric RTP is primarily to avoid issues with NATs and 502 Firewalls by ensuring that the send and receive RTP packet streams, 503 as well as RTCP, are actually bi-directional transport-layer flows. 504 This will keep alive the NAT and firewall pinholes, and help indicate 505 consent that the receive direction is a transport-layer flow the 506 intended recipient actually wants. In addition, it saves resources, 507 specifically ports at the end-points, but also in the network as NAT 508 mappings or firewall state is not unnecessary bloated. The amount of 509 per flow QoS state kept in the network is also reduced. 511 4.8. Choice of RTP Synchronisation Source (SSRC) 513 Implementations are REQUIRED to support signalled RTP synchronisation 514 source (SSRC) identifiers. If SDP is used, this MUST be done using 515 the "a=ssrc:" SDP attribute defined in Section 4.1 and Section 5 of 516 [RFC5576] and the "previous-ssrc" source attribute defined in 517 Section 6.2 of [RFC5576]; other per-SSRC attributes defined in 518 [RFC5576] MAY be supported. 520 While support for signalled SSRC identifiers is mandated, their use 521 in an RTP session is OPTIONAL. Implementations MUST be prepared to 522 accept RTP and RTCP packets using SSRCs that have not been explicitly 523 signalled ahead of time. Implementations MUST support random SSRC 524 assignment, and MUST support SSRC collision detection and resolution, 525 according to [RFC3550]. When using signalled SSRC values, collision 526 detection MUST be performed as described in Section 5 of [RFC5576]. 528 It is often desirable to associate an RTP packet stream with a non- 529 RTP context. For users of the WebRTC API a mapping between SSRCs and 530 MediaStreamTracks are provided per Section 11. For gateways or other 531 usages it is possible to associate an RTP packet stream with an "m=" 532 line in a session description formatted using SDP. If SSRCs are 533 signalled this is straightforward (in SDP the "a=ssrc:" line will be 534 at the media level, allowing a direct association with an "m=" line). 535 If SSRCs are not signalled, the RTP payload type numbers used in an 536 RTP packet stream are often sufficient to associate that packet 537 stream with a signalling context (e.g., if RTP payload type numbers 538 are assigned as described in Section 4.3 of this memo, the RTP 539 payload types used by an RTP packet stream can be compared with 540 values in SDP "a=rtpmap:" lines, which are at the media level in SDP, 541 and so map to an "m=" line). 543 4.9. Generation of the RTCP Canonical Name (CNAME) 545 The RTCP Canonical Name (CNAME) provides a persistent transport-level 546 identifier for an RTP end-point. While the Synchronisation Source 547 (SSRC) identifier for an RTP end-point can change if a collision is 548 detected, or when the RTP application is restarted, its RTCP CNAME is 549 meant to stay unchanged for the duration of a RTCPeerConnection 550 [W3C.WD-webrtc-20130910], so that RTP end-points can be uniquely 551 identified and associated with their RTP packet streams within a set 552 of related RTP sessions. 554 Each RTP end-point MUST have at least one RTCP CNAME, and that RTCP 555 CNAME MUST be unique within the RTCPeerConnection. RTCP CNAMEs 556 identify a particular synchronisation context, i.e., all SSRCs 557 associated with a single RTCP CNAME share a common reference clock. 558 If an end-point has SSRCs that are associated with several 559 unsynchronised reference clocks, and hence different synchronisation 560 contexts, it will need to use multiple RTCP CNAMEs, one for each 561 synchronisation context. 563 Taking the discussion in Section 11 into account, a WebRTC Endpoint 564 MUST NOT use more than one RTCP CNAME in the RTP sessions belonging 565 to single RTCPeerConnection (that is, an RTCPeerConnection forms a 566 synchronisation context). RTP middleboxes MAY generate RTP packet 567 streams associated with more than one RTCP CNAME, to allow them to 568 avoid having to resynchronize media from multiple different end- 569 points part of a multi-party RTP session. 571 The RTP specification [RFC3550] includes guidelines for choosing a 572 unique RTP CNAME, but these are not sufficient in the presence of NAT 573 devices. In addition, long-term persistent identifiers can be 574 problematic from a privacy viewpoint (Section 13). Accordingly, a 575 WebRTC Endpoint MUST generate a new, unique, short-term persistent 576 RTCP CNAME for each RTCPeerConnection, following [RFC7022], with a 577 single exception; if explicitly requested at creation an 578 RTCPeerConnection MAY use the same CNAME as as an existing 579 RTCPeerConnection within their common same-origin context. 581 An WebRTC Endpoint MUST support reception of any CNAME that matches 582 the syntax limitations specified by the RTP specification [RFC3550] 583 and cannot assume that any CNAME will be chosen according to the form 584 suggested above. 586 4.10. Handling of Leap Seconds 588 The guidelines regarding handling of leap seconds to limit their 589 impact on RTP media play-out and synchronization given in [RFC7164] 590 SHOULD be followed. 592 5. WebRTC Use of RTP: Extensions 594 There are a number of RTP extensions that are either needed to obtain 595 full functionality, or extremely useful to improve on the baseline 596 performance, in the WebRTC context. One set of these extensions is 597 related to conferencing, while others are more generic in nature. 598 The following subsections describe the various RTP extensions 599 mandated or suggested for use within WebRTC. 601 5.1. Conferencing Extensions and Topologies 603 RTP is a protocol that inherently supports group communication. 604 Groups can be implemented by having each endpoint send its RTP packet 605 streams to an RTP middlebox that redistributes the traffic, by using 606 a mesh of unicast RTP packet streams between endpoints, or by using 607 an IP multicast group to distribute the RTP packet streams. These 608 topologies can be implemented in a number of ways as discussed in 609 [I-D.ietf-avtcore-rtp-topologies-update]. 611 While the use of IP multicast groups is popular in IPTV systems, the 612 topologies based on RTP middleboxes are dominant in interactive video 613 conferencing environments. Topologies based on a mesh of unicast 614 transport-layer flows to create a common RTP session have not seen 615 widespread deployment to date. Accordingly, WebRTC Endpoints are not 616 expected to support topologies based on IP multicast groups or to 617 support mesh-based topologies, such as a point-to-multipoint mesh 618 configured as a single RTP session (Topo-Mesh in the terminology of 620 [I-D.ietf-avtcore-rtp-topologies-update]). However, a point-to- 621 multipoint mesh constructed using several RTP sessions, implemented 622 in WebRTC using independent RTCPeerConnections 623 [W3C.WD-webrtc-20130910], can be expected to be used in WebRTC, and 624 needs to be supported. 626 WebRTC Endpoints implemented according to this memo are expected to 627 support all the topologies described in 628 [I-D.ietf-avtcore-rtp-topologies-update] where the RTP endpoints send 629 and receive unicast RTP packet streams to and from some peer device, 630 provided that peer can participate in performing congestion control 631 on the RTP packet streams. The peer device could be another RTP 632 endpoint, or it could be an RTP middlebox that redistributes the RTP 633 packet streams to other RTP endpoints. This limitation means that 634 some of the RTP middlebox-based topologies are not suitable for use 635 in WebRTC. Specifically: 637 o Video switching MCUs (Topo-Video-switch-MCU) SHOULD NOT be used, 638 since they make the use of RTCP for congestion control and quality 639 of service reports problematic (see Section 3.8 of 640 [I-D.ietf-avtcore-rtp-topologies-update]). 642 o The Relay-Transport Translator (Topo-PtM-Trn-Translator) topology 643 SHOULD NOT be used because its safe use requires a congestion 644 control algorithm or RTP circuit breaker that handles point to 645 multipoint, which has not yet been standardised. 647 The following topology can be used, however it has some issues worth 648 noting: 650 o Content modifying MCUs with RTCP termination (Topo-RTCP- 651 terminating-MCU) MAY be used. Note that in this RTP Topology, RTP 652 loop detection and identification of active senders is the 653 responsibility of the WebRTC application; since the clients are 654 isolated from each other at the RTP layer, RTP cannot assist with 655 these functions (see section 3.9 of 656 [I-D.ietf-avtcore-rtp-topologies-update]). 658 The RTP extensions described in Section 5.1.1 to Section 5.1.6 are 659 designed to be used with centralised conferencing, where an RTP 660 middlebox (e.g., a conference bridge) receives a participant's RTP 661 packet streams and distributes them to the other participants. These 662 extensions are not necessary for interoperability; an RTP end-point 663 that does not implement these extensions will work correctly, but 664 might offer poor performance. Support for the listed extensions will 665 greatly improve the quality of experience and, to provide a 666 reasonable baseline quality, some of these extensions are mandatory 667 to be supported by WebRTC Endpoints. 669 The RTCP conferencing extensions are defined in Extended RTP Profile 670 for Real-time Transport Control Protocol (RTCP)-Based Feedback (RTP/ 671 AVPF) [RFC4585] and the memo on Codec Control Messages (CCM) in RTP/ 672 AVPF [RFC5104]; they are fully usable by the Secure variant of this 673 profile (RTP/SAVPF) [RFC5124]. 675 5.1.1. Full Intra Request (FIR) 677 The Full Intra Request message is defined in Sections 3.5.1 and 4.3.1 678 of the Codec Control Messages [RFC5104]. It is used to make the 679 mixer request a new Intra picture from a participant in the session. 680 This is used when switching between sources to ensure that the 681 receivers can decode the video or other predictive media encoding 682 with long prediction chains. WebRTC Endpoints that are sending media 683 MUST understand and react to FIR feedback messages they receive, 684 since this greatly improves the user experience when using 685 centralised mixer-based conferencing. Support for sending FIR 686 messages is OPTIONAL. 