idnits 2.17.1 draft-ietf-avtcore-rtp-multi-stream-01.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 (Using the creation date from RFC3550, updated by this document, for RFC5378 checks: 1998-04-07) -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (July 11, 2013) is 3943 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) == Unused Reference: 'I-D.ietf-avtcore-6222bis' is defined on line 533, but no explicit reference was found in the text == Outdated reference: A later version (-06) exists of draft-ietf-avtcore-6222bis-04 == Outdated reference: A later version (-13) exists of draft-ietf-avtcore-multi-media-rtp-session-02 == Outdated reference: A later version (-12) exists of draft-ietf-avtcore-rtp-multi-stream-optimisation-00 == Outdated reference: A later version (-10) exists of draft-ietf-avtcore-rtp-topologies-update-00 == Outdated reference: A later version (-25) exists of draft-ietf-clue-framework-10 == Outdated reference: A later version (-54) exists of draft-ietf-mmusic-sdp-bundle-negotiation-04 Summary: 0 errors (**), 0 flaws (~~), 8 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 AVTCORE J. Lennox 3 Internet-Draft Vidyo 4 Updates: 3550 (if approved) M. Westerlund 5 Intended status: Standards Track Ericsson 6 Expires: January 12, 2014 Q. Wu 7 Huawei 8 C. Perkins 9 University of Glasgow 10 July 11, 2013 12 Sending Multiple Media Streams in a Single RTP Session 13 draft-ietf-avtcore-rtp-multi-stream-01 15 Abstract 17 This document expands and clarifies the behavior of the Real-Time 18 Transport Protocol (RTP) endpoints when they are sending multiple 19 media streams in a single RTP session. In particular, issues 20 involving Real-Time Transport Control Protocol (RTCP) messages are 21 described. 23 This document updates RFC 3550 in regards to handling of multiple 24 SSRCs per endpoint in RTP sessions. 26 Status of This Memo 28 This Internet-Draft is submitted in full conformance with the 29 provisions of BCP 78 and BCP 79. 31 Internet-Drafts are working documents of the Internet Engineering 32 Task Force (IETF). Note that other groups may also distribute 33 working documents as Internet-Drafts. The list of current Internet- 34 Drafts is at http://datatracker.ietf.org/drafts/current/. 36 Internet-Drafts are draft documents valid for a maximum of six months 37 and may be updated, replaced, or obsoleted by other documents at any 38 time. It is inappropriate to use Internet-Drafts as reference 39 material or to cite them other than as "work in progress." 41 This Internet-Draft will expire on January 12, 2014. 43 Copyright Notice 45 Copyright (c) 2013 IETF Trust and the persons identified as the 46 document authors. All rights reserved. 48 This document is subject to BCP 78 and the IETF Trust's Legal 49 Provisions Relating to IETF Documents 50 (http://trustee.ietf.org/license-info) in effect on the date of 51 publication of this document. Please review these documents 52 carefully, as they describe your rights and restrictions with respect 53 to this document. Code Components extracted from this document must 54 include Simplified BSD License text as described in Section 4.e of 55 the Trust Legal Provisions and are provided without warranty as 56 described in the Simplified BSD License. 58 Table of Contents 60 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 61 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 3 62 3. Use Cases For Multi-Stream Endpoints . . . . . . . . . . . . 3 63 3.1. Multiple-Capturer Endpoints . . . . . . . . . . . . . . . 3 64 3.2. Multi-Media Sessions . . . . . . . . . . . . . . . . . . 3 65 3.3. Multi-Stream Mixers . . . . . . . . . . . . . . . . . . . 4 66 4. Multi-Stream Endpoint RTP Media Recommendations . . . . . . . 4 67 5. Multi-Stream Endpoint RTCP Recommendations . . . . . . . . . 4 68 5.1. RTCP Reporting Requirement . . . . . . . . . . . . . . . 5 69 5.2. Initial Reporting Interval . . . . . . . . . . . . . . . 5 70 5.3. Compound RTCP Packets . . . . . . . . . . . . . . . . . . 5 71 6. RTCP Considerations for Streams with Disparate Rates . . . . 7 72 6.1. Timing out SSRCs . . . . . . . . . . . . . . . . . . . . 8 73 6.2. Tuning RTCP transmissions . . . . . . . . . . . . . . . . 