idnits 2.17.1 draft-ietf-payload-flexible-fec-scheme-14.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 : ---------------------------------------------------------------------------- -- 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 (January 3, 2019) is 1933 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: '0-14' is mentioned on line 766, but not defined == Missing Reference: '15-45' is mentioned on line 770, but not defined == Missing Reference: '46-109' is mentioned on line 774, but not defined == Missing Reference: 'RFCXXXX' is mentioned on line 1213, but not defined ** Obsolete normative reference: RFC 4566 (Obsoleted by RFC 8866) -- Obsolete informational reference (is this intentional?): RFC 2326 (Obsoleted by RFC 7826) -- Obsolete informational reference (is this intentional?): RFC 2733 (Obsoleted by RFC 5109) Summary: 1 error (**), 0 flaws (~~), 5 warnings (==), 4 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 PAYLOAD M. Zanaty 3 Internet-Draft Cisco 4 Intended status: Standards Track V. Singh 5 Expires: July 7, 2019 callstats.io 6 A. Begen 7 Networked Media 8 G. Mandyam 9 Qualcomm Inc. 10 January 3, 2019 12 RTP Payload Format for Flexible Forward Error Correction (FEC) 13 draft-ietf-payload-flexible-fec-scheme-14 15 Abstract 17 This document defines new RTP payload formats for the Forward Error 18 Correction (FEC) packets that are generated by the non-interleaved 19 and interleaved parity codes from source media encapsulated in RTP. 20 These parity codes are systematic codes, where a number of FEC repair 21 packets are generated from a set of source packets from one or more 22 source RTP streams. These FEC repair packets are sent in a 23 redundancy RTP stream separate from the source RTP stream(s) that 24 carries the source packets. RTP source packets that were lost in 25 transmission can be reconstructed using the source and repair packets 26 that were received. The non-interleaved and interleaved parity codes 27 which are defined in this specification offer a good protection 28 against random and bursty packet losses, respectively, at a cost of 29 complexity. The RTP payload formats that are defined in this 30 document address scalability issues experienced with the earlier 31 specifications, and offer several improvements. Due to these 32 changes, the new payload formats are not backward compatible with 33 earlier specifications, but endpoints that do not implement this 34 specification can still work by simply ignoring the FEC repair 35 packets. 37 Status of This Memo 39 This Internet-Draft is submitted in full conformance with the 40 provisions of BCP 78 and BCP 79. 42 Internet-Drafts are working documents of the Internet Engineering 43 Task Force (IETF). Note that other groups may also distribute 44 working documents as Internet-Drafts. The list of current Internet- 45 Drafts is at https://datatracker.ietf.org/drafts/current/. 47 Internet-Drafts are draft documents valid for a maximum of six months 48 and may be updated, replaced, or obsoleted by other documents at any 49 time. It is inappropriate to use Internet-Drafts as reference 50 material or to cite them other than as "work in progress." 52 This Internet-Draft will expire on July 7, 2019. 54 Copyright Notice 56 Copyright (c) 2019 IETF Trust and the persons identified as the 57 document authors. All rights reserved. 59 This document is subject to BCP 78 and the IETF Trust's Legal 60 Provisions Relating to IETF Documents 61 (https://trustee.ietf.org/license-info) in effect on the date of 62 publication of this document. Please review these documents 63 carefully, as they describe your rights and restrictions with respect 64 to this document. Code Components extracted from this document must 65 include Simplified BSD License text as described in Section 4.e of 66 the Trust Legal Provisions and are provided without warranty as 67 described in the Simplified BSD License. 69 Table of Contents 71 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 72 1.1. Parity Codes . . . . . . . . . . . . . . . . . . . . . . 4 73 1.1.1. One-Dimensionsal (1-D) Non-interleaved (Row) FEC 74 Protection . . . . . . . . . . . . . . . . . . . . . 6 75 1.1.2. 1-D Interleaved (Column) FEC Protection . . . . . . . 7 76 1.1.3. Use Cases for 1-D FEC Protection . . . . . . . . . . 8 77 1.1.4. Two-Dimensional (2-D) (Row and Column) FEC Protection 10 78 1.1.5. FEC Overhead Considerations . . . . . . . . . . . . . 12 79 2. Requirements Notation . . . . . . . . . . . . . . . . . . . . 13 80 3. Definitions and Notations . . . . . . . . . . . . . . . . . . 13 81 3.1. Definitions . . . . . . . . . . . . . . . . . . . . . . . 13 82 3.2. Notations . . . . . . . . . . . . . . . . . . . . . . . . 14 83 4. Packet Formats . . . . . . . . . . . . . . . . . . . . . . . 14 84 4.1. Source Packets . . . . . . . . . . . . . . . . . . . . . 14 85 4.2. FEC Repair Packets . . . . . . . . . . . . . . . . . . . 15 86 4.2.1. RTP Header of FEC Repair Packets . . . . . . . . . . 16 87 4.2.2. FEC Header of FEC Repair Packets . . . . . . . . . . 17 88 5. Payload Format Parameters . . . . . . . . . . . . . . . . . . 25 89 5.1. Media Type Registration - Parity Codes . . . . . . . . . 25 90 5.1.1. Registration of audio/flexfec . . . . . . . . . . . . 25 91 5.1.2. Registration of video/flexfec . . . . . . . . . . . . 26 92 5.1.3. Registration of text/flexfec . . . . . . . . . . . . 28 93 5.1.4. Registration of application/flexfec . . . . . . . . . 29 94 5.2. Mapping to SDP Parameters . . . . . . . . . . . . . . . . 31 95 5.2.1. Offer-Answer Model Considerations . . . . . . . . . . 31 96 5.2.2. Declarative Considerations . . . . . . . . . . . . . 32 98 6. Protection and Recovery Procedures - Parity Codes . . . . . . 32 99 6.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . 32 100 6.2. Repair Packet Construction . . . . . . . . . . . . . . . 33 101 6.3. Source Packet Reconstruction . . . . . . . . . . . . . . 34 102 6.3.1. Associating the Source and Repair Packets . . . . . . 35 103 6.3.2. Recovering the RTP Header . . . . . . . . . . . . . . 37 104 6.3.3. Recovering the RTP Payload . . . . . . . . . . . . . 38 105 6.3.4. Iterative Decoding Algorithm for the 2-D Parity FEC 106 Protection . . . . . . . . . . . . . . . . . . . . . 38 107 7. Signaling Requirements . . . . . . . . . . . . . . . . . . . 42 108 7.1. SDP Examples . . . . . . . . . . . . . . . . . . . . . . 43 109 7.1.1. Example SDP for Flexible FEC Protection with in-band 110 SSRC mapping . . . . . . . . . . . . . . . . . . . . 43 111 7.1.2. Example SDP for Flexible FEC Protection with explicit 112 signalling in the SDP . . . . . . . . . . . . . . . . 44 113 7.2. On the Use of the RTP Stream Identifier Source 114 Description . . . . . . . . . . . . . . . . . . . . . . . 44 115 8. Congestion Control Considerations . . . . . . . . . . . . . . 45 116 9. Security Considerations . . . . . . . . . . . . . . . . . . . 45 117 10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 46 118 11. