idnits 2.17.1 draft-ietf-avtcore-ecn-for-rtp-05.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 : ---------------------------------------------------------------------------- == There are 3 instances of lines with private range IPv4 addresses in the document. If these are generic example addresses, they should be changed to use any of the ranges defined in RFC 6890 (or successor): 192.0.2.x, 198.51.100.x or 203.0.113.x. -- The document has examples using IPv4 documentation addresses according to RFC6890, but does not use any IPv6 documentation addresses. Maybe there should be IPv6 examples, too? Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (October 31, 2011) is 4554 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) == Missing Reference: 'TBA1' is mentioned on line 699, but not defined == Missing Reference: 'TBA2' is mentioned on line 860, but not defined ** Obsolete normative reference: RFC 5245 (Obsoleted by RFC 8445, RFC 8839) ** Obsolete normative reference: RFC 5389 (Obsoleted by RFC 8489) ** Obsolete normative reference: RFC 6336 (Obsoleted by RFC 8839) -- Obsolete informational reference (is this intentional?): RFC 4347 (Obsoleted by RFC 6347) -- Obsolete informational reference (is this intentional?): RFC 4566 (Obsoleted by RFC 8866) -- Obsolete informational reference (is this intentional?): RFC 4960 (Obsoleted by RFC 9260) -- Obsolete informational reference (is this intentional?): RFC 5751 (Obsoleted by RFC 8551) Summary: 3 errors (**), 0 flaws (~~), 4 warnings (==), 6 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group M. Westerlund 3 Internet-Draft I. Johansson 4 Intended status: Standards Track Ericsson 5 Expires: May 3, 2012 C. Perkins 6 University of Glasgow 7 P. O'Hanlon 8 UCL 9 K. Carlberg 10 G11 11 October 31, 2011 13 Explicit Congestion Notification (ECN) for RTP over UDP 14 draft-ietf-avtcore-ecn-for-rtp-05 16 Abstract 18 This memo specifies how Explicit Congestion Notification (ECN) can be 19 used with the Real-time Transport Protocol (RTP) running over UDP, 20 using RTP Control Protocol (RTCP) as a feedback mechanism. It 21 defines a new RTCP Extended Report (XR) block for periodic ECN 22 feedback, a new RTCP transport feedback message for timely reporting 23 of congestion events, and a Session Traversal Utilities for NAT 24 (STUN) extension used in the optional initialization method using 25 Interactive Connectivity Establishment (ICE). Signalling and 26 procedures for negotiation of capabilities and initialization methods 27 are also defined. 29 Status of this Memo 31 This Internet-Draft is submitted in full conformance with the 32 provisions of BCP 78 and BCP 79. 34 Internet-Drafts are working documents of the Internet Engineering 35 Task Force (IETF). Note that other groups may also distribute 36 working documents as Internet-Drafts. The list of current Internet- 37 Drafts is at http://datatracker.ietf.org/drafts/current/. 39 Internet-Drafts are draft documents valid for a maximum of six months 40 and may be updated, replaced, or obsoleted by other documents at any 41 time. It is inappropriate to use Internet-Drafts as reference 42 material or to cite them other than as "work in progress." 44 This Internet-Draft will expire on May 3, 2012. 46 Copyright Notice 48 Copyright (c) 2011 IETF Trust and the persons identified as the 49 document authors. All rights reserved. 51 This document is subject to BCP 78 and the IETF Trust's Legal 52 Provisions Relating to IETF Documents 53 (http://trustee.ietf.org/license-info) in effect on the date of 54 publication of this document. Please review these documents 55 carefully, as they describe your rights and restrictions with respect 56 to this document. Code Components extracted from this document must 57 include Simplified BSD License text as described in Section 4.e of 58 the Trust Legal Provisions and are provided without warranty as 59 described in the Simplified BSD License. 61 Table of Contents 63 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 64 2. Conventions, Definitions and Acronyms . . . . . . . . . . . . 5 65 3. Discussion, Requirements, and Design Rationale . . . . . . . . 6 66 3.1. Requirements . . . . . . . . . . . . . . . . . . . . . . . 7 67 3.2. Applicability . . . . . . . . . . . . . . . . . . . . . . 8 68 3.3. Interoperability . . . . . . . . . . . . . . . . . . . . . 12 69 4. Overview of Use of ECN with RTP/UDP/IP . . . . . . . . . . . . 13 70 5. RTCP Extensions for ECN feedback . . . . . . . . . . . . . . . 15 71 5.1. RTP/AVPF Transport Layer ECN Feedback packet . . . . . . . 16 72 5.2. RTCP XR Report block for ECN summary information . . . . . 19 73 6. SDP Signalling Extensions for ECN . . . . . . . . . . . . . . 21 74 6.1. Signalling ECN Capability using SDP . . . . . . . . . . . 21 75 6.2. RTCP ECN Feedback SDP Parameter . . . . . . . . . . . . . 25 76 6.3. XR Block ECN SDP Parameter . . . . . . . . . . . . . . . . 25 77 6.4. ICE Parameter to Signal ECN Capability . . . . . . . . . . 26 78 7. Use of ECN with RTP/UDP/IP . . . . . . . . . . . . . . . . . . 26 79 7.1. Negotiation of ECN Capability . . . . . . . . . . . . . . 26 80 7.2. Initiation of ECN Use in an RTP Session . . . . . . . . . 27 81 7.3. Ongoing Use of ECN Within an RTP Session . . . . . . . . . 34 82 7.4. Detecting Failures . . . . . . . . . . . . . . . . . . . . 37 83 8. Processing ECN in RTP Translators and Mixers . . . . . . . . . 40 84 8.1. Transport Translators . . . . . . . . . . . . . . . . . . 40 85 8.2. Fragmentation and Reassembly in Translators . . . . . . . 41 86 8.3. Generating RTCP ECN Feedback in Media Transcoders . . . . 43 87 8.4. Generating RTCP ECN Feedback in Mixers . . . . . . . . . . 44 88 9. Implementation considerations . . . . . . . . . . . . . . . . 44 89 10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 45 90 10.1. SDP Attribute Registration . . . . . . . . . . . . . . . . 45 91 10.2. RTP/AVPF Transport Layer Feedback Message . . . . . . . . 45 92 10.3. RTCP Feedback SDP Parameter . . . . . . . . . . . . . . . 46 93 10.4. RTCP XR Report blocks . . . . . . . . . . . . . . . . . . 46 94 10.5. RTCP XR SDP Parameter . . . . . . . . . . . . . . . . . . 46 95 10.6. STUN attribute . . . . . . . . . . . . . . . . . . . . . . 46 96 10.7. ICE Option . . . . . . . . . . . . . . . . . . . . . . . . 46 97 11. Security Considerations . . . . . . . . . . . . . . . . . . . 46 98 12. Examples of SDP Signalling . . . . . . . . . . . . . . . . . . 49 99 12.1. Basic SDP Offer/Answer . . . . . . . . . . . . . . . . . . 49 100 12.2. Declarative Multicast SDP . . . . . . . . . . . . . . . . 51 101 13. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 52 102 14. References . . . . . . . . . . . . . . . . . . . . . . . . . . 52 103 14.1. Normative References . . . . . . . . . . . . . . . . . . . 52 104 14.2. Informative References . . . . . . . . . . . . . . . . . . 53 105 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 55 107 1. Introduction 109 This memo outlines how Explicit Congestion Notification (ECN) 110 [RFC3168] can be used for Real-time Transport Protocol (RTP) 111 [RFC3550] flows running over UDP/IP which use RTP Control Protocol 112 (RTCP) as a feedback mechanism. The solution consists of feedback of 113 ECN congestion experienced markings to the sender using RTCP, 114 verification of ECN functionality end-to-end, and procedures for how 115 to initiate ECN usage. Since the initiation process has some 116 dependencies on the signalling mechanism used to establish the RTP 117 session, a specification for signalling mechanisms using Session 118 Description Protocol (SDP) [RFC4566] is included. 120 ECN is getting attention as a method to minimise the impact of 121 congestion on real-time multimedia traffic. The use of ECN provides 122 a way for the network to send a congestion control signal to a media 123 transport without having to impair the media. Unlike packet loss, 124 ECN signals unambiguously indicate congestion to the transport as 125 quickly as feedback delays allow, and without confusing congestion 126 with losses that might have occurred for other reasons such as 127 transmission errors, packet-size errors, routing errors, badly 128 implemented middleboxes, policy violations and so forth. 130 The introduction of ECN into the Internet requires changes to both 131 the network and transport layers. At the network layer, IP 132 forwarding has to be updated to allow routers to mark packets, rather 133 than discarding them in times of congestion [RFC3168]. In addition, 134 transport protocols have to be modified to inform the sender that ECN 135 marked packets are being received, so it can respond to the 136 congestion. The Transmission Control Protocol (TCP) [RFC3168], 137 Stream Control Transmission Protocol (SCTP) [RFC4960] and Datagram 138 Congestion Control Protocol (DCCP) [RFC4340] have been updated to 139 support ECN, but to date there is no specification how UDP-based 140 transports, such as RTP [RFC3550], can use ECN. This is due to the 141 lack of feedback mechanisms directly in UDP. Instead the signaling 142 control protocol on top of UDP needs to provide that feedback. For 143 RTP that feedback is provided by RTCP. 145 The remainder of this memo is structured as follows. We start by 146 describing the conventions, definitions and acronyms used in this 147 memo in Section 2, and the design rationale and applicability in 148 Section 3. Section 4 gives an overview of how ECN is used with RTP 149 over UDP. RTCP extensions for ECN feedback are defined in Section 5, 150 and SDP signalling extensions in Section 6. The details of how ECN 151 is used with RTP over UDP are defined in Section 7. In Section 8 we 152 describe how ECN is handled in RTP translators and mixers. Section 9 153 discusses some implementation considerations, Section 10 lists IANA 154 considerations, and Section 11 discusses security considerations. 156 2. Conventions, Definitions and Acronyms 158 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 159 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 160 "OPTIONAL" in this document are to be interpreted as described in RFC 161 2119 [RFC2119]. 163 Abbreviations and Definitions: 165 Sender: A sender of RTP packets carrying an encoded media stream. 166 The sender can change how the media transmission is performed by 167 varying the media coding or packetisation. It is one end-point of 168 the ECN control loop. 170 Receiver: A receiver of RTP packets with the intention to consume 171 the media stream. It sends RTCP feedback on the received stream. 172 It is the other end-point of the ECN control loop. 174 ECN Capable Host: A sender or receiver of a media stream that is 175 capable of setting and/or processing ECN marks. 177 ECN Capable Transport (ECT): A transport flow where both sender and 178 receiver are ECN capable hosts. Packets sent by an ECN Capable 179 Transport will be marked as ECT(0) or ECT(1) on transmission. See 180 [RFC3168] for the definition of the ECT(0) and ECT(1) marks. 182 ECN-CE: ECN Congestion Experienced mark (see [RFC3168]). 184 ECN Capable Packets: Packets with ECN mark set to either ECT(0), 185 ECT(1) or ECN-CE. 187 Not-ECT packets: Packets that are not sent by an ECN capable 188 transport, and are not ECN-CE marked. 190 ECN Oblivious Relay: A router or middlebox that treats ECN Capable 191 Packets no differently from Not-ECT packets. 193 ECN Capable Queue: A queue that supports ECN-CE marking of ECN- 194 Capable Packets to indicate congestion. 196 ECN Blocking Middlebox: A middlebox that discards ECN-Capable 197 Packets. 199 ECN Reverting Middlebox: A middlebox that changes ECN-Capable 200 Packets to Not-ECT packets by removing the ECN mark. 202 Note that RTP mixers or translators that operate in such a manner 203 that they terminate or split the ECN control loop will take on the 204 role of receivers or senders. This is further discussed in 205 Section 3.2. 207 3. Discussion, Requirements, and Design Rationale 209 ECN has been specified for use with TCP [RFC3168], SCTP [RFC4960], 210 and DCCP [RFC4340] transports. These are all unicast protocols which 211 negotiate the use of ECN during the initial connection establishment 212 handshake (supporting incremental deployment, and checking if ECN 213 marked packets pass all middleboxes on the path). ECN-CE marks are 214 immediately echoed back to the sender by the receiving end-point 215 using an additional bit in feedback messages, and the sender then 216 interprets the mark as equivalent to a packet loss for congestion 217 control purposes. 219 If RTP is run over TCP, SCTP, or DCCP, it can use the native ECN 220 support provided by those protocols. This memo does not concern 221 itself further with these use cases. However, RTP is more commonly 222 run over UDP. This combination does not currently support ECN, and 223 we observe that it has significant differences from the other 224 transport protocols for which ECN has been specified. These include: 226 Signalling: RTP relies on separate signalling protocols to negotiate 227 parameters before a session can be created, and doesn't include an 228 in-band handshake or negotiation at session set-up time (i.e., 229 there is no equivalent to the TCP three-way handshake in RTP). 231 Feedback: RTP does not explicitly acknowledge receipt of datagrams. 232 Instead, the RTP Control Protocol (RTCP) provides reception 233 quality feedback, and other back channel communication, for RTP 234 sessions. The feedback interval is generally on the order of 235 seconds, rather than once per network RTT (although the RTP/AVPF 236 profile [RFC4585] allows more rapid feedback in most cases). RTCP 237 is also very much oriented around counting packets, which makes 238 byte counting congestion algorithms difficult to utilize. 240 Congestion Response: While it is possible to adapt the transmission 241 of many audio/visual streams in response to network congestion, 242 and such adaptation is required by [RFC3550], the dynamics of the 243 congestion response may be quite different to those of TCP or 244 other transport protocols. 246 Middleboxes: The RTP framework explicitly supports the concept of 247 mixers and translators, which are middleboxes that are involved in 248 media transport functions. 250 Multicast: RTP is explicitly a group communication protocol, and was 251 designed from the start to support IP multicast (primarily Any 252 Source Multicast (ASM) [RFC1112], although a recent extension 253 supports Source Specific Multicast (SSM) [RFC3569] with unicast 254 feedback [RFC5760]). 256 Application Awareness: When ECN support is provided within the 257 transport protocol, the ability of the application to react to 258 congestion is limited, since it has little visibility into the 259 transport layer. By adding support of ECN to RTP using RTCP 260 feedback, the application is made aware of congestion, allowing a 261 wider range of reactions in response to that loss. 263 Counting vs Detecting Congestion: TCP and the protocols derived from 264 it are mainly designed to respond the same whether they experience 265 a burst of congestion indications within one RTT or just one. 266 Whereas real-time applications may be concerned with the amount of 267 congestion experienced, whether it is distributed smoothly or in 268 bursts. When feedback of ECN was added to TCP [RFC3168], the 269 receiver was designed to flip the echo congestion experienced 270 (ECE) flag to 1 for a whole RTT then flop it back to zero. 271 Whereas ECN feedback in RTCP will need to report a count of how 272 much congestion has been experienced within an RTCP reporting 273 period, irrespective of round trip times. 275 These differences will significantly alter the shape of ECN support 276 in RTP-over-UDP compared to ECN support in TCP, SCTP, and DCCP, but 277 do not invalidate the need for ECN support. 279 ECN support is more important for RTP sessions than, for instance, is 280 the case for TCP. This is because the impact of packet loss in real- 281 time audio-visual media flows is highly visible to users. Effective 282 ECN support for RTP flows running over UDP will allow real-time 283 audio-visual applications to respond to the onset of congestion 284 before routers are forced to drop packets, allowing those 285 applications to control how they reduce their transmission rate, and 286 hence media quality, rather than responding to, and trying to conceal 287 the effects of unpredictable packet loss. Furthermore, widespread 288 deployment for ECN and active queue management in routers, should it 289 occur, can potentially reduce unnecessary queueing delays in routers, 290 lowering the round-trip time and benefiting interactive applications 291 of RTP, such as voice telephony. 