9 74 7. Security Considerations . . . . . . . . . . . . . . . . . . . 11 75 8. Open Issues . . . . . . . . . . . . . . . . . . . . . . . . . 12 76 9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 12 77 10. References . . . . . . . . . . . . . . . . . . . . . . . . . 12 78 10.1. Normative References . . . . . . . . . . . . . . . . . . 12 79 10.2. Informative References . . . . . . . . . . . . . . . . . 13 80 Appendix A. Changes From Earlier Versions . . . . . . . . . . . 14 81 A.1. Changes From WG Draft -00 . . . . . . . . . . . . . . . . 14 82 A.2. Changes From Individual Draft -02 . . . . . . . . . . . . 14 83 A.3. Changes From Individual Draft -01 . . . . . . . . . . . . 14 84 A.4. Changes From Individual Draft -00 . . . . . . . . . . . . 14 85 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 15 87 1. Introduction 89 At the time The Real-Time Transport Protocol (RTP) [RFC3550] was 90 originally written, and for quite some time after, endpoints in RTP 91 sessions typically only transmitted a single media stream per RTP 92 session, where separate RTP sessions were typically used for each 93 distinct media type. 95 Recently, however, a number of scenarios have emerged (discussed 96 further in Section 3) in which endpoints wish to send multiple RTP 97 media streams, distinguished by distinct RTP synchronization source 98 (SSRC) identifiers, in a single RTP session. Although RTP's initial 99 design did consider such scenarios, the specification was not 100 consistently written with such use cases in mind. The specifications 101 are thus somewhat unclear. 103 The purpose of this document is to expand and clarify [RFC3550]'s 104 language for these use cases. The authors believe this does not 105 result in any major normative changes to the RTP specification, 106 however this document defines how the RTP specification is to be 107 interpreted. In these cases, this document updates RFC3550. 109 The document starts with terminology and some use cases where 110 multiple sources will occur. This is followed by some case studies 111 to try to identify issues that exist and need considerations. This 112 is followed by RTP and RTCP recommendations to resolve issues. Next 113 are security considerations and remaining open issues. 115 2. Terminology 117 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 118 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 119 "OPTIONAL" in this document are to be interpreted as described in RFC 120 2119 [RFC2119] and indicate requirement levels for compliant 121 implementations. 123 3. Use Cases For Multi-Stream Endpoints 125 This section discusses several use cases that have motivated the 126 development of endpoints that send multiple streams in a single RTP 127 session. 129 3.1. Multiple-Capturer Endpoints 131 The most straightforward motivation for an endpoint to send multiple 132 media streams in a session is the scenario where an endpoint has 133 multiple capture devices of the same media type and characteristics. 134 For example, telepresence endpoints, of the type described by the 135 CLUE Telepresence Framework [I-D.ietf-clue-framework] is designed, 136 often have multiple cameras or microphones covering various areas of 137 a room. 139 3.2. Multi-Media Sessions 141 Recent work has been done in RTP 142 [I-D.ietf-avtcore-multi-media-rtp-session] and SDP 144 [I-D.ietf-mmusic-sdp-bundle-negotiation] to update RTP's historical 145 assumption that media streams of different media types would always 146 be sent on different RTP sessions. In this work, a single endpoint's 147 audio and video media streams (for example) are instead sent in a 148 single RTP session. 150 3.3. Multi-Stream Mixers 152 There are several RTP topologies which can involve a central device 153 that itself generates multiple media streams in a session. 155 One example is a mixer providing centralized compositing for a multi- 156 capture scenario like that described in Section 3.1. In this case, 157 the centralized node is behaving much like a multi-capturer endpoint, 158 generating several similar and related sources. 