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 46 119 12. References . . . . . . . . . . . . . . . . . . . . . . . . . 46 120 12.1. Normative References . . . . . . . . . . . . . . . . . . 46 121 12.2. Informative References . . . . . . . . . . . . . . . . . 47 122 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 49 124 1. Introduction 126 This document defines new RTP payload formats for the Forward Error 127 Correction (FEC) that is generated by the non-interleaved and 128 interleaved parity codes from a source media encapsulated in RTP 129 [RFC3550]. The type of the source media protected by these parity 130 codes can be audio, video, text or application. The FEC data are 131 generated according to the media type parameters, which are 132 communicated out-of-band (e.g., in SDP). Furthermore, the 133 associations or relationships between the source and repair RTP 134 streams may be communicated in-band or out-of-band. The in-band 135 mechanism is advantageous when the endpoint is adapting the FEC 136 parameters. The out-of-band mechanism may be preferable when the FEC 137 parameters are fixed. While this document fully defines the use of 138 FEC to protect RTP streams, it also leverages several definitions 139 along with the basic source/repair header description from [RFC6363] 140 in their application to the parity codes defined here. 142 The Redundancy RTP Stream [RFC7656] repair packets proposed in this 143 document protect the Source RTP Stream packets that belong to the 144 same RTP session. 146 The RTP payload formats that are defined in this document address the 147 scalability issues experienced with the formats defined in earlier 148 specifications including [RFC2733], [RFC5109] and [SMPTE2022-1]. 150 1.1. Parity Codes 152 Both the non-interleaved and interleaved parity codes use the 153 eXclusive OR (XOR) operation to generate the repair packets. The 154 following steps take place: 156 1. The sender determines a set of source packets to be protected by 157 FEC based on the media type parameters. 159 2. The sender applies the XOR operation on the source packets to 160 generate the required number of repair packets. 162 3. The sender sends the repair packet(s) along with the source 163 packets, in different RTP streams, to the receiver(s). The 164 repair packets may be sent proactively or on-demand based on RTCP 165 feedback messages such as NACK [RFC4585]. 167 At the receiver side, if all of the source packets are successfully 168 received, there is no need for FEC recovery and the repair packets 169 are discarded. However, if there are missing source packets, the 170 repair packets can be used to recover the missing information. 171 Figure 1 and Figure 2 describe example block diagrams for the 172 systematic parity FEC encoder and decoder, respectively. 174 +------------+ 176 +--+ +--+ +--+ +--+ --> | Systematic | --> +--+ +--+ +--+ +--+ 178 +--+ +--+ +--+ +--+ | Parity FEC | +--+ +--+ +--+ +--+ 180 | Encoder | 182 | (Sender) | --> +==+ +==+ 184 +------------+ +==+ +==+ 186 Source Packet: +--+ Repair Packet: +==+ 188 +--+ +==+ 190 Figure 1: Block diagram for systematic parity FEC encoder 192 +------------+ 194 +--+ X X +--+ --> | Systematic | --> +--+ +--+ +--+ +--+ 196 +--+ +--+ | Parity FEC | +--+ +--+ +--+ +--+ 198 | Decoder | 200 +==+ +==+ --> | (Receiver) | 202 +==+ +==+ +------------+ 204 Source Packet: +--+ Repair Packet: +==+ Lost Packet: X 206 +--+ +==+ 208 Figure 2: Block diagram for systematic parity FEC decoder 210 In Figure 2, it is clear that the FEC repair packets have to be 211 received by the endpoint within a certain amount of time for the FEC 212 recovery process to be useful. The repair window is defined as the 213 time that spans a FEC block, which consists of the source packets and 214 the corresponding repair packets. At the receiver side, the FEC 215 decoder SHOULD buffer source and repair packets at least for the 216 duration of the repair window, to allow all the repair packets to 217 arrive. The FEC decoder can start decoding the already received 218 packets sooner; however, it should not register a FEC decoding 219 failure until it waits at least for the duration of the repair 220 window. 222 1.1.1. One-Dimensionsal (1-D) Non-interleaved (Row) FEC Protection 224 Consider a group of D x L source packets that have sequence numbers 225 starting from 1 running to D x L, and a repair packet is generated by 226 applying the XOR operation to every L consecutive packets as sketched 227 in Figure 3. This process is referred to as 1-D non-interleaved FEC 228 protection. As a result of this process, D repair packets are 229 generated, which are referred to as non-interleaved (or row) FEC 230 repair packets. In general D and L represent values that describe 231 how packets are grouped together from a depth and length perspective 232 (respectively) when interleaving all D x L source packets. 234 +--------------------------------------------------+ --- +===+ 236 | S_1 S_2 S3 ... S_L | + |XOR| = |R_1| 238 +--------------------------------------------------+ --- +===+ 240 +--------------------------------------------------+ --- +===+ 242 | S_L+1 S_L+2 S_L+3 ... S_2xL | + |XOR| = |R_2| 244 +--------------------------------------------------+ --- +===+ 246 . . . . . . 248 . . . . . . 250 . . . . . . 252 +--------------------------------------------------+ --- +===+ 254 | S_(D-1)xL+1 S_(D-1)xL+2 S_(D-1)xL+3 ... S_DxL | + |XOR| = |R_D| 256 +--------------------------------------------------+ --- +===+ 258 Figure 3: Generating non-interleaved (row) FEC repair packets 260 1.1.2. 1-D Interleaved (Column) FEC Protection 262 If the XOR operation is applied to the group of the source packets 263 whose sequence numbers are L apart from each other, as sketched in 264 Figure 4. In this case the endpoint generates L repair packets. 265 This process is referred to as 1-D interleaved FEC protection, and 266 the resulting L repair packets are referred to as interleaved (or 267 column) FEC repair packets. 269 +-------------+ +-------------+ +-------------+ +-------+ 271 | S_1 | | S_2 | | S3 | ... | S_L | 273 | S_L+1 | | S_L+2 | | S_L+3 | ... | S_2xL | 275 | . | | . | | | | | 277 | . | | . | | | | | 279 | . | | . | | | | | 281 | S_(D-1)xL+1 | | S_(D-1)xL+2 | | S_(D-1)xL+3 | ... | S_DxL | 283 +-------------+ +-------------+ +-------------+ +-------+ 285 + + + + 287 ------------- ------------- ------------- ------- 289 | XOR | | XOR | | XOR | ... | XOR | 291 ------------- ------------- ------------- ------- 293 = = = = 295 +===+ +===+ +===+ +===+ 297 |C_1| |C_2| |C_3| ... |C_L| 299 +===+ +===+ +===+ +===+ 301 Figure 4: Generating interleaved (column) FEC repair packets 303 1.1.3. Use Cases for 1-D FEC Protection 305 A sender may generate one non-interleaved repair packet out of L 306 consecutive source packets or one interleaved repair packet out of D 307 non-consecutive source packets. Regardless of whether the repair 308 packet is a non-interleaved or an interleaved one, it can provide a 309 full recovery of the missing information if there is only one packet 310 missing among the corresponding source packets. This implies that 311 1-D non-interleaved FEC protection performs better when the source 312 packets are randomly lost. However, if the packet losses occur in 313 bursts, 1-D interleaved FEC protection performs better provided that 314 L is chosen large enough, i.e., L-packet duration is not shorter than 315 the observed burst duration. If the sender generates non-interleaved 316 FEC repair packets and a burst loss hits the source packets, the 317 repair operation fails. This is illustrated in Figure 5. 