293 3.1. Requirements 295 Considering ECN, transport protocols supporting ECN, and RTP based 296 applications one can create a set of requirements that must be 297 satisfied to at least some degree if ECN is to used by RTP over UDP. 299 o REQ 1: A mechanism MUST exist to negotiate and initiate the use of 300 ECN for RTP/UDP/IP sessions so that an RTP sender will not send 301 packets with ECT in the IP header unless it knows that all 302 potential receivers will understand any ECN-CE indications they 303 might receive. 305 o REQ 2: A mechanism MUST exist to feed back the reception of any 306 packets that are ECN-CE marked to the packet sender. 308 o REQ 3: The provided mechanism SHOULD minimise the possibility of 309 cheating (either by the sender or receiver). 311 o REQ 4: Some detection and fallback mechanism SHOULD exist to avoid 312 loss of communication due to the attempted usage of ECN in case an 313 intermediate node clears ECT or drops packets that are ECT marked. 315 o REQ 5: Negotiation of ECN SHOULD NOT significantly increase the 316 time taken to negotiate and set-up the RTP session (an extra RTT 317 before the media can flow is unlikely to be acceptable for some 318 use cases). 320 o REQ 6: Negotiation of ECN SHOULD NOT cause media clipping at the 321 start of a session. 323 The following sections describes how these requirements can be met 324 for RTP over UDP. 326 3.2. Applicability 328 The use of ECN with RTP over UDP is dependent on negotiation of ECN 329 capability between the sender and receiver(s), and validation of ECN 330 support in all elements of the network path(s) traversed. RTP is 331 used in a heterogeneous range of network environments and topologies, 332 with various different signalling protocols. The mechanisms defined 333 here make it possible to verify support for ECN in each of these 334 environments, and irrespective of the topology. 336 Due to the need for each RTP sender that intends to use ECN with RTP 337 to track all participants in the RTP session, the sub-sampling of the 338 group membership as specified by "Sampling of the Group Membership in 339 RTP" [RFC2762] MUST NOT be used. 341 The use of ECN is further dependent on a capability of the RTP media 342 flow to react to congestion signalled by ECN marked packets. 343 Depending on the application, media codec, and network topology, this 344 adaptation can occur in various forms and at various nodes. As an 345 example, the sender can change the media encoding, or the receiver 346 can change the subscription to a layered encoding, or either reaction 347 can be accomplished by a transcoding middlebox. RFC 5117 identifies 348 seven topologies in which RTP sessions may be configured, and which 349 may affect the ability to use ECN: 351 Topo-Point-to-Point: This utilises standard unicast flows. ECN may 352 be used with RTP in this topology in an analogous manner to its 353 use with other unicast transport protocols, with RTCP conveying 354 ECN feedback messages. 356 Topo-Multicast: This is either an any source multicast (ASM) group 357 [RFC3569] with potentially several active senders and multicast 358 RTCP feedback, or a source specific multicast (SSM) group 359 [RFC4607] with a single distribution source and unicast RTCP 360 feedback from receivers. RTCP is designed to scale to large group 361 sizes while avoiding feedback implosion (see Section 6.2 of 362 [RFC3550], [RFC4585], and [RFC5760]), and can be used by a sender 363 to determine if all its receivers, and the network paths to those 364 receivers, support ECN (see Section 7.2). It is somewhat more 365 difficult to determine if all network paths from all senders to 366 all receivers support ECN. Accordingly, we allow ECN to be used 367 by an RTP sender using multicast UDP provided the sender has 368 verified that the paths to all its known receivers support ECN, 369 and irrespective of whether the paths from other senders to their 370 receivers support ECN ("all its known receivers" are all the SSRCs 371 that the RTP sender has received RTP or RTCP from the last five 372 reporting intervals, i.e., they have not timed out). Note that 373 group membership may change during the lifetime of a multicast RTP 374 session, potentially introducing new receivers that are not ECN 375 capable or have a path that doesn't support ECN. Senders must use 376 the mechanisms described in Section 7.4 to check that all 377 receivers, and the network paths traversed to reach those 378 receivers, continue to support ECN, and they need to fallback to 379 non-ECN use if any receivers join that do not. 381 SSM groups that uses unicast RTCP feedback [RFC5760] do need a few 382 extra considerations. This topology can have multiple media 383 senders that provides traffic to the distribution source (DS) and 384 are separated from the DS. There can also be multiple feedback 385 targets. The requirement for using ECN for RTP in this topology 386 is that the media sender must be provided the feedback from the 387 receivers, it may be in aggregated form from the feedback targets. 388 We will not mention this SSM use case in the below text 389 specifically, but when actions are required by the media source, 390 they do apply also to case of SSM where the RTCP feedback goes to 391 the Feedback Target. 393 This specification do support multicast, but has primarily 394 considered smaller multicast groups and is not optimized for 395 larger groups and their needs. 397 Topo-Translator: An RTP translator is an RTP-level middlebox that is 398 invisible to the other participants in the RTP session (although 399 it is usually visible in the associated signalling session). 400 There are two types of RTP translator: those that do not modify 401 the media stream, and are concerned with transport parameters, for 402 example a multicast to unicast gateway; and those that do modify 403 the media stream, for example transcoding between different media 404 codecs. A single RTP session traverses the translator, and the 405 translator must rewrite RTCP messages passing through it to match 406 the changes it makes to the RTP data packets. A legacy, ECN- 407 unaware, RTP translator is expected to ignore the ECN bits on 408 received packets, and to set the ECN bits to not-ECT when sending 409 packets, so causing ECN negotiation on the path containing the 410 translator to fail (any new RTP translator that does not wish to 411 support ECN may do so similarly). An ECN aware RTP translator may 412 act in one of three ways: 414 * If the translator does not modify the media stream, it should 415 copy the ECN bits unchanged from the incoming to the outgoing 416 datagrams, unless it is overloaded and experiencing congestion, 417 in which case it may mark the outgoing datagrams with an ECN-CE 418 mark. Such a translator passes RTCP feedback unchanged. See 419 Section 8.1. 421 * If the translator modifies the media stream to combine or split 422 RTP packets, but does not otherwise transcode the media, it 423 must manage the ECN bits in a way analogous to that described 424 in Section 5.3 of [RFC3168], see Section 8.2 for details. 426 * If the translator is a media transcoder, or otherwise modifies 427 the content of the media stream, the output RTP media stream 428 may have radically different characteristics than the input RTP 429 media stream. Each side of the translator must then be 430 considered as a separate transport connection, with its own ECN 431 processing. This requires the translator interpose itself into 432 the ECN negotiation process, effectively splitting the 433 connection into two parts with their own negotiation. Once 434 negotiation has been completed, the translator must generate 435 RTCP ECN feedback back to the source based on its own 436 reception, and must respond to RTCP ECN feedback received from 437 the receiver(s) (see Section 8.3). 439 It is recognised that ECN and RTCP processing in an RTP translator 440 that modifies the media stream is non-trivial. 442 Topo-Mixer: A mixer is an RTP-level middlebox that aggregates 443 multiple RTP streams, mixing them together to generate a new RTP 444 stream. The mixer is visible to the other participants in the RTP 445 session, and is also usually visible in the associated signalling 446 session. The RTP flows on each side of the mixer are treated 447 independently for ECN purposes, with the mixer generating its own 448 RTCP ECN feedback, and responding to ECN feedback for data it 449 sends. Since unicast transport between the mixer and any end- 450 point are treated independently, it would seem reasonable to allow 451 the transport on one side of the mixer to use ECN, while the 452 transport on the other side of the mixer is not ECN capable, if 453 this is desired. See Section 8.4 for details in how mixers should 454 process ECN. 456 Topo-Video-switch-MCU: A video switching MCU receives several RTP 457 flows, but forwards only one of those flows onwards to the other 458 participants at a time. The flow that is forwarded changes during 459 the session, often based on voice activity. Since only a subset 460 of the RTP packets generated by a sender are forwarded to the 461 receivers, a video switching MCU can break ECN negotiation (the 462 success of the ECN negotiation may depend on the voice activity of 463 the participant at the instant the negotiation takes place - shout 464 if you want ECN). It also breaks congestion feedback and 465 response, since RTP packets are dropped by the MCU depending on 466 voice activity rather than network congestion. This topology is 467 widely used in legacy products, but is NOT RECOMMENDED for new 468 implementations and SHALL NOT be used with ECN. 470 Topo-RTCP-terminating-MCU: In this scenario, each participant runs 471 an RTP point-to-point session between itself and the MCU. Each of 472 these sessions is treated independently for the purposes of ECN 473 and RTCP feedback, potentially with some using ECN and some not. 475 Topo-Asymmetric: It is theoretically possible to build a middlebox 476 that is a combination of an RTP mixer in one direction and an RTP 477 translator in the other. To quote RFC 5117 "This topology is so 478 problematic and it is so easy to get the RTCP processing wrong, 479 that it is NOT RECOMMENDED to implement this topology." 481 These topologies may be combined within a single RTP session. 483 The ECN mechanism defined in this memo is applicable to both sender 484 and receiver controlled congestion algorithms. The mechanism ensures 485 that both senders and receivers will know about ECN-CE markings and 486 any packet losses. Thus the actual decision point for the congestion 487 control is not relevant. This is a great benefit as the rate of an 488 RTP session can be varied in a number of ways, for example a unicast 489 media sender might use TFRC [RFC5348] or some other algorithm, while 490 a multicast session could use a sender based scheme adapting to the 491 lowest common supported rate, or a receiver driven mechanism using 492 layered coding to support more heterogeneous paths. 494 To ensure timely feedback of ECN-CE marked packets when needed, this 495 mechanism requires support for the RTP/AVPF profile [RFC4585] or any 496 of its derivatives, such as RTP/SAVPF [RFC5124]. The standard RTP/ 497 AVP profile [RFC3551] does not allow any early or immediate 498 transmission of RTCP feedback, and has a minimal RTCP interval whose 499 default value (5 seconds) is many times the normal RTT between sender 500 and receiver. 502 3.3. Interoperability 504 The interoperability requirements for this specification are that 505 there is at least one common interoperability point for all 506 implementations. Since initialization using RTP and RTCP 507 (Section 7.2.1) is the one method that works in all cases, although 508 is not optimal for all uses, it is selected as mandatory to implement 509 this initialisation method. This method requires both the RTCP XR 510 extension and the ECN feedback format, which require the RTP/AVPF 511 profile to ensure timely feedback. 513 When one considers all the uses of ECN for RTP it is clear that there 514 exist congestion control mechanisms that are receiver driven only 515 (Section 7.3.3). These congestion control mechanism do not require 516 timely feedback of congestion events to the sender. If such a 517 congestion control mechanism is combined with an initialization 518 method that also doesn't require timely feedback using RTCP, like the 519 leap of faith or the ICE based method then neither the ECN feedback 520 format nor the RTP/AVPF profile would appear to be needed. However, 521 fault detection can be greatly improved by using receiver side 522 detection (Section 7.4.1) and early reporting of such cases using the 523 ECN feedback mechanism. 525 For interoperability we mandate the implementation of the RTP/AVPF 526 profile, with both RTCP extensions and the necessary signalling to 527 support a common operations mode. This specification recommends the 528 use of RTP/AVPF in all cases as negotiation of the common 529 interoperability point requires RTP/AVPF, mixed negotiation of RTP/ 530 AVP and RTP/AVPF depending on other SDP attributes in the same media 531 block is difficult, and the fact that fault detection can be improved 532 when using RTP/AVPF. 534 The use of the ECN feedback format is also recommended, but cases 535 exist where its use is not required due to no need for timely 536 feedback. These will be explicitly noted using the term "no timely 537 feedback required", and generally occur in combination with receiver 538 driven congestion control, and with the leap-of-faith and ICE-based 539 initialization methods. We also note that any receiver driven 540 congestion control solution that still requires RTCP for signalling 541 of any adaptation information to the sender will still require RTP/ 542 AVPF for timeliness. 544 4. Overview of Use of ECN with RTP/UDP/IP 546 The solution for using ECN with RTP over UDP/IP consists of four 547 different pieces that together make the solution work: 549 1. Negotiation of the capability to use ECN with RTP/UDP/IP 551 2. Initiation and initial verification of ECN capable transport 553 3. Ongoing use of ECN within an RTP session 555 4. Handling of dynamic behavior through failure detection, 556 verification and fallback 558 Before an RTP session can be created, a signalling protocol is used 559 to negotiate or at least configure session parameters (see 560 Section 7.1). In some topologies the signalling protocol can also be 561 used to discover the other participants. One of the parameters that 562 must be agreed is the capability of a participant to support ECN. 563 Note that all participants having the capability of supporting ECN 564 does not necessarily imply that ECN is usable in an RTP session, 565 since there may be middleboxes on the path between the participants 566 which don't pass ECN-marked packets (for example, a firewall that 567 blocks traffic with the ECN bits set). This document defines the 568 information that needs to be negotiated, and provides a mapping to 569 SDP for use in both declarative and offer/answer contexts. 571 When a sender joins a session for which all participants claim to 572 support ECN, it must verify if that support is usable. There are 573 three ways in which this verification can be done: 575 o The sender may generate a (small) subset of its RTP data packets 576 with the ECN field set to ECT(0) or ECT(1). Each receiver will 577 then send an RTCP feedback packet indicating the reception of the 578 ECT marked RTP packets. Upon reception of this feedback from each 579 receiver it knows of, the sender can consider ECN functional for 580 its traffic. Each sender does this verification independently. 581 When a new receiver joins an existing RTP session, it will send 582 RTCP reports in the usual manner. If those RTCP reports include 583 ECN information, verification will have succeeded and sources can 584 continue to send ECT packets. If not, verification fails and each 585 sender MUST stop using ECN (see Section 7.2.1 for details). 587 o Alternatively, ECN support can be verified during an initial end- 588 to-end STUN exchange (for example, as part of ICE connection 589 establishment). After having verified connectivity without ECN 590 capability an extra STUN exchange, this time with the ECN field 591 set to ECT(0) or ECT(1), is performed on the candidate path that 592 is about to be used. If successful the path's capability to 593 convey ECN marked packets is verified. A new STUN attribute is 594 defined to convey feedback that the ECT marked STUN request was 595 received (see Section 7.2.2), along with an ICE signalling option 596 (Section 6.4) to indicate that the check is to be performed. 