160 More complicated is the Source Projecting Mixer, see Section 3.6 of 161 [I-D.ietf-avtcore-rtp-topologies-update]. This is a central box that 162 receives media streams from several endpoints, and then selectively 163 forwards modified versions of some of the streams toward the other 164 endpoints it is connected to. Toward one destination, a separate 165 media source appears in the session for every other source connected 166 to the mixer, "projected" from the original streams, but at any given 167 time many of them can appear to be inactive (and thus are receivers, 168 not senders, in RTP). This sort of device is closer to being an RTP 169 mixer than an RTP translator, in that it terminates RTCP reporting 170 about the mixed streams, and it can re-write SSRCs, timestamps, and 171 sequence numbers, as well as the contents of the RTP payloads, and 172 can turn sources on and off at will without appearing to be 173 generating packet loss. Each projected stream will typically 174 preserve its original RTCP source description (SDES) information. 176 4. Multi-Stream Endpoint RTP Media Recommendations 178 While an endpoint MUST (of course) stay within its share of the 179 available session bandwidth, as determined by signalling and 180 congestion control, this need not be applied independently or 181 uniformly to each media stream. In particular, session bandwidth MAY 182 be reallocated among an endpoint's media streams, for example by 183 varying the bandwidth use of a variable-rate codec, or changing the 184 codec used by the media stream, up to the constraints of the 185 session's negotiated (or declared) codecs. This includes enabling or 186 disabling media streams as more or less bandwidth becomes available. 188 5. Multi-Stream Endpoint RTCP Recommendations 189 This section contains a number of different RTCP clarifications or 190 recommendations that enables more efficient and simpler behavior 191 without loss of functionality. 193 The RTP Control Protocol (RTCP) is defined in Section 6 of [RFC3550], 194 but it is largely documented in terms of "participants". In many 195 cases, the specification's recommendations for "participants" are to 196 be interpreted as applying to individual media streams, rather than 197 to endpoints. This section describes several concrete cases where 198 this applies. 200 (tbd: rather than think in terms of media streams, it might be 201 clearer to refer to SSRC values, where a participant with multiple 202 active SSRC values counts as multiple participants, once per SSRC) 204 5.1. RTCP Reporting Requirement 206 For each of an endpoint's media streams, whether or not it is 207 currently sending media, SR/RR and SDES packets MUST be sent at least 208 once per RTCP report interval. (For discussion of the content of SR 209 or RR packets' reception statistic reports, see 210 [I-D.ietf-avtcore-rtp-multi-stream-optimisation].) 212 5.2. Initial Reporting Interval 214 When a new media stream is added to a unicast session, the sentence 215 in [RFC3550]'s Section 6.2 applies: "For unicast sessions ... the 216 delay before sending the initial compound RTCP packet MAY be zero." 217 This applies to individual media sources as well. Thus, endpoints 218 MAY send an initial RTCP packet for an SSRC immediately upon adding 219 it to a unicast session. 221 This allowance also applies, as written, when initially joining a 222 unicast session. However, in this case some caution needs to be 223 exercised if the end-point or mixer has a large number of sources 224 (SSRCs) as this can create a significant burst. How big an issue 225 this depends on the number of source to send initial SR or RR and 226 Session Description CNAME items for in relation to the RTCP 227 bandwidth. 229 (tbd: Maybe some recommendation here? The aim in restricting this to 230 unicast sessions was to avoid this burst of traffic, which the usual 231 RTCP timing and reconsideration rules will prevent) 233 5.3. Compound RTCP Packets 235 Section 6.1 gives the following advice to RTP translators and mixers: 237 It is RECOMMENDED that translators and mixers combine individual 238 RTCP packets from the multiple sources they are forwarding into 239 one compound packet whenever feasible in order to amortize the 240 packet overhead (see Section 7). An example RTCP compound packet 241 as might be produced by a mixer is shown in Fig. 1. If the 242 overall length of a compound packet would exceed the MTU of the 243 network path, it SHOULD be segmented into multiple shorter 244 compound packets to be transmitted in separate packets of the 245 underlying protocol. This does not impair the RTCP bandwidth 246 estimation because each compound packet represents at least one 247 distinct participant. Note that each of the compound packets MUST 248 begin with an SR or RR packet. 