319 +---+ +---+ +===+ 321 | 1 | X X | 4 | |R_1| 323 +---+ +---+ +===+ 325 +---+ +---+ +---+ +---+ +===+ 327 | 5 | | 6 | | 7 | | 8 | |R_2| 329 +---+ +---+ +---+ +---+ +===+ 331 +---+ +---+ +---+ +---+ +===+ 333 | 9 | | 10| | 11| | 12| |R_3| 335 +---+ +---+ +---+ +---+ +===+ 337 Figure 5: Example scenario where 1-D non-interleaved FEC protection 338 fails error recovery (Burst Loss) 340 The sender may generate interleaved FEC repair packets to combat with 341 the bursty packet losses. However, two or more random packet losses 342 may hit the source and repair packets in the same column. In that 343 case, the repair operation fails as well. This is illustrated in 344 Figure 6. Note that it is possible that two burst losses may occur 345 back-to-back, in which case interleaved FEC repair packets may still 346 fail to recover the lost data. 348 +---+ +---+ +---+ 350 | 1 | X | 3 | | 4 | 352 +---+ +---+ +---+ 354 +---+ +---+ +---+ 356 | 5 | X | 7 | | 8 | 358 +---+ +---+ +---+ 360 +---+ +---+ +---+ +---+ 362 | 9 | | 10| | 11| | 12| 364 +---+ +---+ +---+ +---+ 366 +===+ +===+ +===+ +===+ 368 |C_1| |C_2| |C_3| |C_4| 370 +===+ +===+ +===+ +===+ 372 Figure 6: Example scenario where 1-D interleaved FEC protection fails 373 error recovery (Periodic Loss) 375 1.1.4. Two-Dimensional (2-D) (Row and Column) FEC Protection 377 In networks where the source packets are lost both randomly and in 378 bursts, the sender ought to generate both non-interleaved and 379 interleaved FEC repair packets. This type of FEC protection is known 380 as 2-D parity FEC protection. At the expense of generating more FEC 381 repair packets, thus increasing the FEC overhead, 2-D FEC provides 382 superior protection against mixed loss patterns. However, it is 383 still possible for 2-D parity FEC protection to fail to recover all 384 of the lost source packets if a particular loss pattern occurs. An 385 example scenario is illustrated in Figure 7. 387 +---+ +---+ +===+ 389 | 1 | X X | 4 | |R_1| 391 +---+ +---+ +===+ 393 +---+ +---+ +---+ +---+ +===+ 395 | 5 | | 6 | | 7 | | 8 | |R_2| 397 +---+ +---+ +---+ +---+ +===+ 399 +---+ +---+ +===+ 401 | 9 | X X | 12| |R_3| 403 +---+ +---+ +===+ 405 +===+ +===+ +===+ +===+ 407 |C_1| |C_2| |C_3| |C_4| 409 +===+ +===+ +===+ +===+ 411 Figure 7: Example scenario #1 where 2-D parity FEC protection fails 412 error recovery 414 2-D parity FEC protection also fails when at least two rows are 415 missing a source and the FEC packet and the missing source packets 416 (in at least two rows) are aligned in the same column. An example 417 loss pattern is sketched in Figure 8. Similarly, 2-D parity FEC 418 protection cannot repair all missing source packets when at least two 419 columns are missing a source and the FEC packet and the missing 420 source packets (in at least two columns) are aligned in the same row. 422 +---+ +---+ +---+ 424 | 1 | | 2 | X | 4 | X 426 +---+ +---+ +---+ 428 +---+ +---+ +---+ +---+ +===+ 430 | 5 | | 6 | | 7 | | 8 | |R_2| 432 +---+ +---+ +---+ +---+ +===+ 434 +---+ +---+ +---+ 436 | 9 | | 10| X | 12| X 438 +---+ +---+ +---+ 440 +===+ +===+ +===+ +===+ 442 |C_1| |C_2| |C_3| |C_4| 444 +===+ +===+ +===+ +===+ 446 Figure 8: Example scenario #2 where 2-D parity FEC protection fails 447 error recovery 449 1.1.5. FEC Overhead Considerations 451 The overhead is defined as the ratio of the number of bytes belonging 452 to the repair packets to the number of bytes belonging to the 453 protected source packets. 455 Generally, repair packets are larger in size compared to the source 456 packets. Also, not all the source packets are necessarily equal in 457 size. However, assuming that each repair packet carries an equal 458 number of bytes as carried by a source packet, the overhead for 459 different FEC protection methods can be computed as follows: 461 o 1-D Non-interleaved FEC Protection: Overhead = 1/L 462 o 1-D Interleaved FEC Protection: Overhead = 1/D 464 o 2-D Parity FEC Protection: Overhead = 1/L + 1/D 466 where L and D are the number of columns and rows in the source block, 467 respectively. 469 2. Requirements Notation 471 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 472 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 473 "OPTIONAL" in this document are to be interpreted as described in BCP 474 14 [RFC2119] [RFC8174] when, and only when, they appear in all 475 capitals, as shown here. 477 3. Definitions and Notations 479 3.1. Definitions 481 This document uses a number of definitions from [RFC6363]. 483 1-D Non-interleaved Row FEC: A protection scheme that operates on 484 consecutive source packets in the source block, able to recover a 485 single lost source packet per row of the source block. 487 1-D Interleaved Column FEC: A protection scheme that operates on 488 interleaved source packets in the source block, able to recover a 489 single lost source packet per column of the source block. 491 2-D FEC: A protection scheme that combines row and column FEC. 493 Source Block: A set of source packets that are protected by a set 494 of 1-D or 2-D FEC repair packets. 496 FEC Block: A source block and its corresponding FEC repair 497 packets. 499 Repair Window: The time that spans a FEC block, which consists of 500 the source packets and the corresponding FEC repair packets. 502 XOR Parity Codes: A FEC code which uses the eXclusive OR (XOR) 503 parity operation to encode a set of source packets to form a FEC 504 repair packet. 506 3.2. Notations 508 L: Number of columns of the source block (length of each row). 510 D: Number of rows of the source block (depth of each column). 512 bitmask: A 15-bit, 46-bit, or 110-bit mask indicating which source 513 packets are protected by a FEC repair packet. If the bit i in the 514 mask is set to 1, the source packet number N + i is protected by 515 this FEC repair packet, where N is the sequence number base 516 indicated in the FEC repair packet. The most significant bit of 517 the mask corresponds to i=0. The least signficant bit of the mask 518 corresponds to i=14 in the 15-bit mask, i=45 in the 46-bit mask, 519 or i=109 in the 110-bit mask. 521 4. Packet Formats 523 This section describes the formats of the source packets and defines 524 the formats of the FEC repair packets. 