598 o Thirdly, the sender may make a leap of faith that ECN will work. 599 This is only recommended for applications that know they are 600 running in controlled environments where ECN functionality has 601 been verified through other means. In this mode it is assumed 602 that ECN works, and the system reacts to failure indicators if the 603 assumption proved wrong. The use of this method relies on a high 604 confidence that ECN operation will be successful, or an 605 application where failure is not serious. The impact on the 606 network and other users must be considered when making a leap of 607 faith, so there are limitations on when this method is allowed 608 (see Section 7.2.3). 610 The first mechanism, using RTP with RTCP feedback, has the advantage 611 of working for all RTP sessions, but the disadvantages of potential 612 clipping if ECN marked RTP packets are discarded by middleboxes, and 613 slow verification of ECN support. The STUN-based mechanism is faster 614 to verify ECN support, but only works in those scenarios supported by 615 end-to-end STUN, such as within an ICE exchange. The third one, 616 leap-of-faith, has the advantage of avoiding additional tests or 617 complexities and enabling ECN usage from the first media packet. The 618 downside is that if the end-to-end path contains middleboxes that do 619 not pass ECN, the impact on the application can be severe: in the 620 worst case, all media could be lost if a middlebox that discards ECN 621 marked packets is present. A less severe effect, but still requiring 622 reaction, is the presence of a middlebox that re-marks ECT marked 623 packets to non-ECT, possibly marking packets with an ECN-CE mark as 624 non-ECT. This could result in increased levels of congestion due to 625 non-responsiveness, and impact media quality as applications end up 626 relying on packet loss as an indication of congestion. 628 Once ECN support has been verified (or assumed) to work for all 629 receivers, a sender marks all its RTP packets as ECT packets, while 630 receivers rapidly feed back reports on any ECN-CE marks to the sender 631 using RTCP in RTP/AVPF immediate or early feedback mode, unless no 632 timely feedback is required. Each feedback report indicates the 633 receipt of new ECN-CE marks since the last ECN feedback packet, and 634 also counts the total number of ECN-CE marked packets as a cumulative 635 sum. This is the mechanism to provide the fastest possible feedback 636 to senders about ECN-CE marks. On receipt of an ECN-CE marked 637 packet, the system must react to congestion as-if packet loss has 638 been reported. Section 7.3 describes the ongoing use of ECN within 639 an RTP session. 641 This rapid feedback is not optimised for reliability, so another 642 mechanism, RTCP XR ECN summary reports, is used to ensure more 643 reliable, but less timely, reporting of the ECN information. The ECN 644 summary report contains the same information as the ECN feedback 645 format, only packed differently for better efficiency with reports 646 for many sources. It is sent in a compound RTCP packet, along with 647 regular RTCP reception reports. By using cumulative counters for 648 observed ECN-CE, ECT, not-ECT, packet duplication, and packet loss 649 the sender can determine what events have happened since the last 650 report, independently of any RTCP packets having been lost. 652 RTCP reports MUST NOT be ECT marked, since ECT marked traffic may be 653 dropped if the path is not ECN compliant. RTCP is used to provide 654 feedback about what has been transmitted and what ECN markings that 655 are received, so it is important that it is received in cases when 656 ECT marked traffic is not getting through. 658 There are numerous reasons why the path the RTP packets take from the 659 sender to the receiver may change, e.g., mobility, link failure 660 followed by re-routing around it. Such an event may result in the 661 packet being sent through a node that is ECN non-compliant, thus re- 662 marking or dropping packets with ECT set. To prevent this from 663 impacting the application for longer than necessary, the operation of 664 ECN is constantly monitored by all senders (Section 7.4). Both the 665 RTCP XR ECN summary reports and the ECN feedback packets allow the 666 sender to compare the number of ECT(0), ECT(1), and non-ECT marked 667 packets received with the number that were sent, while also reporting 668 ECN-CE marked and lost packets. If these numbers do not agree, it 669 can be inferred that the path does not reliably pass ECN-marked 670 packets. A sender detecting a possible ECN non-compliance issue 671 should then stop sending ECT marked packets to determine if that 672 allows the packets to be correctly delivered. If the issues can be 673 connected to ECN, then ECN usage is suspended. 675 5. RTCP Extensions for ECN feedback 677 This memo defines two new RTCP extensions: one RTP/AVPF [RFC4585] 678 transport layer feedback format for urgent ECN information, and one 679 RTCP XR [RFC3611] ECN summary report block type for regular reporting 680 of the ECN marking information. 682 5.1. RTP/AVPF Transport Layer ECN Feedback packet 684 This RTP/AVPF transport layer feedback format is intended for use in 685 RTP/AVPF early or immediate feedback modes when information needs to 686 urgently reach the sender. Thus its main use is to report reception 687 of an ECN-CE marked RTP packet so that the sender may perform 688 congestion control, or to speed up the initiation procedures by 689 rapidly reporting that the path can support ECN-marked traffic. The 690 feedback format is also defined with reduced size RTCP [RFC5506] in 691 mind, where RTCP feedback packets may be sent without accompanying 692 Sender or Receiver Reports that would contain the Extended Highest 693 Sequence number and the accumulated number of packet losses. Both 694 are important for ECN to verify functionality and keep track of when 695 CE marking does occur. 697 The RTP/AVPF transport layer feedback packet starts with the common 698 header defined by the RTP/AVPF profile [RFC4585] which is reproduced 699 in Figure 1. The FMT field takes the value [TBA1] to indicate that 700 the Feedback Control Information (FCI) contains ECN Feedback report, 701 as defined in Figure 2. 703 0 1 2 3 704 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 705 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 706 |V=2|P| FMT=TBA1| PT=RTPFB=205 | length | 707 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 708 | SSRC of packet sender | 709 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 710 | SSRC of media source | 711 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 712 : Feedback Control Information (FCI) : 713 : : 715 Figure 1: RTP/AVPF Common Packet Format for Feedback Messages 717 0 1 2 3 718 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 719 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 720 | Extended Highest Sequence Number | 721 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 722 | ECT (0) Counter | 723 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 724 | ECT (1) Counter | 725 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 726 | ECN-CE Counter | not-ECT Counter | 727 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 728 | Loss Packet Counter | Duplication Counter | 729 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 731 Figure 2: ECN Feedback Report Format 733 The ECN Feedback Report contains the following fields: 735 Extended Highest Sequence Number: The 32-bit Extended highest 736 sequence number received, as defined by [RFC3550]. Indicates the 737 highest RTP sequence number to which this report relates. 739 ECT(0) Counter: The 32-bit cumulative number of RTP packets with 740 ECT(0) received from this SSRC. 742 ECT(1) Counter: The 32-bit cumulative number of RTP packets with 743 ECT(1) received from this SSRC. 745 ECN-CE Counter: The cumulative number of RTP packets received from 746 this SSRC since the receiver joined the RTP session that were 747 ECN-CE marked, including ECN-CE marks in any duplicate packets. 748 The receiver should keep track of this value using a local 749 representation that is at least 32-bits, and only include the 16- 750 bits with least significance. In other words, the field will wrap 751 if more than 65535 ECN-CE marked packets have been received. 753 not-ECT Counter: The cumulative number of RTP packets received from 754 this SSRC since the receiver joined the RTP session that had an 755 ECN field value of not-ECT. The receiver should keep track of 756 this value using a local representation that is at least 32-bits, 757 and only include the 16-bits with least significance. In other 758 words, the field will wrap if more than 65535 not-ECT packets have 759 been received. 761 Lost Packets Counter: The cumulative number of RTP packets that the 762 receiver expected to receive minus the number of packets it 763 actually received that are not a duplicate of an already received 764 packet, from this SSRC since the receiver joined the RTP session. 766 Note that packets that arrive late are not counted as lost. The 767 receiver should keep track of this value using a local 768 representation that is at least 32-bits, and only include the 16- 769 bits with least significance. In other words, the field will wrap 770 if more than 65535 packets are lost. 772 Duplication Counter: The cumulative number of RTP packets received 773 that are a duplicate of an already received packet from this SSRC 774 since the receiver joined the RTP session. The receiver should 775 keep track of this value using a local representation that is at 776 least 32-bits, and only include the 16-bits with least 777 significance. In other words, the field will wrap if more than 778 65535 duplicate packets have been received. 780 All fields in the ECN Feedback Report are unsigned integers in 781 network byte order. Each ECN Feedback Report corresponds to a single 782 RTP source (SSRC). Multiple sources can be reported by including 783 multiple ECN Feedback Reports packets in an compound RTCP packet. 785 The counters SHALL be initiated to 0 for each new SSRC received. 786 This to enable detection of ECN-CE marks or Packet loss on the 787 initial report from a specific participant. 789 The use of at least 32-bit counters allows even extremely high packet 790 volume applications to not have wrapping of counters within any 791 timescale close to the RTCP reporting intervals. However, 32-bits 792 are not sufficiently large to disregard the fact that wrappings may 793 happen during the life time of a long-lived RTP session. Thus 794 handling of wrapping of these counters MUST be supported. It is 795 recommended that implementations uses local representation of these 796 counters that are longer than 32-bits to enable easy handling of 797 wraps. 799 There is a difference in packet duplication reports between the 800 packet loss counter that is defined in the Receiver Report Block 801 [RFC3550] and that defined here. To avoid holding state for what RTP 802 sequence numbers have been received, [RFC3550] specifies that one can 803 count packet loss by counting the number of received packets and 804 comparing it to the number of packets expected. As a result a packet 805 duplication can hide a packet loss. However, when populating the ECN 806 Feedback report, a receiver needs to track the sequence numbers 807 actually received and count duplicates and packet loss separately to 808 provide a more reliable indication. Reordering may however still 809 result in that packet loss is reported in one report and then removed 810 in the next. 812 The ECN-CE counter is robust for packet duplication. Adding each 813 received ECN-CE marked packet to the counter is not an issue, in fact 814 it is required to ensure complete tracking of the ECN state. If one 815 of the clones was ECN-CE marked that is still an indication of 816 congestion. Packet duplication has potential impact on the ECN 817 verification and thus there is a need to count the duplicates. 819 5.2. RTCP XR Report block for ECN summary information 821 This unilateral XR report block combined with RTCP SR or RR report 822 blocks carries the same information as the ECN Feedback Report and is 823 be based on the same underlying information. However, the ECN 824 Feedback Report is intended to report on an ECN-CE mark as soon as 825 possible, while this extended report is for the regular RTCP 826 reporting and continuous verification of the ECN functionality end- 827 to-end. 829 The ECN Summary report block consists of one RTCP XR report block 830 header, shown in Figure 3 followed by one or more ECN summary report 831 data blocks, as defined in Figure 4. 833 0 1 2 3 834 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 835 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 836 | BT=[TBA2] | Reserved | Block Length | 837 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 839 Figure 3: RTCP XR Report Header 841 0 1 2 3 842 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 843 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 844 | SSRC of Media Sender | 845 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 846 | ECT (0) Counter | 847 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 848 | ECT (1) Counter | 849 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 850 | ECN-CE Counter | not-ECT Counter | 851 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 852 | Loss Packet Counter | Duplication Counter | 853 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 855 Figure 4: RTCP XR ECN Summary Report 857 The RTCP XR ECN Summary Report contains the following fields: 859 BT: Block Type identifying the ECN summary report block. Value is 860 [TBA2]. 862 Reserved: All bits SHALL be set to 0 on transmission and ignored on 863 reception. 865 Block Length: The length of the report block. Used to indicate the 866 number of report data blocks present in the ECN summary report. 867 This length will be 5*n, where n is the number of ECN summary 868 report blocks, since blocks are a fixed size. The block length 869 MAY be zero if there is nothing to report. Receivers MUST discard 870 reports where the block length is not a multiple of five octets, 871 since these cannot be valid. 873 SSRC of Media Sender: The SSRC identifying the media sender this 874 report is for. 876 ECT(0) Counter: as in Section 5.1. 878 ECT(1) Counter: as in Section 5.1. 880 ECN-CE Counter: as in Section 5.1. 882 not-ECT Counter: as in Section 5.1. 884 Loss Packet Counter: as in Section 5.1. 886 Duplication Counter: as in Section 5.1. 888 The Extended Highest Sequence number counter for each SSRC is not 889 present in RTCP XR report, in contrast to the feedback version. The 890 reason is that this summary report will rely on the information sent 891 in the Sender Report (SR) or Receiver Report (RR) blocks part of the 892 same RTCP compound packet. The Extended Highest Sequence number is 893 available from the SR or RR. 895 All the SSRCs that are present in the SR or RR SHOULD also be 896 included in the RTCP XR ECN summary report. In cases where the 897 number of senders are so large that the combination of SR/RR and the 898 ECN summary for all the senders exceed the MTU, then only a subset of 899 the senders SHOULD be included so that the reports for the subset 900 fits within the MTU. The subsets SHOULD be selected round-robin 901 across multiple intervals so that all sources are periodically 902 reported. In case there are no SSRCs that currently are counted as 903 senders in the session, the report block SHALL still be sent with no 904 report block entry and a zero report block length to continuously 905 indicate to the other participants the receiver capability to report 906 ECN information. 908 6. SDP Signalling Extensions for ECN 910 This section defines a number of SDP signalling extensions used in 911 the negotiation of the ECN for RTP support when using SDP. This 912 includes one SDP attribute "ecn-capable-rtp" that negotiates the 913 actual operation of ECN for RTP. Two SDP signalling parameters are 914 defined to indicate the use of the RTCP XR ECN summary block and the 915 RTP/AVPF feedback format for ECN. One ICE option SDP representation 916 is also defined. 918 6.1. Signalling ECN Capability using SDP 920 One new SDP attribute, "a=ecn-capable-rtp", is defined. This is a 921 media level attribute, and MUST NOT be used at the session level. It 922 is not subject to the character set chosen. The aim of this 923 signalling is to indicate the capability of the sender and receivers 924 to support ECN, and to negotiate the method of ECN initiation to be 925 used in the session. The attribute takes a list of initiation 926 methods, ordered in decreasing preference. The defined values for 927 the initiation method are: 929 rtp: Using RTP and RTCP as defined in Section 7.2.1. 931 ice: Using STUN within ICE as defined in Section 7.2.2. 933 leap: Using the leap of faith method as defined in Section 7.2.3. 935 Further methods may be specified in the future, so unknown methods 936 MUST be ignored upon reception. 