688 5.1.2. Picture Loss Indication (PLI) 690 The Picture Loss Indication message is defined in Section 6.3.1 of 691 the RTP/AVPF profile [RFC4585]. It is used by a receiver to tell the 692 sending encoder that it lost the decoder context and would like to 693 have it repaired somehow. This is semantically different from the 694 Full Intra Request above as there could be multiple ways to fulfil 695 the request. WebRTC Endpoints that are sending media MUST understand 696 and react to PLI feedback messages as a loss tolerance mechanism. 697 Receivers MAY send PLI messages. 699 5.1.3. Slice Loss Indication (SLI) 701 The Slice Loss Indication message is defined in Section 6.3.2 of the 702 RTP/AVPF profile [RFC4585]. It is used by a receiver to tell the 703 encoder that it has detected the loss or corruption of one or more 704 consecutive macro blocks, and would like to have these repaired 705 somehow. It is RECOMMENDED that receivers generate SLI feedback 706 messages if slices are lost when using a codec that supports the 707 concept of macro blocks. A sender that receives an SLI feedback 708 message SHOULD attempt to repair the lost slice(s). 710 5.1.4. Reference Picture Selection Indication (RPSI) 712 Reference Picture Selection Indication (RPSI) messages are defined in 713 Section 6.3.3 of the RTP/AVPF profile [RFC4585]. Some video encoding 714 standards allow the use of older reference pictures than the most 715 recent one for predictive coding. If such a codec is in use, and if 716 the encoder has learnt that encoder-decoder synchronisation has been 717 lost, then a known as correct reference picture can be used as a base 718 for future coding. The RPSI message allows this to be signalled. 719 Receivers that detect that encoder-decoder synchronisation has been 720 lost SHOULD generate an RPSI feedback message if codec being used 721 supports reference picture selection. A RTP packet stream sender 722 that receives such an RPSI message SHOULD act on that messages to 723 change the reference picture, if it is possible to do so within the 724 available bandwidth constraints, and with the codec being used. 726 5.1.5. Temporal-Spatial Trade-off Request (TSTR) 728 The temporal-spatial trade-off request and notification are defined 729 in Sections 3.5.2 and 4.3.2 of [RFC5104]. This request can be used 730 to ask the video encoder to change the trade-off it makes between 731 temporal and spatial resolution, for example to prefer high spatial 732 image quality but low frame rate. Support for TSTR requests and 733 notifications is OPTIONAL. 735 5.1.6. Temporary Maximum Media Stream Bit Rate Request (TMMBR) 737 The TMMBR feedback message is defined in Sections 3.5.4 and 4.2.1 of 738 the Codec Control Messages [RFC5104]. This request and its 739 notification message are used by a media receiver to inform the 740 sending party that there is a current limitation on the amount of 741 bandwidth available to this receiver. This can be various reasons 742 for this: for example, an RTP mixer can use this message to limit the 743 media rate of the sender being forwarded by the mixer (without doing 744 media transcoding) to fit the bottlenecks existing towards the other 745 session participants. WebRTC Endpoints that are sending media are 746 REQUIRED to implement support for TMMBR messages, and MUST follow 747 bandwidth limitations set by a TMMBR message received for their SSRC. 748 The sending of TMMBR requests is OPTIONAL. 750 5.2. Header Extensions 752 The RTP specification [RFC3550] provides the capability to include 753 RTP header extensions containing in-band data, but the format and 754 semantics of the extensions are poorly specified. The use of header 755 extensions is OPTIONAL in WebRTC, but if they are used, they MUST be 756 formatted and signalled following the general mechanism for RTP 757 header extensions defined in [RFC5285], since this gives well-defined 758 semantics to RTP header extensions. 760 As noted in [RFC5285], the requirement from the RTP specification 761 that header extensions are "designed so that the header extension may 762 be ignored" [RFC3550] stands. To be specific, header extensions MUST 763 only be used for data that can safely be ignored by the recipient 764 without affecting interoperability, and MUST NOT be used when the 765 presence of the extension has changed the form or nature of the rest 766 of the packet in a way that is not compatible with the way the stream 767 is signalled (e.g., as defined by the payload type). Valid examples 768 of RTP header extensions might include metadata that is additional to 769 the usual RTP information, but that can safely be ignored without 770 compromising interoperability. 772 5.2.1. Rapid Synchronisation 774 Many RTP sessions require synchronisation between audio, video, and 775 other content. This synchronisation is performed by receivers, using 776 information contained in RTCP SR packets, as described in the RTP 777 specification [RFC3550]. This basic mechanism can be slow, however, 778 so it is RECOMMENDED that the rapid RTP synchronisation extensions 779 described in [RFC6051] be implemented in addition to RTCP SR-based 780 synchronisation. The rapid synchronisation extensions use the 781 general RTP header extension mechanism [RFC5285], which requires 782 signalling, but are otherwise backwards compatible. 784 5.2.2. Client-to-Mixer Audio Level 786 The Client to Mixer Audio Level extension [RFC6464] is an RTP header 787 extension used by an endpoint to inform a mixer about the level of 788 audio activity in the packet to which the header is attached. This 789 enables an RTP middlebox to make mixing or selection decisions 790 without decoding or detailed inspection of the payload, reducing the 791 complexity in some types of mixers. It can also save decoding 792 resources in receivers, which can choose to decode only the most 793 relevant RTP packet streams based on audio activity levels. 795 The Client-to-Mixer Audio Level [RFC6464] header extension is 796 RECOMMENDED to be implemented. If this header extension is 797 implemented, it is REQUIRED that implementations are capable of 798 encrypting the header extension according to [RFC6904] since the 799 information contained in these header extensions can be considered 800 sensitive. The use of this encryption is RECOMMENDED, however usage 801 of the encryption can be explicitly disabled through API or 802 signalling. 804 5.2.3. Mixer-to-Client Audio Level 806 The Mixer to Client Audio Level header extension [RFC6465] provides 807 an endpoint with the audio level of the different sources mixed into 808 a common source stream by a RTP mixer. This enables a user interface 809 to indicate the relative activity level of each session participant, 810 rather than just being included or not based on the CSRC field. This 811 is a pure optimisation of non critical functions, and is hence 812 OPTIONAL to implement. If this header extension is implemented, it 813 is REQUIRED that implementations are capable of encrypting the header 814 extension according to [RFC6904] since the information contained in 815 these header extensions can be considered sensitive. It is further 816 RECOMMENDED that this encryption is used, unless the encryption has 817 been explicitly disabled through API or signalling. 819 6. WebRTC Use of RTP: Improving Transport Robustness 821 There are tools that can make RTP packet streams robust against 822 packet loss and reduce the impact of loss on media quality. However, 823 they generally add some overhead compared to a non-robust stream. 824 The overhead needs to be considered, and the aggregate bit-rate MUST 825 be rate controlled to avoid causing network congestion (see 826 Section 7). As a result, improving robustness might require a lower 827 base encoding quality, but has the potential to deliver that quality 828 with fewer errors. The mechanisms described in the following sub- 829 sections can be used to improve tolerance to packet loss. 831 6.1. Negative Acknowledgements and RTP Retransmission 833 As a consequence of supporting the RTP/SAVPF profile, implementations 834 can send negative acknowledgements (NACKs) for RTP data packets 835 [RFC4585]. This feedback can be used to inform a sender of the loss 836 of particular RTP packets, subject to the capacity limitations of the 837 RTCP feedback channel. A sender can use this information to optimise 838 the user experience by adapting the media encoding to compensate for 839 known lost packets. 841 RTP packet stream senders are REQUIRED to understand the Generic NACK 842 message defined in Section 6.2.1 of [RFC4585], but MAY choose to 843 ignore some or all of this feedback (following Section 4.2 of 844 [RFC4585]). Receivers MAY send NACKs for missing RTP packets. 845 Guidelines on when to send NACKs are provided in [RFC4585]. It is 846 not expected that a receiver will send a NACK for every lost RTP 847 packet, rather it needs to consider the cost of sending NACK 848 feedback, and the importance of the lost packet, to make an informed 849 decision on whether it is worth telling the sender about a packet 850 loss event. 852 The RTP Retransmission Payload Format [RFC4588] offers the ability to 853 retransmit lost packets based on NACK feedback. Retransmission needs 854 to be used with care in interactive real-time applications to ensure 855 that the retransmitted packet arrives in time to be useful, but can 856 be effective in environments with relatively low network RTT (an RTP 857 sender can estimate the RTT to the receivers using the information in 858 RTCP SR and RR packets, as described at the end of Section 6.4.1 of 859 [RFC3550]). The use of retransmissions can also increase the forward 860 RTP bandwidth, and can potentially caused increased packet loss if 861 the original packet loss was caused by network congestion. Note, 862 however, that retransmission of an important lost packet to repair 863 decoder state can have lower cost than sending a full intra frame. 864 It is not appropriate to blindly retransmit RTP packets in response 865 to a NACK. The importance of lost packets and the likelihood of them 866 arriving in time to be useful needs to be considered before RTP 867 retransmission is used. 869 Receivers are REQUIRED to implement support for RTP retransmission 870 packets [RFC4588]. Senders MAY send RTP retransmission packets in 871 response to NACKs if the RTP retransmission payload format has been 872 negotiated for the session, and if the sender believes it is useful 873 to send a retransmission of the packet(s) referenced in the NACK. An 874 RTP sender does not need to retransmit every NACKed packet. 