526 4.1. Source Packets 528 The source packets contain the information that identifies the source 529 block and the position within the source block occupied by the 530 packet. Since the source packets that are carried within an RTP 531 stream already contain unique sequence numbers in their RTP headers 532 [RFC3550], the source packets can be identified in a straightforward 533 manner and there is no need to append additional field(s). The 534 primary advantage of not modifying the source packets in any way is 535 that it provides backward compatibility for the receivers that do not 536 support FEC at all. In multicast scenarios, this backward 537 compatibility becomes quite useful as it allows the non-FEC-capable 538 and FEC-capable receivers to receive and interpret the same source 539 packets sent in the same multicast session. 541 The source packets are transmitted as usual without altering them. 542 They are used along with the FEC repair packets to recover any 543 missing source packets, making this scheme a systematic code. 545 The source packets are full RTP packets with optional CSRC list, RTP 546 header extension, and padding. If any of these optional elements are 547 present in the source RTP packet, and that source packet is lost, 548 they are recovered by the FEC repair operation, which recovers the 549 full source RTP packet including these optional elements. 551 4.2. FEC Repair Packets 553 The FEC repair packets will contain information that identifies the 554 source block they pertain to and the relationship between the 555 contained repair packets and the original source block. For this 556 purpose, the RTP header of the repair packets is used, as well as 557 another header within the RTP payload, called the FEC header, as 558 shown in Figure 9. 560 Note that all the source stream packets that are protected by a 561 particular FEC packet need to be in the same RTP session. 563 +------------------------------+ 565 | IP Header | 567 +------------------------------+ 569 | Transport Header | 571 +------------------------------+ 573 | RTP Header | 575 +------------------------------+ ---+ 577 | FEC Header | | 579 +------------------------------+ | RTP Payload 581 | Repair "Payload" | | 583 +------------------------------+ ---+ 585 Figure 9: Format of FEC repair packets 587 The Repair "Payload", which follows the FEC Header, includes repair 588 of everything following the fixed 12-byte RTP header of each source 589 packet, including any CSRC identifier list and header extensions if 590 present. 592 4.2.1. RTP Header of FEC Repair Packets 594 The RTP header is formatted according to [RFC3550] with some further 595 clarifications listed below: 597 Version (V) 2 bits: This MUST be set to 2 (binary 10), as this 598 specification requires all source RTP packets and all FEC repair 599 packets to use RTP version 2. The reason for this restriction is 600 the first 2 bits of the FEC header contain other information (R 601 and F bits) rather than recovering the RTP version field. 603 Padding (P) bit: Source packets can have optional RTP padding, 604 which can be recovered. FEC repair packets can have optional RTP 605 padding, which is independent of the RTP padding of the source 606 pakcets. 608 Extension (X) bit: Source packets can have optional RTP header 609 extensions, which can be recovered. FEC repair packets can have 610 optional RTP header extensions, which are independent of the RTP 611 header extensions of the source packets. 613 CSRC Count (CC) 4 bits, and CSRC List (CSRC_i) 32 bits each: 614 Source packets can have an optional CSRC list and count, which can 615 be recovered. FEC repair packets MUST use the CSRC list and count 616 to specify the SSRC(s) of the source RTP stream(s) protected by 617 this FEC repair packet. 619 Marker (M) bit: This bit is not used for this payload type, and 620 SHALL be set to 0 by senders, and SHALL be ignored by receivers. 622 Payload Type: The (dynamic) payload type for the FEC repair 623 packets is determined through out-of-band means. Note that this 624 document registers new payload formats for the repair packets 625 (Refer to Section 5 for details). According to [RFC3550], an RTP 626 receiver that cannot recognize a payload type must discard it. 627 This provides backward compatibility. If a non-FEC-capable 628 receiver receives a repair packet, it will not recognize the 629 payload type, and hence, will discard the repair packet. 631 Sequence Number (SN): The sequence number follows the standard 632 definition provided in [RFC3550]. definition. Therefore it must 633 be one higher than the sequence number in the previously 634 transmitted repair packet, and the initial value of the sequence 635 number should be random (i.e. unpredictable). 637 Timestamp (TS): The timestamp SHALL be set to a time corresponding 638 to the repair packet's transmission time. Note that the timestamp 639 value has no use in the actual FEC protection process and is 640 usually useful for jitter calculations. 642 Synchronization Source (SSRC): The SSRC value for each repair 643 stream SHALL be randomly assigned as per the guidelines provided 644 in Section 8 of [RFC3550]. This allows the sender to multiplex 645 the source and repair RTP streams in the same RTP session, or 646 multiplex multiple repair streams in an RTP session. The repair 647 streams' SSRC's CNAME MUST be identical to the CNAME of the source 648 RTP stream(s) that this repair stream protects. In cases when the 649 repair stream covers packets from multiple source RTP streams with 650 different CNAME values, any of these CNAME values MAY be used. 652 In some networks, the RTP Source, which produces the source 653 packets and the FEC Source, which generates the repair packets 654 from the source packets may not be the same host. In such 655 scenarios, using the same CNAME for the source and repair RTP 656 streams means that the RTP Source and the FEC Source will share 657 the same CNAME (for this specific source-repair stream 658 association). A common CNAME may be produced based on an 659 algorithm that is known both to the RTP and FEC Source [RFC7022]. 660 This usage is compliant with [RFC3550]. 662 Note that due to the randomness of the SSRC assignments, there is 663 a possibility of SSRC collision. In such cases, the collisions 664 must be resolved as described in [RFC3550]. 666 4.2.2. FEC Header of FEC Repair Packets 668 The format of the FEC header has 3 variants, depending on the values 669 in the first 2 bits (R and F bits) as shown in Figure 10. Two of 670 these variants are meant to describe different methods for deriving 671 the source data from a source packet for a repair packet. This 672 allows for customizing the FEC method to allow for robustness against 673 different levels of burst errors and random packet losses. The third 674 variant is for a straight retransmission of the source packet. 