938 In addition, a number of OPTIONAL parameters may be included in the 939 "a=ecn-capable-rtp" attribute as follows: 941 mode: This parameter signals the endpoint's capability to set and 942 read ECN marks in UDP packets. An examination of various 943 operating systems has shown that end-system support for ECN 944 marking of UDP packets may be symmetric or asymmetric. By this we 945 mean that some systems may allow end points to set the ECN bits in 946 an outgoing UDP packet but not read them, while others may allow 947 applications to read the ECN bits but not set them. This 948 either/or case may produce an asymmetric support for ECN and thus 949 should be conveyed in the SDP signalling. The "mode=setread" 950 state is the ideal condition where an endpoint can both set and 951 read ECN bits in UDP packets. The "mode=setonly" state indicates 952 that an endpoint can set the ECT bit, but cannot read the ECN bits 953 from received UDP packets to determine if upstream congestion 954 occurred. The "mode=readonly" state indicates that the endpoint 955 can read the ECN bits to determine if congestion has occurred for 956 incoming packets, but it cannot set the ECT bits in outgoing UDP 957 packets. When the "mode=" parameter is omitted it is assumed that 958 the node has "setread" capabilities. This option can provide for 959 an early indication that ECN cannot be used in a session. This 960 would be case when both the offerer and answerer set the "mode=" 961 parameter to "setonly" or both set it to "readonly". 963 ect: This parameter makes it possible to express the preferred ECT 964 marking. This is either "random", "0", or "1", with "0" being 965 implied if not specified. The "ect" parameter describes a 966 receiver preference, and is useful in the case where the receiver 967 knows it is behind a link using IP header compression, the 968 efficiency of which would be seriously disrupted if it were to 969 receive packets with randomly chosen ECT marks. It is RECOMMENDED 970 that ECT(0) marking be used. 972 The ABNF [RFC5234] grammar for the "a=ecn-capable-rtp" attribute is 973 shown in Figure 5. 974 ecn-attribute = "a=ecn-capable-rtp:" SP init-list [SP parm-list] 975 init-list = init-value *("," init-value) 976 init-value = "rtp" / "ice" / "leap" / init-ext 977 init-ext = token 978 parm-list = parm-value *(";" SP parm-value) 979 parm-value = mode / ect / parm-ext 980 mode = "mode=" ("setonly" / "setread" / "readonly") 981 ect = "ect=" ("0" / "1" / "random") 982 parm-ext = parm-name "=" parm-value-ext 983 parm-name = token 984 parm-value-ext = token / quoted-string 985 quoted-string = DQUOTE *qdtext DQUOTE 986 qdtext = %x20-21 / %x23-7E / %x80-FF 987 ; any 8-bit ASCII except <"> 989 ; external references: 990 ; token: from RFC 4566 991 ; SP and DQUOTE from RFC 5234 993 Figure 5: ABNF Grammar for the "a=ecn-capable-rtp" attribute 995 6.1.1. Use of "a=ecn-capable-rtp:" with the Offer/Answer Model 997 When SDP is used with the offer/answer model [RFC3264], the party 998 generating the SDP offer MUST insert an "a=ecn-capable-rtp" attribute 999 into the media section of the SDP offer of each RTP session for which 1000 it wishes to use ECN. The attribute includes one or more ECN 1001 initiation methods in a comma separated list in decreasing order of 1002 preference, with any number of optional parameters following. The 1003 answering party compares the list of initiation methods in the offer 1004 with those it supports in order of preference. If there is a match, 1005 and if the receiver wishes to attempt to use ECN in the session, it 1006 includes an "a=ecn-capable-rtp" attribute containing its single 1007 preferred choice of initiation method, and any optional parameters, 1008 in the media sections of the answer. If there is no matching 1009 initiation method capability, or if the receiver does not wish to 1010 attempt to use ECN in the session, it does not include an "a=ecn- 1011 capable-rtp" attribute in its answer. If the attribute is removed in 1012 the answer then ECN MUST NOT be used in any direction for that media 1013 flow. If there are initialization methods that are unknown, they 1014 MUST be ignored on reception and MUST NOT be included in an answer. 1016 The endpoints' capability to set and read ECN marks, as expressed by 1017 the optional "mode=" parameter, determines whether ECN support can be 1018 negotiated for flows in one or both directions: 1020 o If the "mode=setonly" parameter is present in the "a=ecn-capable- 1021 rtp" attribute of the offer and the answering party is also 1022 "mode=setonly", then there is no common ECN capability, and the 1023 answer MUST NOT include the "a=ecn-capable-rtp" attribute. 1024 Otherwise, if the offer is "mode=setonly" then ECN may only be 1025 initiated in the direction from the offering party to the 1026 answering party. 1028 o If the "mode=readonly" parameter is present in the "a=ecn-capable- 1029 rtp" attribute of the offer and the answering party is 1030 "mode=readonly", then there is no common ECN capability, and the 1031 answer MUST NOT include the "a=ecn-capable-rtp" attribute. 1032 Otherwise, if the offer is "mode=readonly" then ECN may only be 1033 initiated in the direction from the answering party to the 1034 offering party. 1036 o If the "mode=setread" parameter is present in the "a=ecn-capable- 1037 rtp" attribute of the offer and the answering party is "setonly", 1038 then ECN may only be initiated in the direction from the answering 1039 party to the offering party. If the offering party is 1040 "mode=setread" but the answering party is "mode=readonly", then 1041 ECN may only be initiated in the direction from the offering party 1042 to the answering party. If both offer and answer are 1043 "mode=setread", then ECN may be initiated in both directions. 1044 Note that "mode=setread" is implied by the absence of a "mode=" 1045 parameter in the offer or the answer. 1047 o An offer that does not include a "mode=" parameter MUST be treated 1048 as-if a "mode=setread" parameter had been included. 1050 In an RTP session using multicast and ECN, participants that intend 1051 to send RTP packets SHOULD support setting ECT marks in RTP packets 1052 (i.e., should be "mode=setonly" or "mode=setread"). Participants 1053 receiving data need the capability to read ECN marks on incoming 1054 packets. It is important that receivers can read ECN marks (are 1055 "mode=readonly" or "mode=setread"), since otherwise no sender in the 1056 multicast session will be able to enable ECN. Accordingly, receivers 1057 that are "mode=setonly" SHOULD NOT join multicast RTP sessions that 1058 use ECN. If session participants that are not aware of the ECN for 1059 RTP signalling are invited to a multicast session, and simply ignore 1060 the signalling attribute, the other party in the offer/answer 1061 exchange SHOULD terminate the SDP dialogue so that the participant 1062 leaves the session. 1064 The "ect=" parameter in the "a=ecn-capable-rtp" attribute is set 1065 independently in the offer and the answer. Its value in the offer 1066 indicates a preference for the sending behaviour of the answering 1067 party, and its value in the answer indicates a sending preference for 1068 the behaviour of the offering party. It will be the senders choice 1069 to honour the receivers preference for what to receive or not. In 1070 multicast sessions, all senders SHOULD set the ECT marks using the 1071 value declared in the "ect=" parameter. 1073 Unknown optional parameters MUST be ignored on reception, and MUST 1074 NOT be included in the answer. That way new parameters may be 1075 introduced and verified to be supported by the other end-point by 1076 having them include it in any answer. 1078 6.1.2. Use of "a=ecn-capable-rtp:" with Declarative SDP 1080 When SDP is used in a declarative manner, for example in a multicast 1081 session using the Session Announcement Protocol (SAP, [RFC2974]), 1082 negotiation of session description parameters is not possible. The 1083 "a=ecn-capable-rtp" attribute MAY be added to the session description 1084 to indicate that the sender will use ECN in the RTP session. The 1085 attribute MUST include a single method of initiation. Participants 1086 MUST NOT join such a session unless they have the capability to 1087 receive ECN-marked UDP packets, implement the method of initiation, 1088 and can generate RTCP ECN feedback. The mode parameter MAY also be 1089 included in declarative usage, to indicate the minimal capability is 1090 required by the consumer of the SDP. So for example in a SSM session 1091 the participants configured with a particular SDP will all be in a 1092 media receive only mode, thus mode=readonly will work as the 1093 capability of reporting on the ECN markings in the received is what 1094 is required. However, using "mode=readonly" also in ASM sessions is 1095 reasonable, unless all senders are required to attempt to use ECN for 1096 their outgoing RTP data traffic, in which case the mode needs to be 1097 set to "setread". 1099 6.1.3. General Use of the "a=ecn-capable-rtp:" Attribute 1101 The "a=ecn-capable-rtp" attribute MAY be used with RTP media sessions 1102 using UDP/IP transport. It MUST NOT be used for RTP sessions using 1103 TCP, SCTP, or DCCP transport, or for non-RTP sessions. 1105 As described in Section 7.3.3, RTP sessions using ECN require rapid 1106 RTCP ECN feedback, unless timely feedback is not required due to a 1107 receiver driven congestion control. To ensure that the sender can 1108 react to ECN-CE marked packets timely feedback is usually required. 1109 Thus, the use of the Extended RTP Profile for RTCP-Based Feedback 1110 (RTP/AVPF) [RFC4585] or other profile that inherits RTP/AVPF's 1111 signalling rules, MUST be signalled unless timely feedback is not 1112 required. If timely feedback is not required it is still RECOMMENDED 1113 to use RTP/AVPF. The signalling of an RTP/AVPF based profile is 1114 likely to be required even if the preferred method of initialization 1115 and the congestion control does not require timely feedback, as the 1116 common interoperable method is likely to be signalled or the improved 1117 fault reaction is desired. 1119 6.2. RTCP ECN Feedback SDP Parameter 1121 A new "nack" feedback parameter "ecn" is defined to indicate the 1122 usage of the RTCP ECN feedback packet format (Section 5.1). The ABNF 1123 [RFC5234] definition of the SDP parameter extension is: 1124 rtcp-fb-nack-param = 1125 rtcp-fb-nack-param /= ecn-fb-par 1126 ecn-fb-par = SP "ecn" 1128 The offer/answer rules for this SDP feedback parameters are specified 1129 in the RTP/AVPF profile [RFC4585]. 1131 6.3. XR Block ECN SDP Parameter 1133 A new unilateral RTCP XR block for ECN summary information is 1134 specified, thus the XR block SDP signalling also needs to be extended 1135 with a parameter. This is done in the same way as for the other XR 1136 blocks. The XR block SDP attribute as defined in Section 5.1 of the 1137 RTCP XR specification [RFC3611] is defined to be extensible. As no 1138 parameter values are needed for this ECN summary block, this 1139 parameter extension consists of a simple parameter name used to 1140 indicate support and intent to use the XR block. 1141 xr-format = 1142 xr-format /= ecn-summary-par 1143 ecn-summary-par = "ecn-sum" 1145 For SDP declarative and offer/answer usage, see the RTCP XR 1146 specification [RFC3611] and its description of how to handle 1147 unilateral parameters. 1149 6.4. ICE Parameter to Signal ECN Capability 1151 One new ICE [RFC5245] option, "rtp+ecn", is defined. This is used 1152 with the SDP session level "a=ice-options" attribute in an SDP offer 1153 to indicate that the initiator of the ICE exchange has the capability 1154 to support ECN for RTP-over-UDP flows (via "a=ice-options: rtp+ecn"). 1155 The answering party includes this same attribute at the session level 1156 in the SDP answer if it also has the capability, and removes the 1157 attribute if it does not wish to use ECN, or doesn't have the 1158 capability to use ECN. If the ICE initiation method (Section 7.2.2) 1159 is actually going to be used, it is also needs to be explicitly 1160 negotiated using the "a=ecn-capable-rtp" attribute. This ICE option 1161 SHALL be included when the ICE initiation method is offered or 1162 declared in the SDP. 1164 Note: This signalling mechanism is not strictly needed as long as 1165 the STUN ECN testing capability is used within the context of this 1166 document. It may however be useful if the ECN verification 1167 capability is used in additional contexts. 1169 7. Use of ECN with RTP/UDP/IP 1171 In the detailed specification of the behaviour below, the different 1172 functions in the general case will first be discussed. In case 1173 special considerations are needed for middleboxes, multicast usage 1174 etc, those will be specially discussed in related subsections. 1176 7.1. Negotiation of ECN Capability 1178 The first stage of ECN negotiation for RTP-over-UDP is to signal the 1179 capability to use ECN. An RTP system that supports ECN and uses SDP 1180 for its signalling MUST implement the SDP extension to signal ECN 1181 capability as described in Section 6.1, the RTCP ECN feedback SDP 1182 parameter defined in Section 6.2, and the XR Block ECN SDP parameter 1183 defined in Section 6.3. It MAY also implement alternative ECN 1184 capability negotiation schemes, such as the ICE extension described 1185 in Section 6.4. Other signalling systems will need to define 1186 signalling parameters corresponding to those defined for SDP. 1188 The "ecn-capable-rtp" SDP attribute MUST be used when employing ECN 1189 for RTP according to this specification in systems using SDP. As the 1190 RTCP XR ECN summary report is required independently of the 1191 initialization method or congestion control scheme, the "rtcp-xr" 1192 attribute with the "ecn-sum" parameter MUST also be used. The 1193 "rtcp-fb" attribute with the "nack" parameter "ecn" MUST be used 1194 whenever the initialization method or a congestion control algorithm 1195 requires timely sender side knowledge of received CE markings. If 1196 the congestion control scheme requires additional signalling, this 1197 should be indicated as appropriate. 1199 7.2. Initiation of ECN Use in an RTP Session 1201 Once the sender and the receiver(s) have agreed that they have the 1202 capability to use ECN within a session, they may attempt to initiate 1203 ECN use. All session participants connected over the same transport 1204 MUST use the same initiation method. RTP mixers or translators can 1205 use different initiation methods to different participants that are 1206 connected over different underlying transports. The mixer or 1207 translator will need to do individual signalling with each 1208 participant to ensure it is consistent with the ECN support in those 1209 cases where it does not function as one end-point for the ECN control 1210 loop. 1212 At the start of the RTP session, when the first packets with ECT are 1213 sent, it is important to verify that IP packets with ECN field values 1214 of ECT or ECN-CE will reach their destination(s). There is some risk 1215 that the use of ECN will result in either reset of the ECN field, or 1216 loss of all packets with ECT or ECN-CE markings. If the path between 1217 the sender and the receivers exhibits either of these behaviours one 1218 needs to stop using ECN immediately to protect both the network and 1219 the application. 1221 The RTP senders and receivers SHALL NOT ECT mark their RTCP traffic 1222 at any time. This is to ensure that packet loss due to ECN marking 1223 will not effect the RTCP traffic and the necessary feedback 1224 information it carries. 