250 Note: To avoid confusion, an RTCP packet is an individual item, such 251 as a Sender Report (SR), Receiver Report (RR), Source Description 252 (SDES), Goodbye (BYE), Application Defined (APP), Feedback [RFC4585] 253 or Extended Report (XR) [RFC3611] packet. A compound packet is the 254 combination of two or more such RTCP packets where the first packet 255 has to be an SR or an RR packet, and which contains a SDES packet 256 containing an CNAME item. Thus the above results in compound RTCP 257 packets that contain multiple SR or RR packets from different sources 258 as well as any of the other packet types. There are no restrictions 259 on the order in which the packets can occur within the compound 260 packet, except the regular compound rule, i.e., starting with an SR 261 or RR. 263 This advice applies to multi-media-stream endpoints as well, with the 264 same restrictions and considerations. (Note, however, that the last 265 sentence does not apply to AVPF [RFC4585] or SAVPF [RFC5124] feedback 266 packets if Reduced-Size RTCP [RFC5506] is in use.) 268 Due to RTCP's randomization of reporting times, there is a fair bit 269 of tolerance in precisely when an endpoint schedules RTCP to be sent. 270 Thus, one potential way of implementing this recommendation would be 271 to randomize all of an endpoint's sources together, with a single 272 randomization schedule, so an MTU's worth of RTCP all comes out 273 simultaneously. 275 (tbd: Multiplexing RTCP packets from multiple different sources might 276 require some adjustment to the calculation of RTCP's avg_rtcp_size, 277 as the RTCP group interval is proportional to avg_rtcp_size times the 278 group size). 280 6. RTCP Considerations for Streams with Disparate Rates 282 It is possible for a single RTP session to carry streams of greatly 283 differing bandwidth. There are two scenarios where this can occur. 284 The first is when a single RTP session carries multiple flows of the 285 same media type, but with very different quality; for example a video 286 switching multi-point conference unit might send a full rate high- 287 definition video stream of the active speaker but only thumbnails for 288 the other participants, all sent in a single RTP session. The second 289 scenarios occurs when audio and video flows are sent in a single RTP 290 session, as discussed in [I-D.ietf-avtcore-multi-media-rtp-session]. 292 An RTP session has a single set of parameters that configure the 293 session bandwidth, the RTCP sender and receiver fractions (e.g., via 294 the SDP "b=RR:" and "b=RS:" lines), and the parameters of the RTP/ 295 AVPF profile [RFC4585] (e.g., trr-int) if that profile (or its secure 296 extension, RTP/SAVPF [RFC5124]) is used. As a consequence, the RTCP 297 reporting interval will be the same for every SSRC in an RTP session. 298 This uniform RTCP reporting interval can result in RTCP reports being 299 sent more often than is considered desirable for a particular media 300 type. For example, if an audio flow is multiplexed with a high 301 quality video flow where the session bandwidth is configured to match 302 the video bandwidth, this can result in the RTCP packets having a 303 greater bandwidth allocation than the audio data rate. If the 304 reduced minimum RTCP interval described in Section 6.2 of [RFC3550] 305 is used in the session, which might be appropriate for video where 306 rapid feedback is wanted, the audio sources could be expected to send 307 RTCP packets more often than they send audio data packets. This is 308 most likely undesirable, and while the mismatch can be reduced 309 through careful tuning of the RTCP parameters, particularly trr_int 310 in RTP/AVPF sessions, it is inherent in the design of the RTCP timing 311 rules, and affects all RTP sessions containing flows with mismatched 312 bandwidth. 