676 0 1 2 3 678 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 1 680 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 682 |R|F|P|X| CC |M| PT recovery | ...varies depending on R/F... | 684 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 686 | | 688 | ...varies depending on R/F... | 690 | | 692 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 694 : Repair "Payload" follows FEC Header : 696 : : 698 Figure 10: FEC Header 700 The Repair "Payload", which follows the FEC Header, includes repair 701 of everything following the fixed 12-byte RTP header of each source 702 packet, including any CSRC identifier list and header extensions if 703 present. 705 +---+---+----------------------------------------------------------+ 707 | R | F | FEC Header variant | 709 +---+---+----------------------------------------------------------+ 711 | 0 | 0 | Flexible FEC Mask fields indicate source packets | 713 | 0 | 1 | Fixed FEC L/D (cols/rows) fields indicate source packets | 715 | 1 | 0 | Retransmission of a single source packet | 717 | 1 | 1 | Invalid, MUST NOT send, MUST ignore if received | 719 +---+---+----------------------------------------------------------+ 721 Figure 11: R and F bit values for FEC Header variants 723 The first variant, when R=0 and F=0, has a mask to signal protected 724 source packets, as shown in Figure 12. 726 The second variant, when R=0 and F=1, has a number of columns (L) and 727 rows (D) to signal protected source packets, as shown in Figure 13. 729 The final variant, when R=1 and F=0, is a retransmission format as 730 shown in Figure 15. 732 No variant uses R=1 and F=1, which is invalid, and MUST NOT be sent 733 by senders, and MUST be ignored by receivers. 735 The FEC header for all variants consists of the following common 736 fields: 738 o The R bit MUST be set to 1 to indicate a retransmission packet, 739 and MUST be set to 0 for FEC repair packets. 741 o The F bit indicates the type of FEC repair packets, as shown in 742 Figure 11, when the R bit is 0. The F bit MUST be set to 0 when 743 the R bit is 1 for retransmission packets. 745 o The P, X, CC, M and PT recovery fields are used to determine the 746 corresponding fields of the recovered packets. 748 4.2.2.1. FEC Header with Flexible Mask 750 When R=0 and F=0, the FEC Header includes flexible mask fields. 752 0 1 2 3 754 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 1 756 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 758 |0|0|P|X| CC |M| PT recovery | length recovery | 760 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 762 | TS recovery | 764 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 766 | SN base_i |k| Mask [0-14] | 768 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 770 |k| Mask [15-45] (optional) | 772 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 774 | Mask [46-109] (optional) | 776 | | 778 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 780 | ... next SN base and Mask for CSRC_i in CSRC list ... | 782 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 784 : Repair "Payload" follows FEC Header : 786 : : 788 Figure 12: FEC Header for F=0 790 o The Length recovery (16 bits) field is used to determine the 791 length of the recovered packets. This length includes all octets 792 following the fixed 12-byte RTP header of source packets, 793 including CSRC list and optional header extension(s) if present. 794 It excludes the fixed 12-byte RTP header of source packets. 796 o The TS recovery (32 bits) field is used to determine the timestamp 797 of the recovered packets. 799 o The CSRC_i (32 bits) field in the RTP Header (not FEC Header) 800 describes the SSRC of the source packets protected by this 801 particular FEC packet. If a FEC packet protects multiple SSRCs 802 (indicated by the CSRC Count > 1 in the RTP Header), there will be 803 multiple blocks of data containing the SN base and Mask fields. 805 o The SN base_i (16 bits) field indicates the lowest sequence 806 number, taking wrap around into account, of the source packets for 807 a particular SSRC (indicated in CSRC_i) protected by this repair 808 packet. 810 o The Mask fields indicate a bitmask of which source packets are 811 protected by this FEC repair packet, where bit j of the mask set 812 to 1 indicates that the source packet with sequence number (SN 813 base_i + j) is protected by this FEC repair packet, where j=0 is 814 the most significant bit in the mask. 816 o The k-bit in the bitmasks indicates if the mask is 15, 46, or 110 817 bits. k=1 denotes that another mask follows, and k=0 denotes that 818 it is the last block of mask. 820 o The Repair "Payload", which follows the FEC Header, includes 821 repair of everything following the fixed 12-byte RTP header of 822 each source packet, including any CSRC identifier list and header 823 extensions if present. 825 4.2.2.2. FEC Header with Fixed L Columns and D Rows 827 When R=0 and F=1, the FEC Header includes L and D fields for fixed 828 columns and rows. The other fields are the same as the prior 829 section. As in the previous section, the CSRC_i (32 bits) field in 830 the RTP Header (not FEC Header) describes the SSRC of the source 831 packets protected by this particular FEC packet. If there are 832 multiple SSRC's protected by the FEC packet, then there will be 833 multiple blocks of data containing an SN base along with L and D 834 fields. 836 0 1 2 3 838 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 1 840 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 842 |0|1|P|X| CC |M| PT recovery | length recovery | 844 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 846 | TS recovery | 848 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 850 | SN base_i | L (columns) | D (rows) | 852 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 854 | ... next SN base and L/D for CSRC_i in CSRC list ... | 856 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 858 : Repair "Payload" follows FEC Header : 860 : : 862 Figure 13: FEC Header for F=1 864 Consequently, the following conditions occur for L and D values: 866 If L=0, D=0, use the optional payload format parameters for L and D. 868 If L>0, D=0, indicates Row FEC, and no column FEC will follow. 870 Hence, FEC = SN, SN+1, SN+2, ... , SN+(L-1), SN+L. 872 If L>0, D=1, indicates Row FEC, and column FEC will follow. 874 Hence, FEC = SN, SN+1, SN+2, ... , SN+(L-1), SN+L will be 876 produced for each row. 878 Then FEC = SN, SN+L, SN+2L, ..., SN+(D-1)L will be produced 880 for each column. 882 After all row FEC's have been sent, then the column FEC's 884 will be sent. 886 If L>0, D>1, indicates column FEC of every L packet 888 in a group of D packets starting at SN base. 890 Hence, FEC = SN+(Lx0), SN+(Lx1), ... , SN+(LxD). 892 Figure 14: Interpreting the L and D field values 894 It should be noted that the flexible mask-based approach may be 895 inefficient for protecting a large number of source packets, or 896 impossible to signal if larger than the largest mask size. In such 897 cases, the fixed columns and rows variant may be more useful. 