1226 An RTP system that supports ECN MUST implement the initiation of ECN 1227 using in-band RTP and RTCP described in Section 7.2.1. It MAY also 1228 implement other mechanisms to initiate ECN support, for example the 1229 STUN-based mechanism described in Section 7.2.2, or use the leap of 1230 faith option if the session supports the limitations provided in 1231 Section 7.2.3. If support for both in-band and out-of-band 1232 mechanisms are signalled, the sender when negotiating SHOULD offer 1233 detection of ECT using STUN with ICE with higher priority than 1234 detection of ECT using RTP and RTCP. 1236 No matter how ECN usage is initiated, the sender MUST continually 1237 monitor the ability of the network, and all its receivers, to support 1238 ECN, following the mechanisms described in Section 7.4. This is 1239 necessary because path changes or changes in the receiver population 1240 may invalidate the ability of the system to use ECN. 1242 7.2.1. Detection of ECT using RTP and RTCP 1244 The ECN initiation phase using RTP and RTCP to detect if the network 1245 path supports ECN comprises three stages. Firstly, the RTP sender 1246 generates some small fraction of its traffic with ECT marks to act as 1247 probe for ECN support. Then, on receipt of these ECT-marked packets, 1248 the receivers send RTCP ECN feedback packets and RTCP ECN summary 1249 reports to inform the sender that their path supports ECN. Finally, 1250 the RTP sender makes the decision to use ECN or not, based on whether 1251 the paths to all RTP receivers have been verified to support ECN. 1253 Generating ECN Probe Packets: During the ECN initiation phase, an 1254 RTP sender SHALL mark a small fraction of its RTP traffic as ECT, 1255 while leaving the reminder of the packets unmarked. The main 1256 reason for only marking some packets is to maintain usable media 1257 delivery during the ECN initiation phase in those cases where ECN 1258 is not supported by the network path. A secondary reason to send 1259 some not-ECT packets are to ensure that the receivers will send 1260 RTCP reports on this sender, even if all ECT marked packets are 1261 lost in transit. The not-ECT packets also provide a base-line to 1262 compare performance parameters against. A fourth reason for only 1263 probing with a small number of packets is to reduce the risk that 1264 significant numbers of congestion markings might be lost if ECT is 1265 cleared to Not-ECT by an ECN-Reverting Middlebox. Then any 1266 resulting lack of congestion response is likely to have little 1267 damaging effect on others. An RTP sender is RECOMMENDED to send a 1268 minimum of two packets with ECT markings per RTCP reporting 1269 interval. In case a random ECT pattern is intended to be used, at 1270 least one packet with ECT(0) and one with ECT(1) should be sent 1271 per reporting interval; in case a single ECT marking is to be 1272 used, only that ECT value SHOULD be sent. The RTP sender SHALL 1273 continue to send some ECT marked traffic as long as the ECN 1274 initiation phase continues. The sender SHOULD NOT mark all RTP 1275 packets as ECT during the ECN initiation phase. 1277 This memo does not mandate which RTP packets are marked with ECT 1278 during the ECN initiation phase. An implementation should insert 1279 ECT marks in RTP packets in a way that minimises the impact on 1280 media quality if those packets are lost. The choice of packets to 1281 mark is clearly very media dependent, but the use of RTP NO-OP 1282 payloads [I-D.ietf-avt-rtp-no-op], if supported, would be an 1283 appropriate choice. For audio formats, if would make sense for 1284 the sender to mark comfort noise packets or similar. For video 1285 formats, packets containing P- or B-frames (rather than I-frames) 1286 would be an appropriate choice. No matter which RTP packets are 1287 marked, those packets MUST NOT be sent in duplicate, with and 1288 without ECT, since the RTP sequence number is used to identify 1289 packets that are received with ECN markings. 1291 Generating RTCP ECN Feedback: If ECN capability has been negotiated 1292 in an RTP session, the receivers in the session MUST listen for 1293 ECT or ECN-CE marked RTP packets, and generate RTCP ECN feedback 1294 packets (Section 5.1) to mark their receipt. An immediate or 1295 early (depending on the RTP/AVPF mode) ECN feedback packet SHOULD 1296 be generated on receipt of the first ECT or ECN-CE marked packet 1297 from a sender that has not previously sent any ECT traffic. Each 1298 regular RTCP report MUST also contain an ECN summary report 1299 (Section 5.2). Reception of subsequent ECN-CE marked packets MUST 1300 result in additional early or immediate ECN feedback packets being 1301 sent unless no timely feedback is required. 1303 Determination of ECN Support: RTP is a group communication protocol, 1304 where members can join and leave the group at any time. This 1305 complicates the ECN initiation phase, since the sender must wait 1306 until it believes the group membership has stabilised before it 1307 can determine if the paths to all receivers support ECN (group 1308 membership changes after the ECN initiation phase has completed 1309 are discussed in Section 7.3). 1311 An RTP sender shall consider the group membership to be stable 1312 after it has been in the session and sending ECT-marked probe 1313 packets for at least three RTCP reporting intervals (i.e., after 1314 sending its third regularly scheduled RTCP packet), and when a 1315 complete RTCP reporting interval has passed without changes to the 1316 group membership. ECN initiation is considered successful when 1317 the group membership is stable, and all known participants have 1318 sent one or more RTCP ECN feedback packets or RTCP XR ECN summary 1319 reports indicating correct receipt of the ECT-marked RTP packets 1320 generated by the sender. 1322 As an optimisation, if an RTP sender is initiating ECN usage 1323 towards a unicast address, then it MAY treat the ECN initiation as 1324 provisionally successful if it receives an RTCP ECN feedback 1325 report or an RTCP XR ECN summary report indicating successful 1326 receipt of the ECT-marked packets, with no negative indications, 1327 from a single RTP receiver (where a single RTP receiver is 1328 considered as all SSRCs used by a single RTCP CNAME). After 1329 declaring provisional success, the sender MAY generate ECT-marked 1330 packets as described in Section 7.3, provided it continues to 1331 monitor the RTCP reports for a period of three RTCP reporting 1332 intervals from the time the ECN initiation started, to check if 1333 there are any other participants in the session. Thus as long as 1334 any additional SSRC that report on the ECN usage are using the 1335 same RTCP CNAME as the previous reports and they are all 1336 indicating functional ECN the sender may continue. If other 1337 participants are detected, i.e., other RTCP CNAMEs, the sender 1338 MUST fallback to only ECT-marking a small fraction of its RTP 1339 packets, while it determines if ECN can be supported following the 1340 full procedure described above. Different RTCP CNAMEs received 1341 over an unicast transport may occur when using translators in a 1342 multi-party RTP session (e.g., when using a centralised conference 1343 bridge). 1345 Note: The above optimization supports peer to peer unicast 1346 transport with several SSRCs multiplexed onto the same flow 1347 (e.g., a single participant with two video cameras, or SSRC 1348 multiplexed RTP retransmission [RFC4588]). It is desirable to 1349 be able to rapidly negotiate ECN support for such a session, 1350 but the optimisation above can fail if there are 1351 implementations that use the same CNAME for different parts of 1352 a distributed implementation that have different transport 1353 characteristics (e.g., if a single logical endpoint is split 1354 across multiple hosts). 1356 ECN initiation is considered to have failed at the instant the 1357 initiating RTP sender received an RTCP packet that doesn't contain 1358 an RTCP ECN feedback report or ECN summary report from any RTP 1359 session participant that has an RTCP RR with an extended RTP 1360 sequence number field that indicates that it should have received 1361 multiple (>3) ECT marked RTP packets. This can be due to failure 1362 to support the ECN feedback format by the receiver or some 1363 middlebox, or the loss of all ECT marked packets. Both indicate a 1364 lack of ECN support. 1366 If the ECN negotiation succeeds, this indicates that the path can 1367 pass some ECN-marked traffic, and that the receivers support ECN 1368 feedback. This does not necessarily imply that the path can robustly 1369 convey ECN feedback; Section 7.3 describes the ongoing monitoring 1370 that must be performed to ensure the path continues to robustly 1371 support ECN. 1373 When a sender or receiver detects ECN failures on paths they should 1374 log these to enable follow up and statistics gathering regarding 1375 broken paths. The logging mechanism used is implementation 1376 dependent. 1378 7.2.2. Detection of ECT using STUN with ICE 1380 This section describes an OPTIONAL method that can be used to avoid 1381 media impact and also ensure an ECN capable path prior to media 1382 transmission. This method is considered in the context where the 1383 session participants are using ICE [RFC5245] to find working 1384 connectivity. We need to use ICE rather than STUN only, as the 1385 verification needs to happen from the media sender to the address and 1386 port on which the receiver is listening. 1388 Note that this method is only applicable to sessions when the remote 1389 destinations are unicast addresses. In addition, transport 1390 translators that do not terminate the ECN control loop and may 1391 distribute received packets to more than one other receiver must 1392 either disallow this method (and use the RTP/RTCP method instead), or 1393 implement additional handling as discussed below. This is because 1394 the ICE initialization method verifies the underlying transport to 1395 one particular address and port. If the receiver at that address and 1396 port intends to use the received packets in a multi-point session 1397 then the tested capabilities and the actual session behavior are not 1398 matched. 1400 To minimise the impact of set-up delay, and to prioritise the fact 1401 that one has a working connectivity rather than necessarily finding 1402 the best ECN capable network path, this procedure is applied after 1403 having performed a successful connectivity check for a candidate, 1404 which is nominated for usage. At that point an additional 1405 connectivity check is performed, sending the "ECN Check" attribute in 1406 a STUN packet that is ECT marked. On reception of the packet, a STUN 1407 server supporting this extension will note the received ECN field 1408 value, and send a STUN/UDP/IP packet in reply with the ECN field set 1409 to not-ECT and including an ECN check attribute. A STUN server that 1410 doesn't understand the extension, or is incapable of reading the ECN 1411 values on incoming STUN packets, should follow the rule in the STUN 1412 specification for unknown comprehension-optional attributes, and 1413 ignore the attribute, resulting in the sender receiving a STUN 1414 response without the ECN Check STUN attribute. 1416 The ECN STUN checks can be lost on the path, for example due to the 1417 ECT marking, but also various other non ECN related reasons causing 1418 packet loss. The goal is to detect when the ECT markings are 1419 rewritten or if it is the ECT marking that causes packet loss so that 1420 the path can be determined as not ECT. Other reasons for packet loss 1421 should not result in a failure to verify the path as ECT. Therefore 1422 a number of retransmissions should be attempted. But, the sender of 1423 ECN STUN checks will also have to set a criteria for when it gives up 1424 testing for ECN capability on the path. As the ICE agent have 1425 successfully verified the path a RTT measurement for this path can be 1426 performed. To have high probability of succesfully verifying the 1427 path it is RECOMMENDED that the client retransmitt the ECN STUN check 1428 4 times. The interval between the retransmissions will be based on 1429 the Ta timer as defined in Section 16.1 for RTP Media Streams in ICE 1430 [RFC5245]. The number of ECN STUN checks needing to be sent will 1431 depend on the number of ECN capable flows (N) that is to be 1432 established. The interval between each transmission of an ECN check 1433 packet MUST be Ta. In other words for a given flow being verified 1434 for ECT the RTO is set to Ta*N. The transmission for that flow is 1435 stopped when an ECN Check STUN response has been received, which 1436 don't indicate to retransmit the request due to temporary error, the 1437 maximum number of retransmissions has been sent. The ICE agent is 1438 recommended to give up on the ECN verification MAX(1.5*RTT, 20 ms) 1439 after the last ECN STUN check was sent. 1441 The STUN ECN check attribute contains one field and a flag, as shown 1442 in Figure 6. The flag indicates whether the echo field contains a 1443 valid value or not. The field is the ECN echo field, and when valid 1444 contains the two ECN bits from the packet it echoes back. The ECN 1445 check attribute is a comprehension optional attribute. 1446 0 1 2 3 1447 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 1448 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1449 | Type | Length | 1450 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1451 | Reserved |ECF|V| 1452 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1454 Figure 6: ECN Check STUN Attribute 1456 V: Valid (1 bit) ECN Echo value field is valid when set to 1, and 1457 invalid when set 0. 1459 ECF: ECN Echo value field (2 bits) contains the ECN field value of 1460 the STUN packet it echoes back when field is valid. If invalid 1461 the content is arbitrary. 1463 Reserved: Reserved bits (29 bits) SHALL be set to 0 on transmission, 1464 and SHALL be ignored on reception. 1466 This attribute MAY be included in any STUN request to request the ECN 1467 field to be echoed back. In STUN requests the V bit SHALL be set to 1468 0. A compliant STUN server receiving a request with the ECN Check 1469 attribute SHALL read the ECN field value of the IP/UDP packet the 1470 request was received in. Upon forming the response the server SHALL 1471 include the ECN Check attribute setting the V bit to valid and 1472 include the read value of the ECN field into the ECF field. If the 1473 STUN responder was unable to ascertain, due to temporary errors, the 1474 ECN value of the STUN request, it SHALL set the V bit in the response 1475 to 0. The STUN client may retry immediately. 1477 The ICE based initialization method does require some special 1478 consideration when used by a translator. This is especially for 1479 transport translators and translators that fragment or reassemble 1480 packets, since they do not separate the ECN control loops between the 1481 end-points and the translator. When using ICE-based initiation, such 1482 a translator must ensure that any participants joining an RTP session 1483 for which ECN has been negotiated are successfully verified in the 1484 direction from the translator to the joining participant. 1485 Alternatively, it must correctly handle remarking of ECT RTP packets 1486 towards that participant. When a new participant joins the session, 1487 the translator will perform a check towards the new participant. If 1488 that is successfully completed the ECT properties of the session are 1489 maintained for the other senders in the session. If the check fails 1490 then the existing senders will now see a participant that fails to 1491 receive ECT. Thus the failure detection in those senders will 1492 eventually detect this. However to avoid misusing the network on the 1493 path from the translator to the new participant, the translator SHALL 1494 remark the traffic intended to be forwarded from ECT to non-ECT. Any 1495 packet intended to be forward that are ECN-CE marked SHALL be discard 1496 and not sent. In cases where the path from a new participant to the 1497 translator fails the ECT check then only that sender will not 1498 contribute any ECT marked traffic towards the translator. 1500 7.2.3. Leap of Faith ECT initiation method 1502 This method for initiating ECN usage is a leap of faith that assumes 1503 that ECN will work on the used path(s). The method is to go directly 1504 to "ongoing use of ECN" as defined in Section 7.3. Thus all RTP 1505 packets MAY be marked as ECT and the failure detection MUST be used 1506 to detect any case when the assumption that the path was ECT capable 1507 is wrong. This method is only recommended for controlled 1508 environments where the whole path(s) between sender and receiver(s) 1509 has been built and verified to be ECT. 1511 If the sender marks all packets as ECT while transmitting on a path 1512 that contains an ECN-blocking middlebox, then receivers downstream of 1513 that middlebox will not receive any RTP data packets from the sender, 1514 and hence will not consider it to be an active RTP SSRC. The sender 1515 can detect this and revert to sending packets without ECT marks, 1516 since RTCP SR/RR packets from such receivers will either not include 1517 a report for sender's SSRC, or will report that no packets have been 1518 received, but this takes at least one RTCP reporting interval. It 1519 should be noted that a receiver might generate its first RTCP packet 1520 immediately on joining a unicast session, or very shortly after 1521 joining a RTP/AVPF session, before it has had chance to receive any 1522 data packets. A sender that receives RTCP SR/RR packet indicating 1523 lack of reception by a receiver SHOULD therefore wait for a second 1524 RTCP report from that receiver to be sure that the lack of reception 1525 is due to ECT-marking. Since this recovery process can take several 1526 tens of seconds, during which time the RTP session is unusable for 1527 media, it is NOT RECOMMENDED that the leap-of-faith ECT initiation 1528 method be used in environments where ECN-blocking middleboxes are 1529 likely to be present. 