314 Having multiple media types in one RTP session also results in more 315 SSRCs being present in this RTP session. This increasing the amount 316 of cross reporting between the SSRCs. From an RTCP perspective, two 317 RTP sessions with half the number of SSRCs in each will be slightly 318 more efficient. If someone needs either the higher efficiency due to 319 the lesser number of SSRCs or the fact that one can't tailor RTCP 320 usage per media type, they need to use independent RTP sessions. 322 When it comes to configuring RTCP the need for regular periodic 323 reporting needs to be weighted against any feedback or control 324 messages being sent. Applications using RTP/AVPF or RTP/SAVPF are 325 RECOMMENDED to consider setting the trr-int parameter to a value 326 suitable for the application's needs, thus potentially reducing the 327 need for regular reporting and thus releasing more bandwidth for use 328 for feedback or control. 330 Another aspect of an RTP session with multiple media types is that 331 the RTCP packets, RTCP Feedback Messages, or RTCP XR metrics used 332 might not be applicable to all media types. Instead, all RTP/RTCP 333 endpoints need to correlate the media type of the SSRC being 334 referenced in a message or packet and only use those that apply to 335 that particular SSRC and its media type. Signalling solutions might 336 have shortcomings when it comes to indicating that a particular set 337 of RTCP reports or feedback messages only apply to a particular media 338 type within an RTP session. 340 6.1. Timing out SSRCs 342 All SSRCs used in an RTP session MUST use the same timeout behaviour 343 to avoid premature timeouts. This will depend on the RTP profile and 344 its configuration. The RTP specification provides several options 345 that can influence the values used when calculating the time 346 interval. To avoid interoperability issues when using this 347 specification, this document makes several clarifications to the 348 calculations. 350 For RTP/AVP, RTP/SAVP, RTP/AVPF, and RTP/SAVPF with T_rr_interval = 351 0, the timeout interval SHALL be calculated using a multiplier of 5, 352 i.e. the timeout interval becomes 5*Td. The Td calculation SHALL be 353 done using a Tmin value of 5 seconds, not the reduced minimal 354 interval even if used to calculate RTCP packet transmission 355 intervals. If using either the RTP/AVPF or RTP/SAVPF profiles with 356 T_rr_interval != 0 then the calculation as specified in Section 3.5.4 357 of RFC 4585 SHALL be used with a multiplier of 5, i.e. Tmin in the 358 Td calculation is the T_rr_interval. 360 Note: If endpoints implementing the RTP/AVP and RTP/AVPF profiles (or 361 their secure variants) are combined in a single RTP session, and the 362 RTP/AVPF endpoints use a non-zero T_rr_interval that is significantly 363 lower than 5 seconds, then there is a risk that the RTP/AVP endpoints 364 will prematurely timeout the RTP/AVPF endpoints due to their 365 different RTCP timeout intervals. Since an RTP session can only use 366 a single RTP profile, this issue ought never occur. If such mixed 367 RTP profiles are used, however, the RTP/AVPF session MUST NOT use a 368 non-zero T_rr_interval that is smaller than 5 seconds. 370 (tbd: it has been suggested that a minimum non-zero T_rr_interval of 371 4 seconds is more appropriate, due to the nature of the timing 372 rules). 374 6.2. Tuning RTCP transmissions 376 This sub-section discusses what tuning can be done to reduce the 377 downsides of the shared RTCP packet intervals. 379 When using the RTP/AVP or RTP/SAVP profiles the tuning one can do is 380 very limited. The controls one has are limited to the RTCP bandwidth 381 values and whether the minimum RTCP interval is scaled according to 382 the bandwidth. As the scheduling algorithm includes both random 383 factors and reconsideration, one can't simply calculate the expected 384 average transmission interval using the formula for Td. But it does 385 indicate the important factors affecting the transmission interval, 386 namely the RTCP bandwidth available for the role (Active Sender or 387 Participant), the average RTCP packet size, and the number of SSRCs 388 classified in the relevant role. Note that if the ratio of senders 389 to total number of session participants is larger than the ratio of 390 RTCP bandwidth for senders in relation to the total RTCP bandwidth, 391 then senders and receivers are treated together. 