876 6.2. Forward Error Correction (FEC) 878 The use of Forward Error Correction (FEC) can provide an effective 879 protection against some degree of packet loss, at the cost of steady 880 bandwidth overhead. There are several FEC schemes that are defined 881 for use with RTP. Some of these schemes are specific to a particular 882 RTP payload format, others operate across RTP packets and can be used 883 with any payload format. It needs to be noted that using redundant 884 encoding or FEC will lead to increased play out delay, which needs to 885 be considered when choosing the redundancy or FEC formats and their 886 respective parameters. 888 If an RTP payload format negotiated for use in a RTCPeerConnection 889 supports redundant transmission or FEC as a standard feature of that 890 payload format, then that support MAY be used in the 891 RTCPeerConnection, subject to any appropriate signalling. 893 There are several block-based FEC schemes that are designed for use 894 with RTP independent of the chosen RTP payload format. At the time 895 of this writing there is no consensus on which, if any, of these FEC 896 schemes is appropriate for use in WebRTC. Accordingly, this memo 897 makes no recommendation on the choice of block-based FEC for WebRTC 898 use. 900 7. WebRTC Use of RTP: Rate Control and Media Adaptation 902 WebRTC will be used in heterogeneous network environments using a 903 variety set of link technologies, including both wired and wireless 904 links, to interconnect potentially large groups of users around the 905 world. As a result, the network paths between users can have widely 906 varying one-way delays, available bit-rates, load levels, and traffic 907 mixtures. Individual end-points can send one or more RTP packet 908 streams to each participant, and there can be several participants. 910 Each of these RTP packet streams can contain different types of 911 media, and the type of media, bit rate, and number of RTP packet 912 streams as well as transport-layer flows can be highly asymmetric. 913 Non-RTP traffic can share the network paths with RTP transport-layer 914 flows. Since the network environment is not predictable or stable, 915 WebRTC Endpoints MUST ensure that the RTP traffic they generate can 916 adapt to match changes in the available network capacity. 918 The quality of experience for users of WebRTC is very dependent on 919 effective adaptation of the media to the limitations of the network. 920 End-points have to be designed so they do not transmit significantly 921 more data than the network path can support, except for very short 922 time periods, otherwise high levels of network packet loss or delay 923 spikes will occur, causing media quality degradation. The limiting 924 factor on the capacity of the network path might be the link 925 bandwidth, or it might be competition with other traffic on the link 926 (this can be non-WebRTC traffic, traffic due to other WebRTC flows, 927 or even competition with other WebRTC flows in the same session). 929 An effective media congestion control algorithm is therefore an 930 essential part of the WebRTC framework. However, at the time of this 931 writing, there is no standard congestion control algorithm that can 932 be used for interactive media applications such as WebRTC's flows. 933 Some requirements for congestion control algorithms for 934 RTCPeerConnections are discussed in [I-D.ietf-rmcat-cc-requirements]. 935 A future version of this memo will mandate the use of a congestion 936 control algorithm that satisfies these requirements. 938 7.1. Boundary Conditions and Circuit Breakers 940 WebRTC Endpoints MUST implement the RTP circuit breaker algorithm 941 that is described in [I-D.ietf-avtcore-rtp-circuit-breakers]. The 942 RTP circuit breaker is designed to enable applications to recognise 943 and react to situations of extreme network congestion. However, 944 since the RTP circuit breaker might not be triggered until congestion 945 becomes extreme, it cannot be considered a substitute for congestion 946 control, and applications MUST also implement congestion control to 947 allow them to adapt to changes in network capacity. Any future RTP 948 congestion control algorithms are expected to operate within the 949 envelope allowed by the circuit breaker. 951 The session establishment signalling will also necessarily establish 952 boundaries to which the media bit-rate will conform. The choice of 953 media codecs provides upper- and lower-bounds on the supported bit- 954 rates that the application can utilise to provide useful quality, and 955 the packetisation choices that exist. In addition, the signalling 956 channel can establish maximum media bit-rate boundaries using, for 957 example, the SDP "b=AS:" or "b=CT:" lines and the RTP/AVPF Temporary 958 Maximum Media Stream Bit Rate (TMMBR) Requests (see Section 5.1.6 of 959 this memo). Signalled bandwidth limitations, such as SDP "b=AS:" or 960 "b=CT:" lines received from the peer, MUST be followed when sending 961 RTP packet streams. A WebRTC Endpoint receiving media SHOULD signal 962 its bandwidth limitations, these limitations have to be based on 963 known bandwidth limitations, for example the capacity of the edge 964 links. 966 7.2. Congestion Control Interoperability and Legacy Systems 968 There are legacy RTP implementations that do not implement RTCP, and 969 hence do not provide any congestion feedback. Congestion control 970 cannot be performed with these end-points. WebRTC Endpoints that 971 need to interwork with such end-points MUST limit their transmission 972 to a low rate, equivalent to a VoIP call using a low bandwidth codec, 973 that is unlikely to cause any significant congestion. 975 When interworking with legacy implementations that support RTCP using 976 the RTP/AVP profile [RFC3551], congestion feedback is provided in 977 RTCP RR packets every few seconds. Implementations that have to 978 interwork with such end-points MUST ensure that they keep within the 979 RTP circuit breaker [I-D.ietf-avtcore-rtp-circuit-breakers] 980 constraints to limit the congestion they can cause. 982 If a legacy end-point supports RTP/AVPF, this enables negotiation of 983 important parameters for frequent reporting, such as the "trr-int" 984 parameter, and the possibility that the end-point supports some 985 useful feedback format for congestion control purpose such as TMMBR 986 [RFC5104]. Implementations that have to interwork with such end- 987 points MUST ensure that they stay within the RTP circuit breaker 988 [I-D.ietf-avtcore-rtp-circuit-breakers] constraints to limit the 989 congestion they can cause, but might find that they can achieve 990 better congestion response depending on the amount of feedback that 991 is available. 993 With proprietary congestion control algorithms issues can arise when 994 different algorithms and implementations interact in a communication 995 session. If the different implementations have made different 996 choices in regards to the type of adaptation, for example one sender 997 based, and one receiver based, then one could end up in situation 998 where one direction is dual controlled, when the other direction is 999 not controlled. This memo cannot mandate behaviour for proprietary 1000 congestion control algorithms, but implementations that use such 1001 algorithms ought to be aware of this issue, and try to ensure that 1002 effective congestion control is negotiated for media flowing in both 1003 directions. If the IETF were to standardise both sender- and 1004 receiver-based congestion control algorithms for WebRTC traffic in 1005 the future, the issues of interoperability, control, and ensuring 1006 that both directions of media flow are congestion controlled would 1007 also need to be considered. 1009 8. WebRTC Use of RTP: Performance Monitoring 1011 As described in Section 4.1, implementations are REQUIRED to generate 1012 RTCP Sender Report (SR) and Reception Report (RR) packets relating to 1013 the RTP packet streams they send and receive. These RTCP reports can 1014 be used for performance monitoring purposes, since they include basic 1015 packet loss and jitter statistics. 1017 A large number of additional performance metrics are supported by the 1018 RTCP Extended Reports (XR) framework [RFC3611][RFC6792]. At the time 1019 of this writing, it is not clear what extended metrics are suitable 1020 for use in WebRTC, so there is no requirement that implementations 1021 generate RTCP XR packets. However, implementations that can use 1022 detailed performance monitoring data MAY generate RTCP XR packets as 1023 appropriate; the use of such packets SHOULD be signalled in advance. 1025 9. WebRTC Use of RTP: Future Extensions 1027 It is possible that the core set of RTP protocols and RTP extensions 1028 specified in this memo will prove insufficient for the future needs 1029 of WebRTC. In this case, future updates to this memo MUST be made 1030 following the Guidelines for Writers of RTP Payload Format 1031 Specifications [RFC2736], How to Write an RTP Payload Format 1032 [I-D.ietf-payload-rtp-howto] and Guidelines for Extending the RTP 1033 Control Protocol [RFC5968], and SHOULD take into account any future 1034 guidelines for extending RTP and related protocols that have been 1035 developed. 1037 Authors of future extensions are urged to consider the wide range of 1038 environments in which RTP is used when recommending extensions, since 1039 extensions that are applicable in some scenarios can be problematic 1040 in others. Where possible, the WebRTC framework will adopt RTP 1041 extensions that are of general utility, to enable easy implementation 1042 of a gateway to other applications using RTP, rather than adopt 1043 mechanisms that are narrowly targeted at specific WebRTC use cases. 1045 10. Signalling Considerations 1047 RTP is built with the assumption that an external signalling channel 1048 exists, and can be used to configure RTP sessions and their features. 