1531 7.3. Ongoing Use of ECN Within an RTP Session 1533 Once ECN has been successfully initiated for an RTP sender, that 1534 sender begins sending all RTP data packets as ECT-marked, and its 1535 receivers send ECN feedback information via RTCP packets. This 1536 section describes procedures for sending ECT-marked data, providing 1537 ECN feedback information via RTCP, and responding to ECN feedback 1538 information. 1540 7.3.1. Transmission of ECT-marked RTP Packets 1542 After a sender has successfully initiated ECN use, it SHOULD mark all 1543 the RTP data packets it sends as ECT. The sender SHOULD mark packets 1544 as ECT(0) unless the receiver expresses a preference for ECT(1) or 1545 random using the "ect" parameter in the "a=ecn-capable-rtp" 1546 attribute. 1548 The sender SHALL NOT include ECT marks on outgoing RTCP packets, and 1549 SHOULD NOT include ECT marks on any other outgoing control messages 1550 (e.g., STUN [RFC5389] packets, DTLS [RFC4347] handshake packets, or 1551 ZRTP [RFC6189] control packets) that are multiplexed on the same UDP 1552 port. For control packets there might be exceptions, like the STUN 1553 based ECN check defined in Section 7.2.2. 1555 7.3.2. Reporting ECN Feedback via RTCP 1557 An RTP receiver that receives a packet with an ECN-CE mark, or that 1558 detects a packet loss, MUST schedule the transmission of an RTCP ECN 1559 feedback packet as soon as possible (subject to the constraints of 1560 [RFC4585] and [RFC3550]) to report this back to the sender unless no 1561 timely feedback is required. The feedback RTCP packet SHALL consist 1562 of at least one ECN feedback packet (Section 5.1) reporting on the 1563 packets received since the last ECN feedback packet, and will contain 1564 (at least) an RTCP SR/RR packet and an SDES packet, unless reduced 1565 size RTCP [RFC5506] is used. The RTP/AVPF profile in early or 1566 immediate feedback mode SHOULD be used where possible, to reduce the 1567 interval before feedback can be sent. To reduce the size of the 1568 feedback message, reduced size RTCP [RFC5506] MAY be used if 1569 supported by the end-points. Both RTP/AVPF and reduced size RTCP 1570 MUST be negotiated in the session set-up signalling before they can 1571 be used. 1573 Every time a regular compound RTCP packet is to be transmitted, an 1574 ECN-capable RTP receiver MUST include an RTCP XR ECN summary report 1575 as described in Section 5.2 as part of the compound packet. 1577 The multicast feedback implosion problem, that occurs when many 1578 receivers simultaneously send feedback to a single sender, must be 1579 considered. The RTP/AVPF transmission rules will limit the amount of 1580 feedback that can be sent, avoiding the implosion problem but also 1581 delaying feedback by varying degrees from nothing up to a full RTCP 1582 reporting interval. As a result, the full extent of a congestion 1583 situation may take some time to reach the sender, although some 1584 feedback should arrive in a reasonably timely manner, allowing the 1585 sender to react on a single or a few reports. 1587 7.3.3. Response to Congestion Notifications 1589 The reception of RTP packets with ECN-CE marks in the IP header is a 1590 notification that congestion is being experienced. The default 1591 reaction on the reception of these ECN-CE marked packets MUST be to 1592 provide the congestion control algorithm with a congestion 1593 notification that triggers the algorithm to react as if packet loss 1594 had occurred. There should be no difference in congestion response 1595 if ECN-CE marks or packet drops are detected. 1597 We note that there MAY be other reactions to ECN-CE specified in the 1598 future. Such an alternative reaction MUST be specified and 1599 considered to be safe for deployment under any restrictions 1600 specified. A potential example for an alternative reaction could be 1601 emergency communications (such as that generated by first responders, 1602 as opposed to the general public) in networks where the user has been 1603 authorized. A more detailed description of these other reactions, as 1604 well as the types of congestion control algorithms used by end-nodes, 1605 is outside of the scope of this document. 1607 Depending on the media format, type of session, and RTP topology 1608 used, there are several different types of congestion control that 1609 can be used: 1611 Sender-Driven Congestion Control: The sender is responsible for 1612 adapting the transmitted bit-rate in response to RTCP ECN 1613 feedback. When the sender receives the ECN feedback data it feeds 1614 this information into its congestion control or bit-rate 1615 adaptation mechanism so that it can react as if packet loss was 1616 reported. The congestion control algorithm to be used is not 1617 specified here, although TFRC [RFC5348] is one example that might 1618 be used. 1620 Receiver-Driven Congestion Control: In a receiver driven congestion 1621 control mechanism, the receivers can react to the ECN-CE marks 1622 themselves without providing ECN-CE feedback to the sender. This 1623 may allow faster response than sender-driven congestion control in 1624 some circumstances and also scale to large number of receivers and 1625 multicast usage. One example of receiver-driven congestion 1626 control is implemented by providing the content in a layered way, 1627 with each layer providing improved media quality but also 1628 increased bandwidth usage. The receiver locally monitors the 1629 ECN-CE marks on received packets to check if it experiences 1630 congestion with the current number of layers. If congestion is 1631 experienced, the receiver drops one layer, so reducing the 1632 resource consumption on the path towards itself. For example, if 1633 a layered media encoding scheme such as H.264 SVC is used, the 1634 receiver may change its layer subscription, and so reduce the bit 1635 rate it receives. The receiver MUST still send RTCP XR ECN 1636 Summary to the sender, even if it can adapt without contact with 1637 the sender, so that the sender can determine if ECN is supported 1638 on the network path. The timeliness of RTCP feedback is less of a 1639 concern with receiver driven congestion control, and regular RTCP 1640 reporting of ECN summary information is sufficient (without using 1641 RTP/AVPF immediate or early feedback). 1643 Hybrid: There might be mechanisms that utilize both some receiver 1644 behaviors and some sender side monitoring, thus requiring both 1645 feedback of congestion events to the sender and taking receiver 1646 decisions and possible signalling to the sender. In this case the 1647 congestion control algorithm needs to use the signalling to 1648 indicate which features of ECN for RTP are required. 1650 Responding to congestion indication in the case of multicast traffic 1651 is a more complex problem than for unicast traffic. The fundamental 1652 problem is diverse paths, i.e., when different receivers don't see 1653 the same path, and thus have different bottlenecks, so the receivers 1654 may get ECN-CE marked packets due to congestion at different points 1655 in the network. This is problematic for sender driven congestion 1656 control, since when receivers are heterogeneous in regards to 1657 capacity, the sender is limited to transmitting at the rate the 1658 slowest receiver can support. This often becomes a significant 1659 limitation as group size grows. Also, as group size increases the 1660 frequency of reports from each receiver decreases, which further 1661 reduces the responsiveness of the mechanism. Receiver-driven 1662 congestion control has the advantage that each receiver can choose 1663 the appropriate rate for its network path, rather than all receivers 1664 having to settle for the lowest common rate. 1666 We note that ECN support is not a silver bullet to improving 1667 performance. The use of ECN gives the chance to respond to 1668 congestion before packets are dropped in the network, improving the 1669 user experience by allowing the RTP application to control how the 1670 quality is reduced. An application which ignores ECN Congestion 1671 Experienced feedback is not immune to congestion: the network will 1672 eventually begin to discard packets if traffic doesn't respond. It 1673 is in the best interest of an application to respond to ECN 1674 congestion feedback promptly, to avoid packet loss. 1676 7.4. Detecting Failures 1678 Senders and receivers can deliberately ignore ECN-CE and thus get a 1679 benefit over behaving flows (cheating). Th ECN Nonce [RFC3540] is an 1680 addition to TCP that attempts to solve this issue as long as the 1681 sender acts on behalf of the network. The assumption about the 1682 senders acting on the behalf of the network may be reduced due to the 1683 nature of peer-to-peer use of RTP. Still a significant portion of 1684 RTP senders are infrastructure devices (for example, streaming media 1685 servers) that do have an interest in protecting both service quality 1686 and the network. Even though there may be cases where nonce can be 1687 applicable also for RTP, it is not included in this specification. 1688 This as a receiver interested in cheating would simple claim to not 1689 support nonce, or even ECN itself. It is, however, worth mentioning 1690 that, as real-time media is commonly sensitive to increased delay and 1691 packet loss, it will be in both the media sender and receivers 1692 interest to minimise the number and duration of any congestion events 1693 as they will adversely affect media quality. 1695 RTP sessions can also suffer from path changes resulting in a non-ECN 1696 compliant node becoming part of the path. That node may perform 1697 either of two actions that has effect on the ECN and application 1698 functionality. The gravest is if the node drops packets with the ECN 1699 field set to ECT(0), ECT(1), or ECN-CE. This can be detected by the 1700 receiver when it receives an RTCP SR packet indicating that a sender 1701 has sent a number of packets that it has not received. The sender 1702 may also detect such a middlebox based on the receiver's RTCP RR 1703 packet, when the extended sequence number is not advanced due to the 1704 failure to receive packets. If the packet loss is less than 100%, 1705 then packet loss reporting in either the ECN feedback information or 1706 RTCP RR will indicate the situation. The other action is to re-mark 1707 a packet from ECT or ECN-CE to not-ECT. That has less dire results, 1708 however it should be detected so that ECN usage can be suspended to 1709 prevent misusing the network. 1711 The RTCP XR ECN summary packet and the ECN feedback packet allow the 1712 sender to compare the number of ECT marked packets of different types 1713 received with the number it actually sent. The number of ECT packets 1714 received, plus the number of ECN-CE marked and lost packets, should 1715 correspond to the number of sent ECT marked packets plus the number 1716 of received duplicates. If these numbers doesn't agree there are two 1717 likely reasons, a translator changing the stream or not carrying the 1718 ECN markings forward, or that some node re-marks the packets. In 1719 both cases the usage of ECN is broken on the path. By tracking all 1720 the different possible ECN field values a sender can quickly detect 1721 if some non-compliant behavior is happening on the path. 1723 Thus packet losses and non-matching ECN field value statistics are 1724 possible indication of issues with using ECN over the path. The next 1725 section defines both sender and receiver reactions to these cases. 1727 7.4.1. Fallback mechanisms 1729 Upon the detection of a potential failure, both the sender and the 1730 receiver can react to mitigate the situation. 1732 A receiver that detects a packet loss burst MAY schedule an early 1733 feedback packet that includes at least the RTCP RR and the ECN 1734 feedback message to report this to the sender. This will speed up 1735 the detection of the loss at the sender, thus triggering sender side 1736 mitigation. 1738 A sender that detects high packet loss rates for ECT-marked packets 1739 SHOULD immediately switch to sending packets as not-ECT to determine 1740 if the losses are potentially due to the ECT markings. If the losses 1741 disappear when the ECT-marking is discontinued, the RTP sender should 1742 go back to initiation procedures to attempt to verify the apparent 1743 loss of ECN capability of the used path. If a re-initiation fails 1744 then the two possible actions exist: 1746 1. Periodically retry the ECN initiation to detect if a path change 1747 occurs to a path that is ECN capable. 1749 2. Renegotiate the session to disable ECN support. This is a choice 1750 that is suitable if the impact of ECT probing on the media 1751 quality is noticeable. If multiple initiations have been 1752 successful, but the following full usage of ECN has resulted in 1753 the fallback procedures, then disabling of the ECN support is 1754 RECOMMENDED. 1756 We foresee the possibility of flapping ECN capability due to several 1757 reasons: video switching MCU or similar middleboxes that selects to 1758 deliver media from the sender only intermittently; load balancing 1759 devices may in worst case result in that some packets take a 1760 different network path than the others; mobility solutions that 1761 switch underlying network path in a transparent way for the sender or 1762 receiver; and membership changes in a multicast group. It is however 1763 appropriate to mention that there are also issues such as re-routing 1764 of traffic due to a flappy route table or excessive reordering and 1765 other issues that are not directly ECN related but nevertheless may 1766 cause problems for ECN. 1768 7.4.2. Interpretation of ECN Summary information 1770 This section contains discussion on how the ECN summary report 1771 information can be used to detect various types of ECN path issues. 1773 We first review the information the RTCP reports provide on a per 1774 source (SSRC) basis: 1776 ECN-CE Counter: The number of RTP packets received so far in the 1777 session with an ECN field set to CE. 1779 ECT (0/1) Counters: The number of RTP packets received so far in the 1780 session with an ECN field set to ECT (0) and ECT (1) respectively. 1782 not-ECT Counter: The number of RTP packets received so far in the 1783 session with an ECN field set to not-ECT. 1785 Lost Packets counter: The number of RTP packets that where expected 1786 based on sequence numbers but never received. 1788 Duplication Counter: The number of received RTP packets that are 1789 duplicates of already received ones. 1791 Extended Highest Sequence number: The highest sequence number seen 1792 when sending this report, but with additional bits, to handle 1793 disambiguation when wrapping the RTP sequence number field. 1795 The counters will be initialised to zero to provide value for the RTP 1796 stream sender from the first report. After the first report, the 1797 changes between the last received report and the previous report are 1798 determined by simply taking the values of the latest minus the 1799 previous, taking wrapping into account. This definition is also 1800 robust to packet losses, since if one report is missing, the 1801 reporting interval becomes longer, but is otherwise equally valid. 1803 In a perfect world, the number of not-ECT packets received should be 1804 equal to the number sent minus the lost packets counter, and the sum 1805 of the ECT(0), ECT(1), and ECN-CE counters should be equal to the 1806 number of ECT marked packet sent. Two issues may cause a mismatch in 1807 these statistics: severe network congestion or unresponsive 1808 congestion control might cause some ECT-marked packets to be lost, 1809 and packet duplication might result in some packets being received, 1810 and counted in the statistics, multiple times (potentially with a 1811 different ECN-mark on each copy of the duplicate). 1813 The rate of packet duplication is tracked, allowing one to take the 1814 duplication into account. The value of the ECN field for duplicates 1815 will also be counted and when comparing the figures one needs to take 1816 some fraction of packet duplicates that are non-ECT and some fraction 1817 of packet duplicates being ECT into account into the calculation. 1818 Thus when only sending non-ECT then the number of sent packets plus 1819 reported duplicates equals the number of received non-ECT. When 1820 sending only ECT then number of sent ECT packets plus duplicates will 1821 equal ECT(0), ECT(1), ECN-CE and packet loss. When sending a mix of 1822 non-ECT and ECT then there is an uncertainty if any duplicate or 1823 packet loss was an non-ECT or ECT. If the packet duplication is 1824 completely independent of the usage of ECN, then the fraction of 1825 packet duplicates should be in relation to the number of non-ECT vs 1826 ECT packet sent during the period of comparison. This relation does 1827 not hold for packet loss, where higher rates of packet loss for non- 1828 ECT is expected than for ECT traffic. 