393 Let's start with some basic observations: 395 a. Unless the scaled minimum RTCP interval is used, then Td prior to 396 randomization and reconsideration can never be less than 5 397 seconds (assuming default Tmin of 5 seconds). 399 b. If the scaled minimum RTCP interval is used, Td can become as low 400 as 360 divided by RTP Session bandwidth in kilobits. In SDP the 401 RTP session bandwidth is signalled using b=AS. An RTP Session 402 bandwidth of 72 kbps results in Tmin being 5 seconds. An RTP 403 session bandwidth of 360 kbps of course gives a Tmin of 1 second, 404 and to achieve a Tmin equal to once every frame for a 25 Hz video 405 stream requires an RTP session bandwidth of 9 Mbps! (The use of 406 the RTP/AVPF or RTP/SAVPF profile allows a smaller Tmin, and 407 hence more frequent RTCP reports, as discussed below). 409 c. Let's calculate the number (n) of SSRCs in the RTP session that 410 5% of the session bandwidth can support to yield a Td value equal 411 to Tmin with minimal scaling. For this calculation we have to 412 make two assumptions. The first is that we will consider most or 413 all SSRC being senders, resulting in everyone sharing the 414 available bandwidth. Secondly we will select an average RTCP 415 packet size. This packet will consist of an SR, containing (n-1) 416 report blocks up to 31 report blocks, and an SDES item with at 417 least a CNAME (17 bytes in size) in it. Such a basic packet will 418 be 800 bytes for n>=32. With these parameters, and as the 419 bandwidth goes up the time interval is proportionally decreased 420 (due to minimal scaling), thus all the example bandwidths 72 421 kbps, 360 kbps and 9 Mbps all support 9 SSRCs. 423 d. The actual transmission interval for a Td value is [0.5*Td/ 424 1.21828,1.5*Td/1.21828], which means that for Td = 5 seconds, the 425 interval is actually [2.052,6.156] and the distribution is not 426 uniform, but rather exponentially-increasing. The probability 427 for sending at time X, given it is within the interval, is 428 probability of picking X in the interval times the probability to 429 randomly picking a number that is <=X within the interval with an 430 uniform probability distribution. This results in that the 431 majority of the probability mass is above the Td value. 433 To conclude, with RTP/AVP and RTP/SAVP the key limitation for small 434 unicast sessions is going to be the Tmin value. Thus the RTP session 435 bandwidth configured in RTCP has to be sufficiently high to reach the 436 reporting goals the application has following the rules for the 437 scaled minimal RTCP interval. 439 When using RTP/AVPF or RTP/SAVPF we get a quite powerful additional 440 tool, the setting of the T_rr_interval which has several effects on 441 the RTCP reporting. First of all as Tmin is set to 0 after the 442 initial transmission, the regular reporting interval is instead 443 determined by the regular bandwidth based calculation and the 444 T_rr_interval. This has the effect that we are no longer restricted 445 by the minimal interval or even the scaling rule for the minimal 446 rule. Instead the RTCP bandwidth and the T_rr_interval are the 447 governing factors. Now it also becomes important to separate between 448 the application's need for regular reports and RTCP feedback packet 449 types. In both regular RTCP mode, as in Early RTCP Mode, the usage 450 of the T_rr_interval prevents regular RTCP packets, i.e. packets 451 without any Feedback packets, to be sent more often than 452 T_rr_interval. This value is a hard as no regular RTCP packet can be 453 sent less than T_rr_interval after the previous regular packet 454 packet. 