1049 The basic configuration of an RTP session consists of the following 1050 parameters: 1052 RTP Profile: The name of the RTP profile to be used in session. The 1053 RTP/AVP [RFC3551] and RTP/AVPF [RFC4585] profiles can interoperate 1054 on basic level, as can their secure variants RTP/SAVP [RFC3711] 1055 and RTP/SAVPF [RFC5124]. The secure variants of the profiles do 1056 not directly interoperate with the non-secure variants, due to the 1057 presence of additional header fields for authentication in SRTP 1058 packets and cryptographic transformation of the payload. WebRTC 1059 requires the use of the RTP/SAVPF profile, and this MUST be 1060 signalled. Interworking functions might transform this into the 1061 RTP/SAVP profile for a legacy use case, by indicating to the 1062 WebRTC Endpoint that the RTP/SAVPF is used and configuring a trr- 1063 int value of 4 seconds. 1065 Transport Information: Source and destination IP address(s) and 1066 ports for RTP and RTCP MUST be signalled for each RTP session. In 1067 WebRTC these transport addresses will be provided by ICE [RFC5245] 1068 that signals candidates and arrives at nominated candidate address 1069 pairs. If RTP and RTCP multiplexing [RFC5761] is to be used, such 1070 that a single port, i.e. transport-layer flow, is used for RTP 1071 and RTCP flows, this MUST be signalled (see Section 4.5). 1073 RTP Payload Types, media formats, and format parameters: The mapping 1074 between media type names (and hence the RTP payload formats to be 1075 used), and the RTP payload type numbers MUST be signalled. Each 1076 media type MAY also have a number of media type parameters that 1077 MUST also be signalled to configure the codec and RTP payload 1078 format (the "a=fmtp:" line from SDP). Section 4.3 of this memo 1079 discusses requirements for uniqueness of payload types. 1081 RTP Extensions: The use of any additional RTP header extensions and 1082 RTCP packet types, including any necessary parameters, MUST be 1083 signalled. This signalling is to ensure that a WebRTC Endpoint's 1084 behaviour, especially when sending, of any extensions is 1085 predictable and consistent. For robustness, and for compatibility 1086 with non-WebRTC systems that might be connected to a WebRTC 1087 session via a gateway, implementations are REQUIRED to ignore 1088 unknown RTCP packets and RTP header extensions (see also 1089 Section 4.1). 1091 RTCP Bandwidth: Support for exchanging RTCP Bandwidth values to the 1092 end-points will be necessary. This SHALL be done as described in 1093 "Session Description Protocol (SDP) Bandwidth Modifiers for RTP 1094 Control Protocol (RTCP) Bandwidth" [RFC3556] if using SDP, or 1095 something semantically equivalent. This also ensures that the 1096 end-points have a common view of the RTCP bandwidth. A common 1097 RTCP bandwidth is important as a too different view of the 1098 bandwidths can lead to failure to interoperate. 1100 These parameters are often expressed in SDP messages conveyed within 1101 an offer/answer exchange. RTP does not depend on SDP or on the offer 1102 /answer model, but does require all the necessary parameters to be 1103 agreed upon, and provided to the RTP implementation. Note that in 1104 WebRTC it will depend on the signalling model and API how these 1105 parameters need to be configured but they will be need to either be 1106 set in the API or explicitly signalled between the peers. 1108 11. WebRTC API Considerations 1110 The WebRTC API [W3C.WD-webrtc-20130910] and the Media Capture and 1111 Streams API [W3C.WD-mediacapture-streams-20130903] defines and uses 1112 the concept of a MediaStream that consists of zero or more 1113 MediaStreamTracks. A MediaStreamTrack is an individual stream of 1114 media from any type of media source like a microphone or a camera, 1115 but also conceptual sources, like a audio mix or a video composition, 1116 are possible. The MediaStreamTracks within a MediaStream need to be 1117 possible to play out synchronised. 1119 A MediaStreamTrack's realisation in RTP in the context of an 1120 RTCPeerConnection consists of a source packet stream identified with 1121 an SSRC within an RTP session part of the RTCPeerConnection. The 1122 MediaStreamTrack can also result in additional packet streams, and 1123 thus SSRCs, in the same RTP session. These can be dependent packet 1124 streams from scalable encoding of the source stream associated with 1125 the MediaStreamTrack, if such a media encoder is used. They can also 1126 be redundancy packet streams, these are created when applying Forward 1127 Error Correction (Section 6.2) or RTP retransmission (Section 6.1) to 1128 the source packet stream. 1130 It is important to note that the same media source can be feeding 1131 multiple MediaStreamTracks. As different sets of constraints or 1132 other parameters can be applied to the MediaStreamTrack, each 1133 MediaStreamTrack instance added to a RTCPeerConnection SHALL result 1134 in an independent source packet stream, with its own set of 1135 associated packet streams, and thus different SSRC(s). It will 1136 depend on applied constraints and parameters if the source stream and 1137 the encoding configuration will be identical between different 1138 MediaStreamTracks sharing the same media source. If the encoding 1139 parameters and constraints are the same, an implementation could 1140 choose to use only one encoded stream to create the different RTP 1141 packet streams. Note that such optimisations would need to take into 1142 account that the constraints for one of the MediaStreamTracks can at 1143 any moment change, meaning that the encoding configurations might no 1144 longer be identical and two different encoder instances would then be 1145 needed. 1147 The same MediaStreamTrack can also be included in multiple 1148 MediaStreams, thus multiple sets of MediaStreams can implicitly need 1149 to use the same synchronisation base. To ensure that this works in 1150 all cases, and does not force an end-point to to disrupt the media by 1151 changing synchronisation base and CNAME during delivery of any 1152 ongoing packet streams, all MediaStreamTracks and their associated 1153 SSRCs originating from the same end-point need to be sent using the 1154 same CNAME within one RTCPeerConnection. This is motivating the 1155 discussion in Section 4.9 to only use a single CNAME. 1157 The requirement on using the same CNAME for all SSRCs that 1158 originate from the same end-point, does not require a middlebox 1159 that forwards traffic from multiple end-points to only use a 1160 single CNAME. 1162 Different CNAMEs normally need to be used for different 1163 RTCPeerConnection instances, as specified in Section 4.9. Having two 1164 communication sessions with the same CNAME could enable tracking of a 1165 user or device across different services (see Section 4.4.1 of 1166 [I-D.ietf-rtcweb-security] for details). A web application can 1167 request that the CNAMEs used in different RTCPeerConnections (within 1168 a same-orign context) be the same, this allows for synchronization of 1169 the endpoint's RTP packet streams across the different 1170 RTCPeerConnections. 1172 Note: this doesn't result in a tracking issue, since the creation 1173 of matching CNAMEs depends on existing tracking. 1175 The above will currently force a WebRTC Endpoint that receives a 1176 MediaStreamTrack on one RTCPeerConnection and adds it as an outgoing 1177 on any RTCPeerConnection to perform resynchronisation of the stream. 1178 This, as the sending party needs to change the CNAME to the one it 1179 uses, which implies that the sender has to use a local system clock 1180 as timebase for the synchronisation. Thus, the relative relation 1181 between the timebase of the incoming stream and the system sending 1182 out needs to defined. This relation also needs monitoring for clock 1183 drift and likely adjustments of the synchronisation. The sending 1184 entity is also responsible for congestion control for its sent 1185 streams. In cases of packet loss the loss of incoming data also 1186 needs to be handled. This leads to the observation that the method 1187 that is least likely to cause issues or interruptions in the outgoing 1188 source packet stream is a model of full decoding, including repair 1189 etc., followed by encoding of the media again into the outgoing 1190 packet stream. Optimisations of this method is clearly possible and 1191 implementation specific. 1193 A WebRTC Endpoint MUST support receiving multiple MediaStreamTracks, 1194 where each of different MediaStreamTracks (and their sets of 1195 associated packet streams) uses different CNAMEs. However, 1196 MediaStreamTracks that are received with different CNAMEs have no 1197 defined synchronisation. 1199 Note: The motivation for supporting reception of multiple CNAMEs 1200 is to allow for forward compatibility with any future changes that 1201 enables more efficient stream handling when end-points relay/ 1202 forward streams. It also ensures that end-points can interoperate 1203 with certain types of multi-stream middleboxes or end-points that 1204 are not WebRTC. 1206 The binding between the WebRTC MediaStreams, MediaStreamTracks and 1207 the SSRC is done as specified in "Cross Session Stream Identification 1208 in the Session Description Protocol" [I-D.ietf-mmusic-msid]. This 1209 document [I-D.ietf-mmusic-msid] also defines, in section 4.1, how to 1210 map unknown source packet stream SSRCs to MediaStreamTracks and 1211 MediaStreams. This later is relevant to handle some cases of legacy 1212 interop. Commonly the RTP Payload Type of any incoming packets will 1213 reveal if the packet stream is a source stream or a redundancy or 1214 dependent packet stream. The association to the correct source 1215 packet stream depends on the payload format in use for the packet 1216 stream. 1218 Finally this specification puts a requirement on the WebRTC API to 1219 realize a method for determining the CSRC list (Section 4.1) as well 1220 as the Mixer-to-Client audio levels (Section 5.2.3) (when supported) 1221 and the basic requirements for this is further discussed in 1222 Section 12.2.1. 1224 12. RTP Implementation Considerations 1226 The following discussion provides some guidance on the implementation 1227 of the RTP features described in this memo. The focus is on a WebRTC 1228 Endpoint implementation perspective, and while some mention is made 1229 of the behaviour of middleboxes, that is not the focus of this memo. 1231 12.1. Configuration and Use of RTP Sessions 1233 A WebRTC Endpoint will be a simultaneous participant in one or more 1234 RTP sessions. Each RTP session can convey multiple media sources, 1235 and can include media data from multiple end-points. In the 1236 following, some ways in which WebRTC Endpoints can configure and use 1237 RTP sessions is outlined. 