899 4.2.2.3. FEC Header for Retransmissions 901 When R=1 and F=0, the FEC packet is a retransmission of a single 902 source packet. Note that the layout of this retransmission packet is 903 different from other FEC repair packets. The sequence number (SN 904 base_i) replaces the length recovery in the FEC header, since the 905 length is already known for a single packet. There are no L, D or 906 Mask fields, since only a single packet is retransmitted, identified 907 by the sequence number in the FEC header. The source packet SSRC is 908 included in the FEC header for retransmissions, not in the RTP header 909 CSRC list as in the FEC header variants with R=0. When performing 910 retransmissions, a single repair packet stream (SSRC) MAY be used for 911 retransmitting packets from multiple source packet streams (SSRCs), 912 as well as transmitting FEC repair packets that protect multiple 913 source packet streams (SSRCs). 915 This FEC header layout is identical to the source RTP (version 2) 916 packet, starting with its RTP header, where the retransmission 917 "payload" is everything following the fixed 12-byte RTP header of the 918 source packet, including CSRC list and extensions if present. 919 Therefore, the only operation needed for sending retransmissions is 920 to prepend a new RTP header to the source packet. 922 0 1 2 3 924 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 1 926 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 928 |1|0| P|X| CC |M| Payload Type| Sequence Number | 930 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 932 | Timestamp | 934 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 936 | SSRC | 938 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 940 : Retransmission "Payload" follows FEC Header : 942 : : 944 Figure 15: FEC Header for Retransmission 946 5. Payload Format Parameters 948 This section provides the media subtype registration for the non- 949 interleaved and interleaved parity FEC. The parameters that are 950 required to configure the FEC encoding and decoding operations are 951 also defined in this section. If no specific FEC code is specified 952 in the subtype, then the FEC code defaults to the parity code defined 953 in this specification. 955 5.1. Media Type Registration - Parity Codes 957 This registration is done using the template defined in [RFC6838] and 958 following the guidance provided in [RFC4855] along with [RFC4856]. 960 Note to the RFC Editor: In the following sections, please replace 961 "XXXX" with the number of this document prior to publication as an 962 RFC. 964 5.1.1. Registration of audio/flexfec 966 Type name: audio 968 Subtype name: flexfec 970 Required parameters: 972 o rate: The RTP timestamp (clock) rate. The rate SHALL be larger 973 than 1000 Hz to provide sufficient resolution to RTCP operations. 974 However, it is RECOMMENDED to select the rate that matches the 975 rate of the protected source RTP stream. 977 o repair-window: The time that spans the source packets and the 978 corresponding repair packets. The size of the repair window is 979 specified in microseconds. 981 Optional parameters: 983 o L: indicates the number of columns of the source block that are 984 protected by this FEC block and it applies to all the source 985 SSRCs. L is a positive integer. 987 o D: indicates the number of rows of the source block that are 988 protected by this FEC block and it applies to all the source 989 SSRCs. D is a positive integer. 991 o ToP: indicates the type of protection applied by the sender: 0 for 992 1-D interleaved FEC protection, 1 for 1-D non-interleaved FEC 993 protection, 2 for 2-D parity FEC protection, and 3 for 994 retransmission. There can only be one value listed for ToP. The 995 absence of the ToP field means that all protection types are 996 allowed. 998 Note that both L and D in the optional parameters should follow the 999 value pairings stated in Section 4.2.2.2 if included. 1001 Encoding considerations: This media type is framed (See Section 4.8 1002 in the template document [RFC6838]) and contains binary data. 1004 Security considerations: See Section 9 of [RFCXXXX]. 1006 Interoperability considerations: None. 1008 Published specification: [RFCXXXX]. 1010 Applications that use this media type: Multimedia applications that 1011 want to improve resiliency against packet loss by sending redundant 1012 data in addition to the source media. 1014 Fragment identifier considerations: None. 1016 Additional information: None. 1018 Person & email address to contact for further information: IESG 1019 and IETF Audio/Video Transport Payloads Working 1020 Group (or it's successor as delegated by the IESG). 1022 Intended usage: COMMON. 1024 Restriction on usage: This media type depends on RTP framing, and 1025 hence, is only defined for transport via RTP [RFC3550]. 1027 Author: IESG . 1029 Change controller: IETF Audio/Video Transport Payloads Working Group 1030 delegated from the IESG (or it's successor as delegated by the IESG). 1032 5.1.2. Registration of video/flexfec 1034 Type name: video 1036 Subtype name: flexfec 1038 Required parameters: 1040 o rate: The RTP timestamp (clock) rate. The rate SHALL be larger 1041 than 1000 Hz to provide sufficient resolution to RTCP operations. 1043 However, it is RECOMMENDED to select the rate that matches the 1044 rate of the protected source RTP stream. 1046 o repair-window: The time that spans the source packets and the 1047 corresponding repair packets. The size of the repair window is 1048 specified in microseconds. 1050 Optional parameters: 1052 o L: indicates the number of columns of the source block that are 1053 protected by this FEC block and it applies to all the source 1054 SSRCs. L is a positive integer. 1056 o D: indicates the number of rows of the source block that are 1057 protected by this FEC block and it applies to all the source 1058 SSRCs. D is a positive integer. 1060 o ToP: indicates the type of protection applied by the sender: 0 for 1061 1-D interleaved FEC protection, 1 for 1-D non-interleaved FEC 1062 protection, 2 for 2-D parity FEC protection, and 3 for 1063 retransmission. There can only be one value listed for ToP. The 1064 absence of the ToP field means that all protection types are 1065 allowed. 1067 Note that both L and D in the optional parameters should follow the 1068 value pairings stated in Section 4.2.2.2 if included. 1070 Encoding considerations: This media type is framed (See Section 4.8 1071 in the template document [RFC6838]) and contains binary data. 1073 Security considerations: See Section 9 of [RFCXXXX]. 1075 Interoperability considerations: None. 1077 Published specification: [RFCXXXX]. 1079 Applications that use this media type: Multimedia applications that 1080 want to improve resiliency against packet loss by sending redundant 1081 data in addition to the source media. 1083 Fragment identifier considerations: None. 1085 Additional information: None. 1087 Person & email address to contact for further information: IESG 1088 and IETF Audio/Video Transport Payloads Working Group 1089 (or it's successor as delegated by the IESG). 1091 Intended usage: COMMON. 1093 Restriction on usage: This media type depends on RTP framing, and 1094 hence, is only defined for transport via RTP [RFC3550]. 