456 So applications that have a use for feedback packets for some media 457 streams, for example video streams, but don't want frequent regular 458 reporting for audio, could configure the T_rr_interval to a value so 459 that the regular reporting for both audio and video is at a level 460 that is considered acceptable for the audio. They could then use 461 feedback packets, which will include RTCP SR/RR packets, unless 462 reduced-size RTCP feedback packets [RFC5506] are used, and can 463 include other report information in addition to the feedback packet 464 that needs to be sent. That way the available RTCP bandwidth can be 465 focused for the use which provides the most utility for the 466 application. 468 Using T_rr_interval still requires one to determine suitable values 469 for the RTCP bandwidth value, in fact it might make it even more 470 important, as this is more likely to affect the RTCP behaviour and 471 performance than when using RTP/AVP, as there are fewer limitations 472 affecting the RTCP transmission. 474 When using T_rr_interval, i.e. having it be non zero, there are 475 configurations that have to be avoided. If the resulting Td value is 476 smaller but close to T_rr_interval then the interval in which the 477 actual regular RTCP packet transmission falls into becomes very 478 large, from 0.5 times T_rr_interval up to 2.73 times the 479 T_rr_interval. Therefore for configuration where one intends to have 480 Td smaller than T_rr_interval, then Td is RECOMMENDED to be targeted 481 at values less than 1/4th of T_rr_interval which results in that the 482 range becomes [0.5*T_rr_interval, 1.81*T_rr_interval]. 484 With RTP/AVPF, using a T_rr_interval of 0 or with another low value 485 significantly lower than Td still has utility, and different 486 behaviour compared to RTP/AVP. This avoids the Tmin limitations of 487 RTP/AVP, thus allowing more frequent regular RTCP reporting. In fact 488 this will result that the RTCP traffic becomes as high as the 489 configured values. 491 (tbd: a future version of this memo will include examples of how to 492 choose RTCP parameters for common scenarios) 494 There exists no method within the specification for using different 495 regular RTCP reporting intervals depending on the media type or 496 individual media stream. 498 7. Security Considerations 499 In the secure RTP protocol (SRTP) [RFC3711], the cryptographic 500 context of a compound SRTCP packet is the SSRC of the sender of the 501 first RTCP (sub-)packet. This could matter in some cases, especially 502 for keying mechanisms such as Mikey [RFC3830] which use per-SSRC 503 keying. 505 Other than that, the standard security considerations of RTP apply; 506 sending multiple media streams from a single endpoint does not appear 507 to have different security consequences than sending the same number 508 of streams. 510 8. Open Issues 512 At this stage this document contains a number of open issues. The 513 below list tries to summarize the issues: 515 1. Further clarifications on how to handle the RTCP scheduler when 516 sending multiple sources in one compound packet. 518 2. How is the RTCP avg_rtcp_size be calculated when RTCP packets are 519 routinely multiplexed among multiple RTCP senders? 521 3. Do we need to provide a recommendation for unicast session 522 joiners with many sources to not use 0 initial minimal interval 523 from bit-rate burst perspective? 525 9. IANA Considerations 527 No IANA actions needed. 529 10. References 531 10.1. Normative References 533 [I-D.ietf-avtcore-6222bis] 534 Begen, A., Perkins, C., Wing, D., and E. Rescorla, 535 "Guidelines for Choosing RTP Control Protocol (RTCP) 536 Canonical Names (CNAMEs)", draft-ietf-avtcore-6222bis-04 537 (work in progress), June 2013. 539 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 540 Requirement Levels", BCP 14, RFC 2119, March 1997. 542 [RFC3550] Schulzrinne, H., Casner, S., Frederick, R., and V. 543 Jacobson, "RTP: A Transport Protocol for Real-Time 544 Applications", STD 64, RFC 3550, July 2003. 546 [RFC3711] Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K. 547 Norrman, "The Secure Real-time Transport Protocol (SRTP)", 548 RFC 3711, March 2004. 550 [RFC4585] Ott, J., Wenger, S., Sato, N., Burmeister, C., and J. Rey, 551 "Extended RTP Profile for Real-time Transport Control 552 Protocol (RTCP)-Based Feedback (RTP/AVPF)", RFC 4585, July 553 2006. 555 [RFC5124] Ott, J. and E. Carrara, "Extended Secure RTP Profile for 556 Real-time Transport Control Protocol (RTCP)-Based Feedback 557 (RTP/SAVPF)", RFC 5124, February 2008. 