1239 12.1.1. Use of Multiple Media Sources Within an RTP Session 1241 RTP is a group communication protocol, and every RTP session can 1242 potentially contain multiple RTP packet streams. There are several 1243 reasons why this might be desirable: 1245 Multiple media types: Outside of WebRTC, it is common to use one RTP 1246 session for each type of media sources (e.g., one RTP session for 1247 audio sources and one for video sources, each sent over different 1248 transport layer flows). However, to reduce the number of UDP 1249 ports used, the default in WebRTC is to send all types of media in 1250 a single RTP session, as described in Section 4.4, using RTP and 1251 RTCP multiplexing (Section 4.5) to further reduce the number of 1252 UDP ports needed. This RTP session then uses only one bi- 1253 directional transport-layer flow, but will contain multiple RTP 1254 packet streams, each containing a different type of media. A 1255 common example might be an end-point with a camera and microphone 1256 that sends two RTP packet streams, one video and one audio, into a 1257 single RTP session. 1259 Multiple Capture Devices: A WebRTC Endpoint might have multiple 1260 cameras, microphones, or other media capture devices, and so might 1261 want to generate several RTP packet streams of the same media 1262 type. Alternatively, it might want to send media from a single 1263 capture device in several different formats or quality settings at 1264 once. Both can result in a single end-point sending multiple RTP 1265 packet streams of the same media type into a single RTP session at 1266 the same time. 1268 Associated Repair Data: An end-point might send a RTP packet stream 1269 that is somehow associated with another stream. For example, it 1270 might send an RTP packet stream that contains FEC or 1271 retransmission data relating to another stream. Some RTP payload 1272 formats send this sort of associated repair data as part of the 1273 source packet stream, while others send it as a separate packet 1274 stream. 1276 Layered or Multiple Description Coding: An end-point can use a 1277 layered media codec, for example H.264 SVC, or a multiple 1278 description codec, that generates multiple RTP packet streams, 1279 each with a distinct RTP SSRC, within a single RTP session. 1281 RTP Mixers, Translators, and Other Middleboxes: An RTP session, in 1282 the WebRTC context, is a point-to-point association between an 1283 end-point and some other peer device, where those devices share a 1284 common SSRC space. The peer device might be another WebRTC 1285 Endpoint, or it might be an RTP mixer, translator, or some other 1286 form of media processing middlebox. In the latter cases, the 1287 middlebox might send mixed or relayed RTP streams from several 1288 participants, that the WebRTC Endpoint will need to render. Thus, 1289 even though a WebRTC Endpoint might only be a member of a single 1290 RTP session, the peer device might be extending that RTP session 1291 to incorporate other end-points. WebRTC is a group communication 1292 environment and end-points need to be capable of receiving, 1293 decoding, and playing out multiple RTP packet streams at once, 1294 even in a single RTP session. 1296 12.1.2. Use of Multiple RTP Sessions 1298 In addition to sending and receiving multiple RTP packet streams 1299 within a single RTP session, a WebRTC Endpoint might participate in 1300 multiple RTP sessions. There are several reasons why a WebRTC 1301 Endpoint might choose to do this: 1303 To interoperate with legacy devices: The common practice in the non- 1304 WebRTC world is to send different types of media in separate RTP 1305 sessions, for example using one RTP session for audio and another 1306 RTP session, on a separate transport layer flow, for video. All 1307 WebRTC Endpoints need to support the option of sending different 1308 types of media on different RTP sessions, so they can interwork 1309 with such legacy devices. This is discussed further in 1310 Section 4.4. 1312 To provide enhanced quality of service: Some network-based quality 1313 of service mechanisms operate on the granularity of transport 1314 layer flows. If it is desired to use these mechanisms to provide 1315 differentiated quality of service for some RTP packet streams, 1316 then those RTP packet streams need to be sent in a separate RTP 1317 session using a different transport-layer flow, and with 1318 appropriate quality of service marking. This is discussed further 1319 in Section 12.1.3. 1321 To separate media with different purposes: An end-point might want 1322 to send RTP packet streams that have different purposes on 1323 different RTP sessions, to make it easy for the peer device to 1324 distinguish them. For example, some centralised multiparty 1325 conferencing systems display the active speaker in high 1326 resolution, but show low resolution "thumbnails" of other 1327 participants. Such systems might configure the end-points to send 1328 simulcast high- and low-resolution versions of their video using 1329 separate RTP sessions, to simplify the operation of the RTP 1330 middlebox. In the WebRTC context this is currently possible by 1331 establishing multiple WebRTC MediaStreamTracks that have the same 1332 media source in one (or more) RTCPeerConnection. Each 1333 MediaStreamTrack is then configured to deliver a particular media 1334 quality and thus media bit-rate, and will produce an independently 1335 encoded version with the codec parameters agreed specifically in 1336 the context of that RTCPeerConnection. The RTP middlebox can 1337 distinguish packets corresponding to the low- and high-resolution 1338 streams by inspecting their SSRC, RTP payload type, or some other 1339 information contained in RTP payload, RTP header extension or RTCP 1340 packets, but it can be easier to distinguish the RTP packet 1341 streams if they arrive on separate RTP sessions on separate 1342 transport-layer flows. 1344 To directly connect with multiple peers: A multi-party conference 1345 does not need to use an RTP middlebox. Rather, a multi-unicast 1346 mesh can be created, comprising several distinct RTP sessions, 1347 with each participant sending RTP traffic over a separate RTP 1348 session (that is, using an independent RTCPeerConnection object) 1349 to every other participant, as shown in Figure 1. This topology 1350 has the benefit of not requiring an RTP middlebox node that is 1351 trusted to access and manipulate the media data. The downside is 1352 that it increases the used bandwidth at each sender by requiring 1353 one copy of the RTP packet streams for each participant that are 1354 part of the same session beyond the sender itself. 1356 +---+ +---+ 1357 | A |<--->| B | 1358 +---+ +---+ 1359 ^ ^ 1360 \ / 1361 \ / 1362 v v 1363 +---+ 1364 | C | 1365 +---+ 1367 Figure 1: Multi-unicast using several RTP sessions 1369 The multi-unicast topology could also be implemented as a single 1370 RTP session, spanning multiple peer-to-peer transport layer 1371 connections, or as several pairwise RTP sessions, one between each 1372 pair of peers. To maintain a coherent mapping between the 1373 relation between RTP sessions and RTCPeerConnection objects it is 1374 recommend that this is implemented as several individual RTP 1375 sessions. The only downside is that end-point A will not learn of 1376 the quality of any transmission happening between B and C, since 1377 it will not see RTCP reports for the RTP session between B and C, 1378 whereas it would it all three participants were part of a single 1379 RTP session. Experience with the Mbone tools (experimental RTP- 1380 based multicast conferencing tools from the late 1990s) has showed 1381 that RTCP reception quality reports for third parties can be 1382 presented to users in a way that helps them understand asymmetric 1383 network problems, and the approach of using separate RTP sessions 1384 prevents this. However, an advantage of using separate RTP 1385 sessions is that it enables using different media bit-rates and 1386 RTP session configurations between the different peers, thus not 1387 forcing B to endure the same quality reductions if there are 1388 limitations in the transport from A to C as C will. It is 1389 believed that these advantages outweigh the limitations in 1390 debugging power. 1392 To indirectly connect with multiple peers: A common scenario in 1393 multi-party conferencing is to create indirect connections to 1394 multiple peers, using an RTP mixer, translator, or some other type 1395 of RTP middlebox. Figure 2 outlines a simple topology that might 1396 be used in a four-person centralised conference. The middlebox 1397 acts to optimise the transmission of RTP packet streams from 1398 certain perspectives, either by only sending some of the received 1399 RTP packet stream to any given receiver, or by providing a 1400 combined RTP packet stream out of a set of contributing streams. 1402 +---+ +-------------+ +---+ 1403 | A |<---->| |<---->| B | 1404 +---+ | RTP mixer, | +---+ 1405 | translator, | 1406 | or other | 1407 +---+ | middlebox | +---+ 1408 | C |<---->| |<---->| D | 1409 +---+ +-------------+ +---+ 1411 Figure 2: RTP mixer with only unicast paths 1413 There are various methods of implementation for the middlebox. If 1414 implemented as a standard RTP mixer or translator, a single RTP 1415 session will extend across the middlebox and encompass all the 1416 end-points in one multi-party session. Other types of middlebox 1417 might use separate RTP sessions between each end-point and the 1418 middlebox. A common aspect is that these RTP middleboxes can use 1419 a number of tools to control the media encoding provided by a 1420 WebRTC Endpoint. This includes functions like requesting the 1421 breaking of the encoding chain and have the encoder produce a so 1422 called Intra frame. Another is limiting the bit-rate of a given 1423 stream to better suit the mixer view of the multiple down-streams. 1424 Others are controlling the most suitable frame-rate, picture 1425 resolution, the trade-off between frame-rate and spatial quality. 1426 The middlebox has the responsibility to correctly perform 1427 congestion control, source identification, manage synchronisation 1428 while providing the application with suitable media optimisations. 