1830 Detecting clearing of ECN field: If the ratio between ECT and not-ECT 1831 transmitted in the reports has become all not-ECT, or has 1832 substantially changed towards not-ECT, then this is clearly an 1833 indication that the path results in clearing of the ECT field. 1835 Dropping of ECT packets: To determine if the packet drop ratio is 1836 different between not-ECT and ECT marked transmission requires a mix 1837 of transmitted traffic. The sender should compare if the delivery 1838 percentage (delivered / transmitted) between ECT and not-ECT is 1839 significantly different. Care must be taken if the number of packets 1840 are low in either of the categories. One must also take into account 1841 the level of CE marking. A CE marked packet would have been dropped 1842 unless it was ECT marked. Thus, the packet loss level for not-ECT 1843 should be approximately equal to the loss rate for ECT when counting 1844 the CE marked packets as lost ones. A sender performing this 1845 calculation needs to ensure that the difference is statistically 1846 significant. 1848 If erroneous behavior is detected, it should be logged to enable 1849 follow up and statistics gathering. 1851 8. Processing ECN in RTP Translators and Mixers 1853 RTP translators and mixers that support ECN for RTP are required to 1854 process, and potentially modify or generate ECN marking in RTP 1855 packets. They also need to process, and potentially modify or 1856 generate RTCP ECN feedback packets for the translated and/or mixed 1857 streams. This includes both downstream RTCP reports generated by the 1858 media sender, and also reports generated by the receivers, flowing 1859 upstream back towards the sender. 1861 8.1. Transport Translators 1863 Some translators only perform transport level translations, like 1864 copying packets from one address domain, like unicast to multicast. 1865 It may also perform relaying like copying an incoming packet to a 1866 number of unicast receivers. This section details the ECN related 1867 actions for RTP and RTCP. 1869 For the RTP data packets the translator, which does not modify the 1870 media stream, SHOULD copy the ECN bits unchanged from the incoming to 1871 the outgoing datagrams, unless the translator itself is overloaded 1872 and experiencing congestion, in which case it may mark the outgoing 1873 datagrams with an ECN-CE mark. 1875 A Transport translator does not modify RTCP packets. It however MUST 1876 perform the corresponding transport translation of the RTCP packets 1877 as it does with RTP packets being sent from the same source/ 1878 end-point. 1880 8.2. Fragmentation and Reassembly in Translators 1882 An RTP translator may fragment or reassemble RTP data packets without 1883 changing the media encoding, and without reference to the congestion 1884 state of the networks it bridges. An example of this might be to 1885 combine packets of a voice-over-IP stream coded with one 20ms frame 1886 per RTP packet into new RTP packets with two 20ms frames per packet, 1887 thereby reducing the header overheads and so stream bandwidth, at the 1888 expense of an increase in latency. If multiple data packets are re- 1889 encoded into one, or vice versa, the RTP translator MUST assign new 1890 sequence numbers to the outgoing packets. Losses in the incoming RTP 1891 packet stream may also induce corresponding gaps in the outgoing RTP 1892 sequence numbers. An RTP translator MUST rewrite RTCP packets to 1893 make the corresponding changes to their sequence numbers, and to 1894 reflect the impact of the fragmentation or reassembly. This section 1895 describes how that rewriting is to be done for RTCP ECN feedback 1896 packets. Section 7.2 of [RFC3550] describes general procedures for 1897 other RTCP packet types. 1899 The processing of arriving RTP packets for this case is as follows. 1900 If an ECN marked packet is split into two, then both the outgoing 1901 packets MUST be ECN marked identically to the original; if several 1902 ECN marked packets are combined into one, the outgoing packet MUST be 1903 either ECN-CE marked or dropped if any of the incoming packets are 1904 ECN-CE marked. If the outgoing combined packet is not ECN-CE marked, 1905 then it MUST be ECT marked if any of the incoming packets were ECT 1906 marked. 1908 RTCP ECN feedback packets (Section 5.1) contain seven fields that are 1909 rewritten in an RTP translator that fragments or reassembles packets: 1910 the extended highest sequence number, the duplication counter, the 1911 lost packets counter, the ECN-CE counter, and not-ECT counter, the 1912 ECT(0) counter, and the ECT(1) counter. The RTCP XR report block for 1913 ECN summary information (Section 5.2) includes all of these fields 1914 except the extended highest sequence number which is present in the 1915 report block in an SR or RR packet. The procedures for rewriting 1916 these fields are the same for both RTCP ECN feedback packet and the 1917 RTCP XR ECN summary packet. 1919 When receiving an RTCP ECN feedback packet for the translated stream, 1920 an RTP translator first determines the range of packets to which the 1921 report corresponds. The extended highest sequence number in the RTCP 1922 ECN feedback packet (or in the RTCP SR/RR packet contained within the 1923 compound packet, in the case of RTCP XR ECN summary reports) 1924 specifies the end sequence number of the range. For the first RTCP 1925 ECN feedback packet received, the initial extended sequence number of 1926 the range may be determined by subtracting the sum of the lost 1927 packets counter, the ECN-CE counter, the not-ECT counter, the ECT(0) 1928 counter and the ECT(1) counter minus the duplication counter, from 1929 the extended highest sequence number. For subsequent RTCP ECN 1930 feedback packets, the starting sequence number may be determined as 1931 being one after the extended highest sequence number of the previous 1932 RTCP ECN feedback packet received from the same SSRC. These values 1933 are in the sequence number space of the translated packets. 1935 Based on its knowledge of the translation process, the translator 1936 determines the sequence number range for the corresponding original, 1937 pre-translation, packets. The extended highest sequence number in 1938 the RTCP ECN feedback packet is rewritten to match the final sequence 1939 number in the pre-translation sequence number range. 1941 The translator then determines the ratio, R, of the number of packets 1942 in the translated sequence number space (numTrans) to the number of 1943 packets in the pre-translation sequence number space (numOrig) such 1944 that R = numTrans / numOrig. The counter values in the RTCP ECN 1945 feedback report are then scaled by dividing each of them by R. For 1946 example, if the translation process combines two RTP packets into 1947 one, then numOrig will be twice numTrans, giving R=0.5, and the 1948 counters in the translated RTCP ECN feedback packet will be twice 1949 those in the original. 1951 The ratio, R, may have a value that leads to non-integer multiples of 1952 the counters when translating the RTCP packet. For example, a VoIP 1953 translator that combines two adjacent RTP packets into one if they 1954 contain active speech data, but passes comfort noise packets 1955 unchanged, would have an R values of between 0.5 and 1.0 depending on 1956 the amount of active speech. Since the counter values in the 1957 translated RTCP report are integer values, rounding will be necessary 1958 in this case. 1960 When rounding counter values in the translated RTCP packet, the 1961 translator should try to ensure that they sum to the number of RTP 1962 packets in the pre-translation sequence number space (numOrig). The 1963 translator should also try to ensure that no non-zero counter is 1964 rounded to a zero value, unless the pre-translated values are zero, 1965 since that will lose information that a particular type of event has 1966 occurred. It is recognised that it may be impossible to satisfy both 1967 of these constraints; in such cases, it is better to ensure that no 1968 non-zero counter is mapped to a zero value, since this preserves 1969 congestion adaptation and helps the RTCP-based ECN initiation 1970 process. 1972 One should be aware of the impact this type of translators have on 1973 the measurement of packet duplication. A translator performing 1974 aggregation and most likely also an fragmenting translator will 1975 suppress any duplication happening prior to itself. Thus the reports 1976 and what is being scaled will only represent packet duplication 1977 happening from the translator to the receiver reporting on the flow. 1979 It should be noted that scaling the RTCP counter values in this way 1980 is meaningful only on the assumption that the level of congestion in 1981 the network is related to the number of packets being sent. This is 1982 likely to be a reasonable assumption in the type of environment where 1983 RTP translators that fragment or reassemble packets are deployed, as 1984 their entire purpose is to change the number of packets being sent to 1985 adapt to known limitations of the network, but is not necessarily 1986 valid in general. 1988 The rewritten RTCP ECN feedback report is sent from the other side of 1989 the translator to that which it arrived (as part of a compound RTCP 1990 packet containing other translated RTCP packets, where appropriate). 1992 8.3. Generating RTCP ECN Feedback in Media Transcoders 1994 An RTP translator that acts as a media transcoder cannot directly 1995 forward RTCP packets corresponding to the transcoded stream, since 1996 those packets will relate to the non-transcoded stream, and will not 1997 be useful in relation to the transcoded RTP flow. Such a transcoder 1998 will need to interpose itself into the RTCP flow, acting as a proxy 1999 for the receiver to generate RTCP feedback in the direction of the 2000 sender relating to the pre-transcoded stream, and acting in place of 2001 the sender to generate RTCP relating to the transcoded stream, to be 2002 sent towards the receiver. This section describes how this proxying 2003 is to be done for RTCP ECN feedback packets. Section 7.2 of 2004 [RFC3550] describes general procedures for other RTCP packet types. 2006 An RTP translator acting as a media transcoder in this manner does 2007 not have its own SSRC, and hence is not visible to other entities at 2008 the RTP layer. RTCP ECN feedback packets and RTCP XR report blocks 2009 for ECN summary information that are received from downstream relate 2010 to the translated stream, and so must be processed by the translator 2011 as if it were the original media source. These reports drive the 2012 congestion control loop and media adaptation between the translator 2013 and the downstream receiver. If there are multiple downstream 2014 receivers, a logically separate transcoder instance must be used for 2015 each receiver, and must process RTCP ECN feedback and summary reports 2016 independently to the other transcoder instances. An RTP translator 2017 acting as a media transcoder in this manner MUST NOT forward RTCP ECN 2018 feedback packets or RTCP XR ECN summary reports from downstream 2019 receivers in the upstream direction. 2021 An RTP translator acting as a media transcoder will generate RTCP 2022 reports upstream towards the original media sender, based on the 2023 reception quality of the original media stream at the translator. 2024 The translator will run a separate congestion control loop and media 2025 adaptation between itself and the media sender for each of its 2026 downstream receivers, and must generate RTCP ECN feedback packets and 2027 RTCP XR ECN summary reports for that congestion control loop using 2028 the SSRC of that downstream receiver. 2030 8.4. Generating RTCP ECN Feedback in Mixers 2032 An RTP mixer terminates one-or-more RTP flows, combines them into a 2033 single outgoing media stream, and transmits that new stream as a 2034 separate RTP flow. A mixer has its own SSRC, and is visible to other 2035 participants in the session at the RTP layer. 2037 An ECN-aware RTP mixer must generate RTCP ECN feedback packets and 2038 RTCP XR report blocks for ECN summary information relating to the RTP 2039 flows it terminates, in exactly the same way it would if it were an 2040 RTP receiver. These reports form part of the congestion control loop 2041 between the mixer and the media senders generating the streams it is 2042 mixing. A separate control loop runs between each sender and the 2043 mixer. 2045 An ECN-aware RTP mixer will negotiate and initiate the use of ECN on 2046 the mixed RTP flows it generates, and will accept and process RTCP 2047 ECN feedback reports and RTCP XR report blocks for ECN relating to 2048 those mixed flows as if it were a standard media sender. A 2049 congestion control loop runs between the mixer and its receivers, 2050 driven in part by the ECN reports received. 2052 An RTP mixer MUST NOT forward RTCP ECN feedback packets or RTCP XR 2053 ECN summary reports from downstream receivers in the upstream 2054 direction. 2056 9. Implementation considerations 2058 To allow the use of ECN with RTP over UDP, an RTP implementation 2059 desiring to support receiving ECN controlled media streams must 2060 support reading the value of the ECT bits on received UDP datagrams, 2061 and an RTP implementation desiring to support sending ECN controlled 2062 media streams must support setting the ECT bits in outgoing UDP 2063 datagrams. The standard Berkeley sockets API pre-dates the 2064 specification of ECN, and does not provide the functionality which is 2065 required for this mechanism to be used with UDP flows, making this 2066 specification difficult to implement portably. 2068 10. IANA Considerations 2070 Note to RFC Editor: please replace "RFC XXXX" below with the RFC 2071 number of this memo, and remove this note. 2073 10.1. SDP Attribute Registration 2075 Following the guidelines in [RFC4566], the IANA is requested to 2076 register one new SDP attribute: 2078 o Contact name, email address and telephone number: Authors of 2079 RFCXXXX 2081 o Attribute-name: ecn-capable-rtp 2083 o Type of attribute: media-level 2085 o Subject to charset: no 2087 This attribute defines the ability to negotiate the use of ECT (ECN 2088 capable transport) for RTP flows running over UDP/IP. This attribute 2089 should be put in the SDP offer if the offering party wishes to 2090 receive an ECT flow. The answering party should include the 2091 attribute in the answer if it wish to receive an ECT flow. If the 2092 answerer does not include the attribute then ECT MUST be disabled in 2093 both directions. 2095 10.2. RTP/AVPF Transport Layer Feedback Message 2097 The IANA is requested to register one new RTP/AVPF Transport Layer 2098 Feedback Message in the table of FMT values for RTPFB Payload Types 2099 [RFC4585] as defined in Section 5.1: 2101 Name: RTCP-ECN-FB 2102 Long name: RTCP ECN Feedback 2103 Value: TBA1 2104 Reference: RFC XXXX 2106 10.3. RTCP Feedback SDP Parameter 2108 The IANA is requested to register one new SDP "rtcp-fb" attribute 2109 "nack" parameter "ecn" in the SDP ("ack" and "nack" Attribute Values) 2110 registry. 2111 Value name: ecn 2112 Long name: Explicit Congestion Notification 2113 Usable with: nack 2114 Reference: RFC XXXX 2116 10.4. RTCP XR Report blocks 2118 The IANA is requested to register one new RTCP XR Block Type as 2119 defined in Section 5.2: 2121 Block Type: TBA2 2122 Name: ECN Summary Report 2123 Reference: RFC XXXX 2125 10.5. RTCP XR SDP Parameter 2127 The IANA is requested to register one new RTCP XR SDP Parameter "ecn- 2128 sum" in the "RTCP XR SDP Parameters" registry. 2129 Parameter name XR block (block type and name) 2130 -------------- ------------------------------------ 2131 ecn-sum TBA2 ECN Summary Report Block 2133 10.6. STUN attribute 2135 A new STUN [RFC5389] attribute in the Comprehension-optional range 2136 under IETF Review (0x8000-0xFFFF) is request to be assigned to the 2137 STUN attribute defined in Section 7.2.2. The STUN attribute registry 2138 can currently be found at: http://www.iana.org/assignments/ 2139 stun-parameters/stun-parameters.xhtml. 2141 10.7. ICE Option 2143 A new ICE option "rtp+ecn" is registered in the registry that "IANA 2144 Registry for Interactive Connectivity Establishment (ICE) Options" 2145 [RFC6336] creates. 2147 11. Security Considerations 2149 The use of ECN with RTP over UDP as specified in this document has 2150 the following known security issues that need to be considered. 