1096 Author: IESG . 1098 Change controller: IETF Audio/Video Transport Payloads Working Group 1099 delegated from the IESG (or it's successor as delegated by the IESG). 1101 5.1.3. Registration of text/flexfec 1103 Type name: text 1105 Subtype name: flexfec 1107 Required parameters: 1109 o rate: The RTP timestamp (clock) rate. The rate SHALL be larger 1110 than 1000 Hz to provide sufficient resolution to RTCP operations. 1111 However, it is RECOMMENDED to select the rate that matches the 1112 rate of the protected source RTP stream. 1114 o repair-window: The time that spans the source packets and the 1115 corresponding repair packets. The size of the repair window is 1116 specified in microseconds. 1118 Optional parameters: 1120 o L: indicates the number of columns of the source block that are 1121 protected by this FEC block and it applies to all the source 1122 SSRCs. L is a positive integer. 1124 o D: indicates the number of rows of the source block that are 1125 protected by this FEC block and it applies to all the source 1126 SSRCs. D is a positive integer. 1128 o ToP: indicates the type of protection applied by the sender: 0 for 1129 1-D interleaved FEC protection, 1 for 1-D non-interleaved FEC 1130 protection, 2 for 2-D parity FEC protection, and 3 for 1131 retransmission. There can only be one value listed for ToP. The 1132 absence of the ToP field means that all protection types are 1133 allowed. 1135 Note that both L and D in the optional parameters should follow the 1136 value pairings stated in Section 4.2.2.2 if included. 1138 Encoding considerations: This media type is framed (See Section 4.8 1139 in the template document [RFC6838]) and contains binary data. 1141 Security considerations: See Section 9 of [RFCXXXX]. 1143 Interoperability considerations: None. 1145 Published specification: [RFCXXXX]. 1147 Applications that use this media type: Multimedia applications that 1148 want to improve resiliency against packet loss by sending redundant 1149 data in addition to the source media. 1151 Fragment identifier considerations: None. 1153 Additional information: None. 1155 Person & email address to contact for further information: IESG 1156 and IETF Audio/Video Transport Payloads Working 1157 Group (or it's successor as delegated by the IESG). 1159 Intended usage: COMMON. 1161 Restriction on usage: This media type depends on RTP framing, and 1162 hence, is only defined for transport via RTP [RFC3550]. 1164 Author: IESG . 1166 Change controller: IETF Audio/Video Transport Payloads Working Group 1167 delegated from the IESG (or it's successor as delegated by the IESG). 1169 5.1.4. Registration of application/flexfec 1171 Type name: application 1173 Subtype name: flexfec 1175 Required parameters: 1177 o rate: The RTP timestamp (clock) rate. The rate SHALL be larger 1178 than 1000 Hz to provide sufficient resolution to RTCP operations. 1179 However, it is RECOMMENDED to select the rate that matches the 1180 rate of the protected source RTP stream. 1182 o repair-window: The time that spans the source packets and the 1183 corresponding repair packets. The size of the repair window is 1184 specified in microseconds. 1186 Optional parameters: 1188 o L: indicates the number of columns of the source block that are 1189 protected by this FEC block and it applies to all the source 1190 SSRCs. L is a positive integer. 1192 o D: indicates the number of rows of the source block that are 1193 protected by this FEC block and it applies to all the source 1194 SSRCs. D is a positive integer. 1196 o ToP: indicates the type of protection applied by the sender: 0 for 1197 1-D interleaved FEC protection, 1 for 1-D non-interleaved FEC 1198 protection, 2 for 2-D parity FEC protection, and 3 for 1199 retransmission. There can only be one value listed for ToP. The 1200 absence of the ToP field means that all protection types are 1201 allowed. 1203 Note that both L and D in the optional parameters should follow the 1204 value pairings stated in Section 4.2.2.2 if included. 1206 Encoding considerations: This media type is framed (See Section 4.8 1207 in the template document [RFC6838]) and contains binary data. 1209 Security considerations: See Section 9 of [RFCXXXX]. 1211 Interoperability considerations: None. 1213 Published specification: [RFCXXXX]. 1215 Applications that use this media type: Multimedia applications that 1216 want to improve resiliency against packet loss by sending redundant 1217 data in addition to the source media. 1219 Fragment identifier considerations: None. 1221 Additional information: None. 1223 Person & email address to contact for further information: IESG 1224 and IETF Audio/Video Transport Payloads Working Group 1225 (or it's successor as delegated by the IESG). 1227 Intended usage: COMMON. 1229 Restriction on usage: This media type depends on RTP framing, and 1230 hence, is only defined for transport via RTP [RFC3550]. 1232 Author: IESG . 1234 Change controller: IETF Audio/Video Transport Payloads Working Group 1235 delegated from the IESG (or it's successor as delegated by the IESG). 1237 5.2. Mapping to SDP Parameters 1239 Applications that use the RTP transport commonly use Session 1240 Description Protocol (SDP) [RFC4566] to describe their RTP sessions. 1241 The information that is used to specify the media types in an RTP 1242 session has specific mappings to the fields in an SDP description. 1243 This section provides these mappings for the media subtypes 1244 registered by this document. Note that if an application does not 1245 use SDP to describe the RTP sessions, an appropriate mapping must be 1246 defined and used to specify the media types and their parameters for 1247 the control/description protocol employed by the application. 1249 The mapping of the media type specification for "non-interleaved- 1250 parityfec" and "interleaved-parityfec" and their parameters in SDP is 1251 as follows: 1253 o The media type (e.g., "application") goes into the "m=" line as 1254 the media name. 1256 o The media subtype goes into the "a=rtpmap" line as the encoding 1257 name. The RTP clock rate parameter ("rate") also goes into the 1258 "a=rtpmap" line as the clock rate. 1260 o The remaining required payload-format-specific parameters go into 1261 the "a=fmtp" line by copying them directly from the media type 1262 string as a semicolon-separated list of parameter=value pairs. 1264 SDP examples are provided in Section 7.1. 1266 5.2.1. Offer-Answer Model Considerations 1268 When offering 1-D interleaved parity FEC over RTP using SDP in an 1269 Offer/Answer model [RFC3264], the following considerations apply: 1271 o Each combination of the L and D parameters produces a different 1272 FEC data and is not compatible with any other combination. A 1273 sender application may desire to offer multiple offers with 1274 different sets of L and D values as long as the parameter values 1275 are valid. The receiver SHOULD choose the offer that has a 1276 sufficient amount of interleaving. If multiple such offers exist, 1277 the receiver may choose the offer that has the lowest overhead or 1278 the one that requires the smallest amount of buffering. The 1279 selection depends on the application requirements. 