1429 The middlebox also has to be a trusted node when it comes to 1430 security, since it manipulates either the RTP header or the media 1431 itself (or both) received from one end-point, before sending it on 1432 towards the end-point(s), thus they need to be able to decrypt and 1433 then re-encrypt the RTP packet stream before sending it out. 1435 RTP Mixers can create a situation where an end-point experiences a 1436 situation in-between a session with only two end-points and 1437 multiple RTP sessions. Mixers are expected to not forward RTCP 1438 reports regarding RTP packet streams across themselves. This is 1439 due to the difference in the RTP packet streams provided to the 1440 different end-points. The original media source lacks information 1441 about a mixer's manipulations prior to sending it the different 1442 receivers. This scenario also results in that an end-point's 1443 feedback or requests goes to the mixer. When the mixer can't act 1444 on this by itself, it is forced to go to the original media source 1445 to fulfil the receivers request. This will not necessarily be 1446 explicitly visible any RTP and RTCP traffic, but the interactions 1447 and the time to complete them will indicate such dependencies. 1449 Providing source authentication in multi-party scenarios is a 1450 challenge. In the mixer-based topologies, end-points source 1451 authentication is based on, firstly, verifying that media comes 1452 from the mixer by cryptographic verification and, secondly, trust 1453 in the mixer to correctly identify any source towards the end- 1454 point. In RTP sessions where multiple end-points are directly 1455 visible to an end-point, all end-points will have knowledge about 1456 each others' master keys, and can thus inject packets claimed to 1457 come from another end-point in the session. Any node performing 1458 relay can perform non-cryptographic mitigation by preventing 1459 forwarding of packets that have SSRC fields that came from other 1460 end-points before. For cryptographic verification of the source, 1461 SRTP would require additional security mechanisms, for example 1462 TESLA for SRTP [RFC4383], that are not part of the base WebRTC 1463 standards. 1465 To forward media between multiple peers: It is sometimes desirable 1466 for an end-point that receives an RTP packet stream to be able to 1467 forward that RTP packet stream to a third party. The are some 1468 obvious security and privacy implications in supporting this, but 1469 also potential uses. This is supported in the W3C API by taking 1470 the received and decoded media and using it as media source that 1471 is re-encoding and transmitted as a new stream. 1473 At the RTP layer, media forwarding acts as a back-to-back RTP 1474 receiver and RTP sender. The receiving side terminates the RTP 1475 session and decodes the media, while the sender side re-encodes 1476 and transmits the media using an entirely separate RTP session. 1477 The original sender will only see a single receiver of the media, 1478 and will not be able to tell that forwarding is happening based on 1479 RTP-layer information since the RTP session that is used to send 1480 the forwarded media is not connected to the RTP session on which 1481 the media was received by the node doing the forwarding. 1483 The end-point that is performing the forwarding is responsible for 1484 producing an RTP packet stream suitable for onwards transmission. 1485 The outgoing RTP session that is used to send the forwarded media 1486 is entirely separate to the RTP session on which the media was 1487 received. This will require media transcoding for congestion 1488 control purpose to produce a suitable bit-rate for the outgoing 1489 RTP session, reducing media quality and forcing the forwarding 1490 end-point to spend the resource on the transcoding. The media 1491 transcoding does result in a separation of the two different legs 1492 removing almost all dependencies, and allowing the forwarding end- 1493 point to optimise its media transcoding operation. The cost is 1494 greatly increased computational complexity on the forwarding node. 1495 Receivers of the forwarded stream will see the forwarding device 1496 as the sender of the stream, and will not be able to tell from the 1497 RTP layer that they are receiving a forwarded stream rather than 1498 an entirely new RTP packet stream generated by the forwarding 1499 device. 1501 12.1.3. Differentiated Treatment of RTP Packet Streams 1503 There are use cases for differentiated treatment of RTP packet 1504 streams. Such differentiation can happen at several places in the 1505 system. First of all is the prioritization within the end-point 1506 sending the media, which controls, both which RTP packet streams that 1507 will be sent, and their allocation of bit-rate out of the current 1508 available aggregate as determined by the congestion control. 1510 It is expected that the WebRTC API [W3C.WD-webrtc-20130910] will 1511 allow the application to indicate relative priorities for different 1512 MediaStreamTracks. These priorities can then be used to influence 1513 the local RTP processing, especially when it comes to congestion 1514 control response in how to divide the available bandwidth between the 1515 RTP packet streams. Any changes in relative priority will also need 1516 to be considered for RTP packet streams that are associated with the 1517 main RTP packet streams, such as redundant streams for RTP 1518 retransmission and FEC. The importance of such redundant RTP packet 1519 streams is dependent on the media type and codec used, in regards to 1520 how robust that codec is to packet loss. However, a default policy 1521 might to be to use the same priority for redundant RTP packet stream 1522 as for the source RTP packet stream. 1524 Secondly, the network can prioritize transport-layer flows and sub- 1525 flows, including RTP packet streams. Typically, differential 1526 treatment includes two steps, the first being identifying whether an 1527 IP packet belongs to a class that has to be treated differently, the 1528 second consisting of the actual mechanism to prioritize packets. 1529 This is done according to three methods: 1531 DiffServ: The end-point marks a packet with a DiffServ code point to 1532 indicate to the network that the packet belongs to a particular 1533 class. 1535 Flow based: Packets that need to be given a particular treatment are 1536 identified using a combination of IP and port address. 1538 Deep Packet Inspection: A network classifier (DPI) inspects the 1539 packet and tries to determine if the packet represents a 1540 particular application and type that is to be prioritized. 1542 Flow-based differentiation will provide the same treatment to all 1543 packets within a transport-layer flow, i.e., relative prioritization 1544 is not possible. Moreover, if the resources are limited it might not 1545 be possible to provide differential treatment compared to best-effort 1546 for all the RTP packet streams used in a WebRTC session. When flow- 1547 based differentiation is available, the WebRTC Endpoint needs to know 1548 about it so that it can provide the separation of the RTP packet 1549 streams onto different UDP flows to enable a more granular usage of 1550 flow based differentiation. That way at least providing different 1551 prioritization of audio and video if desired by application. 1553 DiffServ assumes that either the end-point or a classifier can mark 1554 the packets with an appropriate DSCP so that the packets are treated 1555 according to that marking. If the end-point is to mark the traffic 1556 two requirements arise in the WebRTC context: 1) The WebRTC Endpoint 1557 has to know which DSCP to use and that it can use them on some set of 1558 RTP packet streams. 2) The information needs to be propagated to the 1559 operating system when transmitting the packet. Details of this 1560 process are outside the scope of this memo and are further discussed 1561 in "DSCP and other packet markings for RTCWeb QoS" 1562 [I-D.ietf-tsvwg-rtcweb-qos]. 1564 For packet based marking schemes it might be possible to mark 1565 individual RTP packets differently based on the relative priority of 1566 the RTP payload. For example video codecs that have I, P, and B 1567 pictures could prioritise any payloads carrying only B frames less, 1568 as these are less damaging to loose. However, depending on the QoS 1569 mechanism and what markings that are applied, this can result in not 1570 only different packet drop probabilities but also packet reordering, 1571 see [I-D.ietf-tsvwg-rtcweb-qos] for further discussion. As a default 1572 policy all RTP packets related to a RTP packet stream ought to be 1573 provided with the same prioritization; per-packet prioritization is 1574 outside the scope of this memo, but might be specified elsewhere in 1575 future. 1577 It is also important to consider how RTCP packets associated with a 1578 particular RTP packet stream need to be marked. RTCP compound 1579 packets with Sender Reports (SR), ought to be marked with the same 1580 priority as the RTP packet stream itself, so the RTCP-based round- 1581 trip time (RTT) measurements are done using the same transport-layer 1582 flow priority as the RTP packet stream experiences. RTCP compound 1583 packets containing RR packet ought to be sent with the priority used 1584 by the majority of the RTP packet streams reported on. RTCP packets 1585 containing time-critical feedback packets can use higher priority to 1586 improve the timeliness and likelihood of delivery of such feedback. 1588 12.2. Media Source, RTP Packet Streams, and Participant Identification 1590 12.2.1. Media Source Identification 1592 Each RTP packet stream is identified by a unique synchronisation 1593 source (SSRC) identifier. The SSRC identifier is carried in each of 1594 the RTP packets comprising a RTP packet stream, and is also used to 1595 identify that stream in the corresponding RTCP reports. The SSRC is 1596 chosen as discussed in Section 4.8. The first stage in 1597 demultiplexing RTP and RTCP packets received on a single transport 1598 layer flow at a WebRTC Endpoint is to separate the RTP packet streams 1599 based on their SSRC value; once that is done, additional 1600 demultiplexing steps can determine how and where to render the media. 