2152 External threats to the RTP and RTCP traffic: 2154 Denial of Service affecting RTCP: An attacker that can modify the 2155 traffic between the media sender and a receiver can achieve either 2156 of two things: 1) Report a lot of packets as being Congestion 2157 Experience marked, thus forcing the sender into a congestion 2158 response; or 2) Ensure that the sender disable the usage of ECN by 2159 reporting failures to receive ECN by changing the counter fields. 2160 This can also be accomplished by injecting false RTCP packets to 2161 the media sender. Reporting a lot of ECN-CE marked traffic is 2162 likely the more efficient denial of service tool as that may 2163 likely force the application to use lowest possible bit-rates. 2164 The prevention against an external threat is to integrity protect 2165 the RTCP feedback information and authenticate the sender. 2167 Information leakage: The ECN feedback mechanism exposes the 2168 receivers perceived packet loss, what packets it considers to be 2169 ECN-CE marked and its calculation of the ECN-none. This is mostly 2170 not considered as sensitive information. If it is considered 2171 sensitive the RTCP feedback should be encrypted. 2173 Changing the ECN bits: An on-path attacker that sees the RTP packet 2174 flow from sender to receiver and who has the capability to change 2175 the packets can rewrite ECT into ECN-CE thus forcing the sender or 2176 receiver to take congestion control response. This denial of 2177 service against the media quality in the RTP session is impossible 2178 for an end-point to protect itself against. Only network 2179 infrastructure nodes can detect this illicit re-marking. It will 2180 be mitigated by turning off ECN, however, if the attacker can 2181 modify its response to drop packets the same vulnerability exist. 2183 Denial of Service affecting the session set-up signalling: If an 2184 attacker can modify the session signalling it can prevent the 2185 usage of ECN by removing the signalling attributes used to 2186 indicate that the initiator is capable and willing to use ECN with 2187 RTP/UDP. This attack can be prevented by authentication and 2188 integrity protection of the signalling. We do note that any 2189 attacker that can modify the signalling has more interesting 2190 attacks they can perform than prevent the usage of ECN, like 2191 inserting itself as a middleman in the media flows enabling wire- 2192 tapping also for an off-path attacker. 2194 The following are threats that exist from misbehaving senders or 2195 receivers: 2197 Receivers cheating: A receiver may attempt to cheat and fail to 2198 report reception of ECN-CE marked packets. The benefit for a 2199 receiver cheating in its reporting would be to get an unfair bit- 2200 rate share across the resource bottleneck. It is far from certain 2201 that a receiver would be able to get a significant larger share of 2202 the resources. That assumes a high enough level of aggregation 2203 that there are flows to acquire shares from. The risk of cheating 2204 is that failure to react to congestion results in packet loss and 2205 increased path delay. 2207 Receivers misbehaving: A receiver may prevent the usage of ECN in an 2208 RTP session by reporting itself as non ECN capable, forcing the 2209 sender to turn off usage of ECN. In a point-to-point scenario 2210 there is little incentive to do this as it will only affect the 2211 receiver. Thus failing to utilise an optimisation. For multi- 2212 party session there exist some motivation why a receiver would 2213 misbehave as it can prevent also the other receivers from using 2214 ECN. As an insider into the session it is difficult to determine 2215 if a receiver is misbehaving or simply incapable, making it 2216 basically impossible in the incremental deployment phase of ECN 2217 for RTP usage to determine this. If additional information about 2218 the receivers and the network is known it might be possible to 2219 deduce that a receiver is misbehaving. If it can be determined 2220 that a receiver is misbehaving, the only response is to exclude it 2221 from the RTP session and ensure that is does not any longer have 2222 any valid security context to affect the session. 2224 Misbehaving Senders: The enabling of ECN gives the media packets a 2225 higher degree of probability to reach the receiver compared to 2226 not-ECT marked ones on a ECN capable path. However, this is no 2227 magic bullet and failure to react to congestion will most likely 2228 only slightly delay a network buffer over-run, in which its 2229 session also will experience packet loss and increased delay. 2230 There is some possibility that the media senders traffic will push 2231 other traffic out of the way without being affected too 2232 negatively. However, we do note that a media sender still needs 2233 to implement congestion control functions to prevent the media 2234 from being badly affected by congestion events. Thus the 2235 misbehaving sender is getting a unfair share. This can only be 2236 detected and potentially prevented by network monitoring and 2237 administrative entities. See Section 7 of [RFC3168] for more 2238 discussion of this issue. 2240 We note that the end-point security functions needed to prevent an 2241 external attacker from inferring with the signalling are source 2242 authentication and integrity protection. To prevent information 2243 leakage from the feedback packets encryption of the RTCP is also 2244 needed. For RTP there exist multiple solutions possible depending on 2245 the application context. Secure RTP (SRTP) [RFC3711] does satisfy 2246 the requirement to protect this mechanism despite only providing 2247 authentication if a entity is within the security context or not. 2248 IPsec [RFC4301] and DTLS [RFC4347] can also provide the necessary 2249 security functions. 2251 The signalling protocols used to initiate an RTP session also need to 2252 be source authenticated and integrity protected to prevent an 2253 external attacker from modifying any signalling. Here an appropriate 2254 mechanism to protect the used signalling needs to be used. For SIP/ 2255 SDP ideally S/MIME [RFC5751] would be used. However, with the 2256 limited deployment a minimal mitigation strategy is to require use of 2257 SIPS (SIP over TLS) [RFC3261] [RFC5630] to at least accomplish hop- 2258 by-hop protection. 2260 We do note that certain mitigation methods will require network 2261 functions. 2263 12. Examples of SDP Signalling 2265 This section contain a few different examples of the signalling 2266 mechanism defined in this specification in an SDP context. If there 2267 are discrepancies between these examples and the specification text, 2268 the specification text is definitive. 2270 12.1. Basic SDP Offer/Answer 2272 This example is a basic offer/answer SDP exchange, assumed done by 2273 SIP (not shown). The intention is to establish a basic audio session 2274 point to point between two users. 2276 The Offer: 2278 v=0 2279 o=jdoe 3502844782 3502844782 IN IP4 10.0.1.4 2280 s=VoIP call 2281 i=SDP offer for VoIP call with ICE and ECN for RTP 2282 b=AS:128 2283 b=RR:2000 2284 b=RS:2500 2285 a=ice-pwd:YH75Fviy6338Vbrhrlp8Yh 2286 a=ice-ufrag:9uB6 2287 a=ice-options:rtp+ecn 2288 t=0 0 2289 m=audio 45664 RTP/AVPF 97 98 99 2290 c=IN IP4 192.0.2.3 2291 a=rtpmap:97 G719/48000/1 2292 a=fmtp:97 maxred=160 2293 a=rtpmap:98 AMR-WB/16000/1 2294 a=fmtp:98 octet-align=1; mode-change-capability=2 2295 a=rtpmap:99 PCMA/8000/1 2296 a=maxptime:160 2297 a=ptime:20 2298 a=ecn-capable-rtp: ice rtp ect=0 mode=setread 2299 a=rtcp-fb:* nack ecn 2300 a=rtcp-fb:* trr-int 1000 2301 a=rtcp-xr:ecn-sum 2302 a=rtcp-rsize 2303 a=candidate:1 1 UDP 2130706431 10.0.1.4 8998 typ host 2304 a=candidate:2 1 UDP 1694498815 192.0.2.3 45664 typ srflx raddr 2305 10.0.1.4 rport 8998 2307 This SDP offer offers a single media stream with 3 media payload 2308 types. It proposes to use ECN with RTP, with the ICE based 2309 initialization as being preferred over the RTP/RTCP one. Leap of 2310 faith is not suggested to be used. The offerer is capable of both 2311 setting and reading the ECN bits. In addition the use of both the 2312 RTCP ECN feedback packet and the RTCP XR ECN summary report are 2313 supported. ICE is also proposed with two candidates. It also 2314 supports reduced size RTCP and can to use it. 2316 The Answer: 2318 v=0 2319 o=jdoe 3502844783 3502844783 IN IP4 198.51.100.235 2320 s=VoIP call 2321 i=SDP offer for VoIP call with ICE and ECN for RTP 2322 b=AS:128 2323 b=RR:2000 2324 b=RS:2500 2325 a=ice-pwd:asd88fgpdd777uzjYhagZg 2326 a=ice-ufrag:8hhY 2327 a=ice-options:rtp+ecn 2328 t=0 0 2329 m=audio 53879 RTP/AVPF 97 99 2330 c=IN IP4 198.51.100.235 2331 a=rtpmap:97 G719/48000/1 2332 a=fmtp:97 maxred=160 2333 a=rtpmap:99 PCMA/8000/1 2334 a=maxptime:160 2335 a=ptime:20 2336 a=ecn-capable-rtp: ice ect=0 mode=readonly 2337 a=rtcp-fb:* nack ecn 2338 a=rtcp-fb:* trr-int 1000 2339 a=rtcp-xr:ecn-sum 2340 a=candidate:1 1 UDP 2130706431 198.51.100.235 53879 typ host 2342 The answer confirms that only one media stream will be used. One RTP 2343 Payload type was removed. ECN capability was confirmed, and the 2344 initialization method will be ICE. However, the answerer is only 2345 capable of reading the ECN bits, which means that ECN can only be 2346 used for RTP flowing from the offerer to the answerer. ECT always 2347 set to 0 will be used in both directions. Both the RTCP ECN feedback 2348 packet and the RTCP XR ECN summary report will be used. Reduced size 2349 RTCP will not be used as the answerer has not indicated support for 2350 it in the answer. 2352 12.2. Declarative Multicast SDP 2354 The below session describes an any source multicast using session 2355 with a single media stream. 2357 v=0 2358 o=jdoe 3502844782 3502844782 IN IP4 198.51.100.235 2359 s=Multicast SDP session using ECN for RTP 2360 i=Multicasted audio chat using ECN for RTP 2361 b=AS:128 2362 t=3502892703 3502910700 2363 m=audio 56144 RTP/AVPF 97 2364 c=IN IP4 233.252.0.212/127 2365 a=rtpmap:97 g719/48000/1 2366 a=fmtp:97 maxred=160 2367 a=maxptime:160 2368 a=ptime:20 2369 a=ecn-capable-rtp: rtp mode=readonly; ect=0 2370 a=rtcp-fb:* nack ecn 2371 a=rtcp-fb:* trr-int 1500 2372 a=rtcp-xr:ecn-sum 2374 In the above example, as this is declarative we need to require 2375 certain functionality. As it is ASM the initialization method that 2376 can work here is the RTP/RTCP based one. So that is indicated. The 2377 ECN setting and reading capability to take part of this session is at 2378 least read. If one is capable of setting that is good, but not 2379 required as one can skip using ECN for anything one sends oneself. 2380 The ECT value is recommended to be set to 0 always. The ECN usage in 2381 this session requires both ECN feedback and the XR ECN summary 2382 report, so their use is also indicated. 2384 13. Acknowledgments 2386 The authors wish to thank the following persons for their reviews and 2387 comments: Thomas Belling, Bob Briscoe, Roni Even, Kevin P. Flemming, 2388 Thomas Frankkila, Christian Groves, Christer Holmgren, Cullen 2389 Jennings Tom Van Caenegem, Simo Veikkolainen, Bill Ver Steeg, Dan 2390 Wing, Qin Wu, and Lei Zhu. 2392 14. References 2394 14.1. Normative References 2396 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 2397 Requirement Levels", BCP 14, RFC 2119, March 1997. 2399 [RFC3168] Ramakrishnan, K., Floyd, S., and D. Black, "The Addition 2400 of Explicit Congestion Notification (ECN) to IP", 2401 RFC 3168, September 2001. 2403 [RFC3550] Schulzrinne, H., Casner, S., Frederick, R., and V. 2404 Jacobson, "RTP: A Transport Protocol for Real-Time 2405 Applications", STD 64, RFC 3550, July 2003. 2407 [RFC3611] Friedman, T., Caceres, R., and A. Clark, "RTP Control 2408 Protocol Extended Reports (RTCP XR)", RFC 3611, 2409 November 2003. 2411 [RFC5234] Crocker, D. and P. Overell, "Augmented BNF for Syntax 2412 Specifications: ABNF", STD 68, RFC 5234, January 2008. 2414 [RFC5245] Rosenberg, J., "Interactive Connectivity Establishment 2415 (ICE): A Protocol for Network Address Translator (NAT) 2416 Traversal for Offer/Answer Protocols", RFC 5245, 2417 April 2010. 2419 [RFC5348] Floyd, S., Handley, M., Padhye, J., and J. Widmer, "TCP 2420 Friendly Rate Control (TFRC): Protocol Specification", 2421 RFC 5348, September 2008. 2423 [RFC5389] Rosenberg, J., Mahy, R., Matthews, P., and D. Wing, 2424 "Session Traversal Utilities for NAT (STUN)", RFC 5389, 2425 October 2008. 2427 [RFC6336] Westerlund, M. and C. Perkins, "IANA Registry for 2428 Interactive Connectivity Establishment (ICE) Options", 2429 RFC 6336, July 2011. 2431 14.2. Informative References 2433 [I-D.ietf-avt-rtp-no-op] 2434 Andreasen, F., "A No-Op Payload Format for RTP", 2435 draft-ietf-avt-rtp-no-op-04 (work in progress), May 2007. 2437 [RFC1112] Deering, S., "Host extensions for IP multicasting", STD 5, 2438 RFC 1112, August 1989. 2440 [RFC2762] Rosenberg, J. and H. Schulzrinne, "Sampling of the Group 2441 Membership in RTP", RFC 2762, February 2000. 2443 [RFC2974] Handley, M., Perkins, C., and E. Whelan, "Session 2444 Announcement Protocol", RFC 2974, October 2000. 2446 [RFC3261] Rosenberg, J., Schulzrinne, H., Camarillo, G., Johnston, 2447 A., Peterson, J., Sparks, R., Handley, M., and E. 2448 Schooler, "SIP: Session Initiation Protocol", RFC 3261, 2449 June 2002. 2451 [RFC3264] Rosenberg, J. and H. Schulzrinne, "An Offer/Answer Model 2452 with Session Description Protocol (SDP)", RFC 3264, 2453 June 2002. 2455 [RFC3540] Spring, N., Wetherall, D., and D. Ely, "Robust Explicit 2456 Congestion Notification (ECN) Signaling with Nonces", 2457 RFC 3540, June 2003. 2459 [RFC3551] Schulzrinne, H. and S. Casner, "RTP Profile for Audio and 2460 Video Conferences with Minimal Control", STD 65, RFC 3551, 2461 July 2003. 2463 [RFC3569] Bhattacharyya, S., "An Overview of Source-Specific 2464 Multicast (SSM)", RFC 3569, July 2003. 2466 [RFC3711] Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K. 2467 Norrman, "The Secure Real-time Transport Protocol (SRTP)", 2468 RFC 3711, March 2004. 2470 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 2471 Internet Protocol", RFC 4301, December 2005. 2473 [RFC4340] Kohler, E., Handley, M., and S. Floyd, "Datagram 2474 Congestion Control Protocol (DCCP)", RFC 4340, March 2006. 2476 [RFC4347] Rescorla, E. and N. Modadugu, "Datagram Transport Layer 2477 Security", RFC 4347, April 2006. 2479 [RFC4566] Handley, M., Jacobson, V., and C. Perkins, "SDP: Session 2480 Description Protocol", RFC 4566, July 2006. 2482 [RFC4585] Ott, J., Wenger, S., Sato, N., Burmeister, C., and J. Rey, 2483 "Extended RTP Profile for Real-time Transport Control 2484 Protocol (RTCP)-Based Feedback (RTP/AVPF)", RFC 4585, 2485 July 2006. 2487 [RFC4588] Rey, J., Leon, D., Miyazaki, A., Varsa, V., and R. 2488 Hakenberg, "RTP Retransmission Payload Format", RFC 4588, 2489 July 2006. 2491 [RFC4607] Holbrook, H. and B. Cain, "Source-Specific Multicast for 2492 IP", RFC 4607, August 2006. 2494 [RFC4960] Stewart, R., "Stream Control Transmission Protocol", 2495 RFC 4960, September 2007. 2497 [RFC5124] Ott, J. and E. Carrara, "Extended Secure RTP Profile for 2498 Real-time Transport Control Protocol (RTCP)-Based Feedback 2499 (RTP/SAVPF)", RFC 5124, February 2008. 2501 [RFC5506] Johansson, I. and M. Westerlund, "Support for Reduced-Size 2502 Real-Time Transport Control Protocol (RTCP): Opportunities 2503 and Consequences", RFC 5506, April 2009. 2505 [RFC5630] Audet, F., "The Use of the SIPS URI Scheme in the Session 2506 Initiation Protocol (SIP)", RFC 5630, October 2009. 2508 [RFC5751] Ramsdell, B. and S. Turner, "Secure/Multipurpose Internet 2509 Mail Extensions (S/MIME) Version 3.2 Message 2510 Specification", RFC 5751, January 2010. 2512 [RFC5760] Ott, J., Chesterfield, J., and E. Schooler, "RTP Control 2513 Protocol (RTCP) Extensions for Single-Source Multicast 2514 Sessions with Unicast Feedback", RFC 5760, February 2010. 2516 [RFC6189] Zimmermann, P., Johnston, A., and J. Callas, "ZRTP: Media 2517 Path Key Agreement for Unicast Secure RTP", RFC 6189, 2518 April 2011. 2520 Authors' Addresses 2522 Magnus Westerlund 2523 Ericsson 2524 Farogatan 6 2525 SE-164 80 Kista 2526 Sweden 2528 Phone: +46 10 714 82 87 2529 Email: magnus.westerlund@ericsson.com 2531 Ingemar Johansson 2532 Ericsson 2533 Laboratoriegrand 11 2534 SE-971 28 Lulea 2535 SWEDEN 2537 Phone: +46 73 0783289 2538 Email: ingemar.s.johansson@ericsson.com 2539 Colin Perkins 2540 University of Glasgow 2541 School of Computing Science 2542 Glasgow G12 8QQ 2543 United Kingdom 2545 Email: csp@csperkins.org 2547 Piers O'Hanlon 2548 University College London 2549 Computer Science Department 2550 Gower Street 2551 London WC1E 6BT 2552 United Kingdom 2554 Email: p.ohanlon@cs.ucl.ac.uk 2556 Ken Carlberg 2557 G11 2558 1600 Clarendon Blvd 2559 Arlington VA 2560 USA 2562 Email: carlberg@g11.org.uk