559 [RFC5506] Johansson, I. and M. Westerlund, "Support for Reduced-Size 560 Real-Time Transport Control Protocol (RTCP): Opportunities 561 and Consequences", RFC 5506, April 2009. 563 10.2. Informative References 565 [I-D.ietf-avtcore-multi-media-rtp-session] 566 Westerlund, M., Perkins, C., and J. Lennox, "Multiple 567 Media Types in an RTP Session", draft-ietf-avtcore-multi- 568 media-rtp-session-02 (work in progress), February 2013. 570 [I-D.ietf-avtcore-rtp-multi-stream-optimisation] 571 Lennox, J., Westerlund, M., Wu, Q., and C. Perkins, 572 "Sending Multiple Media Streams in a Single RTP Session: 573 Grouping RTCP Reception Statistics and Other Feedback ", 574 draft-ietf-avtcore-rtp-multi-stream-optimisation-00 (work 575 in progress), July 2013. 577 [I-D.ietf-avtcore-rtp-topologies-update] 578 Westerlund, M. and S. Wenger, "RTP Topologies", draft- 579 ietf-avtcore-rtp-topologies-update-00 (work in progress), 580 April 2013. 582 [I-D.ietf-clue-framework] 583 Duckworth, M., Pepperell, A., and S. Wenger, "Framework 584 for Telepresence Multi-Streams", draft-ietf-clue- 585 framework-10 (work in progress), May 2013. 587 [I-D.ietf-mmusic-sdp-bundle-negotiation] 588 Holmberg, C., Alvestrand, H., and C. Jennings, 589 "Multiplexing Negotiation Using Session Description 590 Protocol (SDP) Port Numbers", draft-ietf-mmusic-sdp- 591 bundle-negotiation-04 (work in progress), June 2013. 593 [RFC3611] Friedman, T., Caceres, R., and A. Clark, "RTP Control 594 Protocol Extended Reports (RTCP XR)", RFC 3611, November 595 2003. 597 [RFC3830] Arkko, J., Carrara, E., Lindholm, F., Naslund, M., and K. 598 Norrman, "MIKEY: Multimedia Internet KEYing", RFC 3830, 599 August 2004. 601 Appendix A. Changes From Earlier Versions 603 Note to the RFC-Editor: please remove this section prior to 604 publication as an RFC. 606 A.1. Changes From WG Draft -00 608 o Split the Reporting Group Extension from this draft into draft- 609 ietf-avtcore-rtp-multi-stream-optimization-00. 611 o Added RTCP tuning considerations from draft-ietf-avtcore-multi- 612 media-rtp-session-02. 614 A.2. Changes From Individual Draft -02 616 o Resubmitted as working group draft. 618 o Updated references. 620 A.3. Changes From Individual Draft -01 622 o Merged with draft-wu-avtcore-multisrc-endpoint-adver. 624 o Changed how Reporting Groups are indicated in RTCP, to make it 625 clear which source(s) is the group's reporting sources. 627 o Clarified the rules for when sources can be placed in the same 628 reporting group. 630 o Clarified that mixers and translators need to pass reporting group 631 SDES information if they are forwarding RR and SR traffic from 632 members of a reporting group. 634 A.4. Changes From Individual Draft -00 636 o Added the Reporting Group semantic to explicitly indicate which 637 sources come from a single endpoint, rather than leaving it 638 implicit. 640 o Specified that Reporting Group semantics (as they now are) apply 641 to AVPF and XR, as well as to RR/SR report blocks. 643 o Added a description of the cascaded source-projecting mixer, along 644 with a calculation of its RTCP overhead if reporting groups are 645 not in use. 647 o Gave some guidance on how the flexibility of RTCP randomization 648 allows some freedom in RTCP multiplexing. 650 o Clarified the language of several of the recommendations. 652 o Added an open issue discussing how avg_rtcp_size ought to be 653 calculated for multiplexed RTCP. 655 o Added an open issue discussing how RTCP bandwidths are to be 656 chosen for sessions where source bandwidths greatly differ. 658 Authors' Addresses 660 Jonathan Lennox 661 Vidyo, Inc. 662 433 Hackensack Avenue 663 Seventh Floor 664 Hackensack, NJ 07601 665 US 667 Email: jonathan@vidyo.com 669 Magnus Westerlund 670 Ericsson 671 Farogatan 6 672 SE-164 80 Kista 673 Sweden 675 Phone: +46 10 714 82 87 676 Email: magnus.westerlund@ericsson.com 678 Qin Wu 679 Huawei 680 101 Software Avenue, Yuhua District 681 Nanjing, Jiangsu 210012 682 China 684 Email: sunseawq@huawei.com 685 Colin Perkins 686 University of Glasgow 687 School of Computing Science 688 Glasgow G12 8QQ 689 United Kingdom 691 Email: csp@csperkins.org