1602 RTP allows a mixer, or other RTP-layer middlebox, to combine encoded 1603 streams from multiple media sources to form a new encoded stream from 1604 a new media source (the mixer). The RTP packets in that new RTP 1605 packet stream can include a Contributing Source (CSRC) list, 1606 indicating which original SSRCs contributed to the combined source 1607 stream. As described in Section 4.1, implementations need to support 1608 reception of RTP data packets containing a CSRC list and RTCP packets 1609 that relate to sources present in the CSRC list. The CSRC list can 1610 change on a packet-by-packet basis, depending on the mixing operation 1611 being performed. Knowledge of what media sources contributed to a 1612 particular RTP packet can be important if the user interface 1613 indicates which participants are active in the session. Changes in 1614 the CSRC list included in packets needs to be exposed to the WebRTC 1615 application using some API, if the application is to be able to track 1616 changes in session participation. It is desirable to map CSRC values 1617 back into WebRTC MediaStream identities as they cross this API, to 1618 avoid exposing the SSRC/CSRC name space to WebRTC applications. 1620 If the mixer-to-client audio level extension [RFC6465] is being used 1621 in the session (see Section 5.2.3), the information in the CSRC list 1622 is augmented by audio level information for each contributing source. 1623 It is desirable to expose this information to the WebRTC application 1624 using some API, after mapping the CSRC values to WebRTC MediaStream 1625 identities, so it can be exposed in the user interface. 1627 12.2.2. SSRC Collision Detection 1629 The RTP standard requires RTP implementations to have support for 1630 detecting and handling SSRC collisions, i.e., resolve the conflict 1631 when two different end-points use the same SSRC value (see section 1632 8.2 of [RFC3550]). This requirement also applies to WebRTC 1633 Endpoints. There are several scenarios where SSRC collisions can 1634 occur: 1636 o In a point-to-point session where each SSRC is associated with 1637 either of the two end-points and where the main media carrying 1638 SSRC identifier will be announced in the signalling channel, a 1639 collision is less likely to occur due to the information about 1640 used SSRCs. If SDP is used, this information is provided by 1641 Source-Specific SDP Attributes [RFC5576]. Still, collisions can 1642 occur if both end-points start using a new SSRC identifier prior 1643 to having signalled it to the peer and received acknowledgement on 1644 the signalling message. The Source-Specific SDP Attributes 1645 [RFC5576] contains a mechanism to signal how the end-point 1646 resolved the SSRC collision. 1648 o SSRC values that have not been signalled could also appear in an 1649 RTP session. This is more likely than it appears, since some RTP 1650 functions use extra SSRCs to provide their functionality. For 1651 example, retransmission data might be transmitted using a separate 1652 RTP packet stream that requires its own SSRC, separate to the SSRC 1653 of the source RTP packet stream [RFC4588]. In those cases, an 1654 end-point can create a new SSRC that strictly doesn't need to be 1655 announced over the signalling channel to function correctly on 1656 both RTP and RTCPeerConnection level. 1658 o Multiple end-points in a multiparty conference can create new 1659 sources and signal those towards the RTP middlebox. In cases 1660 where the SSRC/CSRC are propagated between the different end- 1661 points from the RTP middlebox collisions can occur. 1663 o An RTP middlebox could connect an end-point's RTCPeerConnection to 1664 another RTCPeerConnection from the same end-point, thus forming a 1665 loop where the end-point will receive its own traffic. While it 1666 is clearly considered a bug, it is important that the end-point is 1667 able to recognise and handle the case when it occurs. This case 1668 becomes even more problematic when media mixers, and so on, are 1669 involved, where the stream received is a different stream but 1670 still contains this client's input. 1672 These SSRC/CSRC collisions can only be handled on RTP level as long 1673 as the same RTP session is extended across multiple 1674 RTCPeerConnections by a RTP middlebox. To resolve the more generic 1675 case where multiple RTCPeerConnections are interconnected, 1676 identification of the media source(s) part of a MediaStreamTrack 1677 being propagated across multiple interconnected RTCPeerConnection 1678 needs to be preserved across these interconnections. 1680 12.2.3. Media Synchronisation Context 1682 When an end-point sends media from more than one media source, it 1683 needs to consider if (and which of) these media sources are to be 1684 synchronized. In RTP/RTCP, synchronisation is provided by having a 1685 set of RTP packet streams be indicated as coming from the same 1686 synchronisation context and logical end-point by using the same RTCP 1687 CNAME identifier. 1689 The next provision is that the internal clocks of all media sources, 1690 i.e., what drives the RTP timestamp, can be correlated to a system 1691 clock that is provided in RTCP Sender Reports encoded in an NTP 1692 format. By correlating all RTP timestamps to a common system clock 1693 for all sources, the timing relation of the different RTP packet 1694 streams, also across multiple RTP sessions can be derived at the 1695 receiver and, if desired, the streams can be synchronized. The 1696 requirement is for the media sender to provide the correlation 1697 information; it is up to the receiver to use it or not. 1699 13. Security Considerations 1701 The overall security architecture for WebRTC is described in 1702 [I-D.ietf-rtcweb-security-arch], and security considerations for the 1703 WebRTC framework are described in [I-D.ietf-rtcweb-security]. These 1704 considerations also apply to this memo. 1706 The security considerations of the RTP specification, the RTP/SAVPF 1707 profile, and the various RTP/RTCP extensions and RTP payload formats 1708 that form the complete protocol suite described in this memo apply. 1709 It is not believed there are any new security considerations 1710 resulting from the combination of these various protocol extensions. 1712 The Extended Secure RTP Profile for Real-time Transport Control 1713 Protocol (RTCP)-Based Feedback [RFC5124] (RTP/SAVPF) provides 1714 handling of fundamental issues by offering confidentiality, integrity 1715 and partial source authentication. A mandatory to implement media 1716 security solution is created by combing this secured RTP profile and 1717 DTLS-SRTP keying [RFC5764] as defined by Section 5.5 of 1718 [I-D.ietf-rtcweb-security-arch]. 1720 RTCP packets convey a Canonical Name (CNAME) identifier that is used 1721 to associate RTP packet streams that need to be synchronised across 1722 related RTP sessions. Inappropriate choice of CNAME values can be a 1723 privacy concern, since long-term persistent CNAME identifiers can be 1724 used to track users across multiple WebRTC calls. Section 4.9 of 1725 this memo provides guidelines for generation of untraceable CNAME 1726 values that alleviate this risk. 1728 Some potential denial of service attacks exist if the RTCP reporting 1729 interval is configured to an inappropriate value. This could be done 1730 by configuring the RTCP bandwidth fraction to an excessively large or 1731 small value using the SDP "b=RR:" or "b=RS:" lines [RFC3556], or some 1732 similar mechanism, or by choosing an excessively large or small value 1733 for the RTP/AVPF minimal receiver report interval (if using SDP, this 1734 is the "a=rtcp-fb:... trr-int" parameter) [RFC4585]. The risks are 1735 as follows: 1737 1. the RTCP bandwidth could be configured to make the regular 1738 reporting interval so large that effective congestion control 1739 cannot be maintained, potentially leading to denial of service 1740 due to congestion caused by the media traffic; 1742 2. the RTCP interval could be configured to a very small value, 1743 causing endpoints to generate high rate RTCP traffic, potentially 1744 leading to denial of service due to the non-congestion controlled 1745 RTCP traffic; and 1747 3. RTCP parameters could be configured differently for each 1748 endpoint, with some of the endpoints using a large reporting 1749 interval and some using a smaller interval, leading to denial of 1750 service due to premature participant timeouts due to mismatched 1751 timeout periods which are based on the reporting interval (this 1752 is a particular concern if endpoints use a small but non-zero 1753 value for the RTP/AVPF minimal receiver report interval (trr-int) 1754 [RFC4585], as discussed in Section 6.1 of 1755 [I-D.ietf-avtcore-rtp-multi-stream]). 1757 Premature participant timeout can be avoided by using the fixed (non- 1758 reduced) minimum interval when calculating the participant timeout 1759 (see Section 4.1 of this memo and Section 6.1 of 1760 [I-D.ietf-avtcore-rtp-multi-stream]). To address the other concerns, 1761 endpoints SHOULD ignore parameters that configure the RTCP reporting 1762 interval to be significantly longer than the default five second 1763 interval specified in [RFC3550] (unless the media data rate is so low 1764 that the longer reporting interval roughly corresponds to 5% of the 1765 media data rate), or that configure the RTCP reporting interval small 1766 enough that the RTCP bandwidth would exceed the media bandwidth. 1768 The guidelines in [RFC6562] apply when using variable bit rate (VBR) 1769 audio codecs such as Opus (see Section 4.3 for discussion of mandated 1770 audio codecs). The guidelines in [RFC6562] also apply, but are of 1771 lesser importance, when using the client-to-mixer audio level header 1772 extensions (Section 5.2.2) or the mixer-to-client audio level header 1773 extensions (Section 5.2.3). The use of the encryption of the header 1774 extensions are RECOMMENDED, unless there are known reasons, like RTP 1775 middleboxes or third party monitoring that will greatly benefit from 1776 the information, and this has been expressed using API or signalling. 1777 If further evidence are produced to show that information leakage is 1778 significant from audio level indications, then use of encryption 1779 needs to be mandated at that time. 1781 14. IANA Considerations 1783 This memo makes no request of IANA. 1785 Note to RFC Editor: this section is to be removed on publication as 1786 an RFC. 1788 15. Acknowledgements 1790 The authors would like to thank Bernard Aboba, Harald Alvestrand, 1791 Cary Bran, Ben Campbell, Charles Eckel, Alex Eleftheriadis, Christian 1792 Groves, Cullen Jennings, Olle Johansson, Suhas Nandakumar, Dan 1793 Romascanu, Jim Spring, Martin Thomson, and the other members of the 1794 IETF RTCWEB working group for their valuable feedback. 1796 16. References 1798 16.1. Normative References 1800 [I-D.ietf-avtcore-multi-media-rtp-session] 1801 Westerlund, M., Perkins, C., and J. Lennox, "Sending 1802 Multiple Types of Media in a Single RTP Session", draft- 1803 ietf-avtcore-multi-media-rtp-session-06 (work in 1804 progress), October 2014. 1806 [I-D.ietf-avtcore-rtp-circuit-breakers] 1807 Perkins, C. and V. Singh, "Multimedia Congestion Control: 1808 Circuit Breakers for Unicast RTP Sessions", draft-ietf- 1809 avtcore-rtp-circuit-breakers-06 (work in progress), July 1810 2014. 1812 [I-D.ietf-avtcore-rtp-multi-stream-optimisation] 1813 Lennox, J., Westerlund, M., Wu, Q., and C. Perkins, 1814 "Sending Multiple Media Streams in a Single RTP Session: 1815 Grouping RTCP Reception Statistics and Other Feedback ", 1816 draft-ietf-avtcore-rtp-multi-stream-optimisation-00 (work 1817 in progress), July 2013. 1819 [I-D.ietf-avtcore-rtp-multi-stream] 1820 Lennox, J., Westerlund, M., Wu, W., and C. Perkins, 1821 "Sending Multiple Media Streams in a Single RTP Session", 1822 draft-ietf-avtcore-rtp-multi-stream-05 (work in progress), 1823 July 2014. 1825 [I-D.ietf-rtcweb-security-arch] 1826 Rescorla, E., "WebRTC Security Architecture", draft-ietf- 1827 rtcweb-security-arch-10 (work in progress), July 2014. 1829 [I-D.ietf-rtcweb-security] 1830 Rescorla, E., "Security Considerations for WebRTC", draft- 1831 ietf-rtcweb-security-07 (work in progress), July 2014. 1833 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1834 Requirement Levels", BCP 14, RFC 2119, March 1997. 1836 [RFC2736] Handley, M. and C. Perkins, "Guidelines for Writers of RTP 1837 Payload Format Specifications", BCP 36, RFC 2736, December 1838 1999. 1840 [RFC3550] Schulzrinne, H., Casner, S., Frederick, R., and V. 1841 Jacobson, "RTP: A Transport Protocol for Real-Time 1842 Applications", STD 64, RFC 3550, July 2003. 1844 [RFC3551] Schulzrinne, H. and S. Casner, "RTP Profile for Audio and 1845 Video Conferences with Minimal Control", STD 65, RFC 3551, 1846 July 2003. 1848 [RFC3556] Casner, S., "Session Description Protocol (SDP) Bandwidth 1849 Modifiers for RTP Control Protocol (RTCP) Bandwidth", RFC 1850 3556, July 2003. 1852 [RFC3711] Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K. 1853 Norrman, "The Secure Real-time Transport Protocol (SRTP)", 1854 RFC 3711, March 2004. 1856 [RFC4566] Handley, M., Jacobson, V., and C. Perkins, "SDP: Session 1857 Description Protocol", RFC 4566, July 2006. 1859 [RFC4585] Ott, J., Wenger, S., Sato, N., Burmeister, C., and J. Rey, 1860 "Extended RTP Profile for Real-time Transport Control 1861 Protocol (RTCP)-Based Feedback (RTP/AVPF)", RFC 4585, July 1862 2006. 1864 [RFC4588] Rey, J., Leon, D., Miyazaki, A., Varsa, V., and R. 1865 Hakenberg, "RTP Retransmission Payload Format", RFC 4588, 1866 July 2006. 1868 [RFC4961] Wing, D., "Symmetric RTP / RTP Control Protocol (RTCP)", 1869 BCP 131, RFC 4961, July 2007. 1871 [RFC5104] Wenger, S., Chandra, U., Westerlund, M., and B. Burman, 1872 "Codec Control Messages in the RTP Audio-Visual Profile 1873 with Feedback (AVPF)", RFC 5104, February 2008. 1875 [RFC5124] Ott, J. and E. Carrara, "Extended Secure RTP Profile for 1876 Real-time Transport Control Protocol (RTCP)-Based Feedback 1877 (RTP/SAVPF)", RFC 5124, February 2008. 1879 [RFC5285] Singer, D. and H. Desineni, "A General Mechanism for RTP 1880 Header Extensions", RFC 5285, July 2008. 1882 [RFC5506] Johansson, I. and M. Westerlund, "Support for Reduced-Size 1883 Real-Time Transport Control Protocol (RTCP): Opportunities 1884 and Consequences", RFC 5506, April 2009. 1886 [RFC5761] Perkins, C. and M. Westerlund, "Multiplexing RTP Data and 1887 Control Packets on a Single Port", RFC 5761, April 2010. 1889 [RFC5764] McGrew, D. and E. Rescorla, "Datagram Transport Layer 1890 Security (DTLS) Extension to Establish Keys for the Secure 1891 Real-time Transport Protocol (SRTP)", RFC 5764, May 2010. 1893 [RFC6051] Perkins, C. and T. Schierl, "Rapid Synchronisation of RTP 1894 Flows", RFC 6051, November 2010. 1896 [RFC6464] Lennox, J., Ivov, E., and E. Marocco, "A Real-time 1897 Transport Protocol (RTP) Header Extension for Client-to- 1898 Mixer Audio Level Indication", RFC 6464, December 2011. 1900 [RFC6465] Ivov, E., Marocco, E., and J. Lennox, "A Real-time 1901 Transport Protocol (RTP) Header Extension for Mixer-to- 1902 Client Audio Level Indication", RFC 6465, December 2011. 1904 [RFC6562] Perkins, C. and JM. Valin, "Guidelines for the Use of 1905 Variable Bit Rate Audio with Secure RTP", RFC 6562, March 1906 2012. 1908 [RFC6904] Lennox, J., "Encryption of Header Extensions in the Secure 1909 Real-time Transport Protocol (SRTP)", RFC 6904, April 1910 2013. 1912 [RFC7007] Terriberry, T., "Update to Remove DVI4 from the 1913 Recommended Codecs for the RTP Profile for Audio and Video 1914 Conferences with Minimal Control (RTP/AVP)", RFC 7007, 1915 August 2013. 1917 [RFC7022] Begen, A., Perkins, C., Wing, D., and E. Rescorla, 1918 "Guidelines for Choosing RTP Control Protocol (RTCP) 1919 Canonical Names (CNAMEs)", RFC 7022, September 2013. 1921 [RFC7160] Petit-Huguenin, M. and G. Zorn, "Support for Multiple 1922 Clock Rates in an RTP Session", RFC 7160, April 2014. 1924 [RFC7164] Gross, K. and R. Brandenburg, "RTP and Leap Seconds", RFC 1925 7164, March 2014. 1927 16.2. Informative References 1929 [I-D.ietf-avtcore-multiplex-guidelines] 1930 Westerlund, M., Perkins, C., and H. Alvestrand, 1931 "Guidelines for using the Multiplexing Features of RTP to 1932 Support Multiple Media Streams", draft-ietf-avtcore- 1933 multiplex-guidelines-03 (work in progress), October 2014. 1935 [I-D.ietf-avtcore-rtp-topologies-update] 1936 Westerlund, M. and S. Wenger, "RTP Topologies", draft- 1937 ietf-avtcore-rtp-topologies-update-04 (work in progress), 1938 August 2014. 1940 [I-D.ietf-avtext-rtp-grouping-taxonomy] 1941 Lennox, J., Gross, K., Nandakumar, S., and G. Salgueiro, 1942 "A Taxonomy of Grouping Semantics and Mechanisms for Real- 1943 Time Transport Protocol (RTP) Sources", draft-ietf-avtext- 1944 rtp-grouping-taxonomy-02 (work in progress), June 2014. 1946 [I-D.ietf-mmusic-msid] 1947 Alvestrand, H., "WebRTC MediaStream Identification in the 1948 Session Description Protocol", draft-ietf-mmusic-msid-07 1949 (work in progress), October 2014. 1951 [I-D.ietf-mmusic-sdp-bundle-negotiation] 1952 Holmberg, C., Alvestrand, H., and C. Jennings, 1953 "Negotiating Media Multiplexing Using the Session 1954 Description Protocol (SDP)", draft-ietf-mmusic-sdp-bundle- 1955 negotiation-12 (work in progress), October 2014. 1957 [I-D.ietf-payload-rtp-howto] 1958 Westerlund, M., "How to Write an RTP Payload Format", 1959 draft-ietf-payload-rtp-howto-13 (work in progress), 1960 January 2014. 1962 [I-D.ietf-rmcat-cc-requirements] 1963 Jesup, R., "Congestion Control Requirements For RMCAT", 1964 draft-ietf-rmcat-cc-requirements-06 (work in progress), 1965 October 2014. 1967 [I-D.ietf-rtcweb-audio] 1968 Valin, J. and C. Bran, "WebRTC Audio Codec and Processing 1969 Requirements", draft-ietf-rtcweb-audio-06 (work in 1970 progress), September 2014. 1972 [I-D.ietf-rtcweb-overview] 1973 Alvestrand, H., "Overview: Real Time Protocols for 1974 Browser-based Applications", draft-ietf-rtcweb-overview-12 1975 (work in progress), October 2014. 1977 [I-D.ietf-rtcweb-use-cases-and-requirements] 1978 Holmberg, C., Hakansson, S., and G. Eriksson, "Web Real- 1979 Time Communication Use-cases and Requirements", draft- 1980 ietf-rtcweb-use-cases-and-requirements-14 (work in 1981 progress), February 2014. 1983 [I-D.ietf-tsvwg-rtcweb-qos] 1984 Dhesikan, S., Jennings, C., Druta, D., Jones, P., and J. 1985 Polk, "DSCP and other packet markings for RTCWeb QoS", 1986 draft-ietf-tsvwg-rtcweb-qos-02 (work in progress), June 1987 2014. 1989 [RFC3611] Friedman, T., Caceres, R., and A. Clark, "RTP Control 1990 Protocol Extended Reports (RTCP XR)", RFC 3611, November 1991 2003. 1993 [RFC4383] Baugher, M. and E. Carrara, "The Use of Timed Efficient 1994 Stream Loss-Tolerant Authentication (TESLA) in the Secure 1995 Real-time Transport Protocol (SRTP)", RFC 4383, February 1996 2006. 1998 [RFC5245] Rosenberg, J., "Interactive Connectivity Establishment 1999 (ICE): A Protocol for Network Address Translator (NAT) 2000 Traversal for Offer/Answer Protocols", RFC 5245, April 2001 2010. 2003 [RFC5576] Lennox, J., Ott, J., and T. Schierl, "Source-Specific 2004 Media Attributes in the Session Description Protocol 2005 (SDP)", RFC 5576, June 2009. 2007 [RFC5968] Ott, J. and C. Perkins, "Guidelines for Extending the RTP 2008 Control Protocol (RTCP)", RFC 5968, September 2010. 2010 [RFC6263] Marjou, X. and A. Sollaud, "Application Mechanism for 2011 Keeping Alive the NAT Mappings Associated with RTP / RTP 2012 Control Protocol (RTCP) Flows", RFC 6263, June 2011. 2014 [RFC6792] Wu, Q., Hunt, G., and P. Arden, "Guidelines for Use of the 2015 RTP Monitoring Framework", RFC 6792, November 2012. 2017 [W3C.WD-mediacapture-streams-20130903] 2018 Burnett, D., Bergkvist, A., Jennings, C., and A. 2019 Narayanan, "Media Capture and Streams", World Wide Web 2020 Consortium WD WD-mediacapture-streams-20130903, September 2021 2013, . 2024 [W3C.WD-webrtc-20130910] 2025 Bergkvist, A., Burnett, D., Jennings, C., and A. 2026 Narayanan, "WebRTC 1.0: Real-time Communication Between 2027 Browsers", World Wide Web Consortium WD WD- 2028 webrtc-20130910, September 2013, 2029 . 2031 Authors' Addresses 2033 Colin Perkins 2034 University of Glasgow 2035 School of Computing Science 2036 Glasgow G12 8QQ 2037 United Kingdom 2039 Email: csp@csperkins.org 2040 URI: http://csperkins.org/ 2042 Magnus Westerlund 2043 Ericsson 2044 Farogatan 6 2045 SE-164 80 Kista 2046 Sweden 2048 Phone: +46 10 714 82 87 2049 Email: magnus.westerlund@ericsson.com 2050 Joerg Ott 2051 Aalto University 2052 School of Electrical Engineering 2053 Espoo 02150 2054 Finland 2056 Email: jorg.ott@aalto.fi