1281 o The value for the repair-window parameter depends on the L and D 1282 values and cannot be chosen arbitrarily. More specifically, L and 1283 D values determine the lower limit for the repair-window size. 1284 The upper limit of the repair-window size does not depend on the L 1285 and D values. 1287 o Although combinations with the same L and D values but with 1288 different repair-window sizes produce the same FEC data, such 1289 combinations are still considered different offers. The size of 1290 the repair-window is related to the maximum delay between the 1291 transmission of a source packet and the associated repair packet. 1292 This directly impacts the buffering requirement on the receiver 1293 side and the receiver must consider this when choosing an offer. 1295 o Any unknown option in the offer MUST be ignored and deleted from 1296 the answer. If FEC is not desired by the receiver, it can be 1297 deleted from the answer. 1299 5.2.2. Declarative Considerations 1301 In declarative usage, like SDP in the Real-time Streaming Protocol 1302 (RTSP, for RTSP 1.0 see [RFC2326] and for RTSP 2.0 see [RFC7826]) or 1303 the Session Announcement Protocol (SAP) [RFC2974], the following 1304 considerations apply: 1306 o The payload format configuration parameters are all declarative 1307 and a participant MUST use the configuration that is provided for 1308 the session. 1310 o More than one configuration may be provided (if desired) by 1311 declaring multiple RTP payload types. In that case, the receivers 1312 should choose the repair stream that is best for them. 1314 6. Protection and Recovery Procedures - Parity Codes 1316 This section provides a complete specification of the 1-D and 2-D 1317 parity codes and their RTP payload formats. It does not apply to the 1318 single packet retransmission format (R=1 in the FEC Header). 1320 6.1. Overview 1322 The following sections specify the steps involved in generating the 1323 repair packets and reconstructing the missing source packets from the 1324 repair packets. 1326 6.2. Repair Packet Construction 1328 The RTP Header of a repair packet is formed based on the guidelines 1329 given in Section 4.2. 1331 The FEC Header and Repair "Payload" of repair packets are formed by 1332 applying the XOR operation on the bit strings that are generated from 1333 the individual source packets protected by this particular repair 1334 packet. The set of the source packets that are associated with a 1335 given repair packet can be computed by the formula given in 1336 Section 6.3.1. 1338 The bit string is formed for each source packet by concatenating the 1339 following fields together in the order specified: 1341 o The first 16 bits of the RTP header (16 bits). 1343 o Unsigned network-ordered 16-bit representation of the source 1344 packet length in bytes minus 12 (for the fixed RTP header), i.e., 1345 the sum of the lengths of all the following if present: the CSRC 1346 list, extension header, RTP payload and RTP padding (16 bits). 1348 o The timestamp of the RTP header (32 bits). 1350 o All octets after the fixed 12-byte RTP header. (Note the SSRC 1351 field is skipped.) 1353 The FEC bit string is generated by applying the parity operation on 1354 the bit strings produced from the source packets. The FEC header is 1355 generated from the FEC bit string as follows: 1357 o The first (most significant) 2 bits in the FEC bit string, which 1358 contain the RTP version field, are skipped. The R and F bits in 1359 the FEC header are set to the appropriate value, i.e., it depends 1360 on the chosen format variant. As a consequence of overwriting the 1361 RTP version field with the R and F bits, this payload format only 1362 supports RTP version 2. 1364 o The next bit in the FEC bit string is written into the P recovery 1365 bit in the FEC header. 1367 o The next bit in the FEC bit string is written into the X recovery 1368 bit in the FEC header. 1370 o The next 4 bits of the FEC bit string are written into the CC 1371 recovery field in the FEC header. 1373 o The next bit is written into the M recovery bit in the FEC header. 1375 o The next 7 bits of the FEC bit string are written into the PT 1376 recovery field in the FEC header. 1378 o The next 16 bits are written into the length recovery field in the 1379 FEC header. 1381 o The next 32 bits of the FEC bit string are written into the TS 1382 recovery field in the FEC header. 1384 o The lowest Sequence Number of the source packets protected by this 1385 repair packet is written into the Sequence Number Base field in 1386 the FEC header. This needs to be repeated for each SSRC that has 1387 packets included in the source block. 1389 o Depending on the chosen FEC header variant, the mask(s) are set 1390 when F=0, or the L and D values are set when F=1. This needs to 1391 be repeated for each SSRC that has packets included in the source 1392 block. 1394 o The rest of the FEC bit string, which contains everything after 1395 the fixed 12-byte RTP header of the source packet, is written into 1396 the Repair "Payload" following the FEC header, where "Payload" 1397 refers to everything after the fixed 12-byte RTP header, including 1398 extensions, CSRC list, true payloads, and padding. 1400 If the lengths of the source packets are not equal, each shorter 1401 packet MUST be padded to the length of the longest packet by adding 1402 octet 0's at the end. 1404 Due to this possible padding and mandatory FEC header, a repair 1405 packet has a larger size than the source packets it protects. This 1406 may cause problems if the resulting repair packet size exceeds the 1407 Maximum Transmission Unit (MTU) size of the path over which the 1408 repair stream is sent. 1410 6.3. Source Packet Reconstruction 1412 This section describes the recovery procedures that are required to 1413 reconstruct the missing source packets. The recovery process has two 1414 steps. In the first step, the FEC decoder determines which source 1415 and repair packets should be used in order to recover a missing 1416 packet. In the second step, the decoder recovers the missing packet, 1417 which consists of an RTP header and RTP payload. 1419 The following describes the RECOMMENDED algorithms for the first and 1420 second steps. Based on the implementation, different algorithms MAY 1421 be adopted. However, the end result MUST be identical to the one 1422 produced by the algorithms described below. 1424 Note that the same algorithms are used by the 1-D parity codes, 1425 regardless of whether the FEC protection is applied over a column or 1426 a row. The 2-D parity codes, on the other hand, usually require 1427 multiple iterations of the procedures described here. This iterative 1428 decoding algorithm is further explained in Section 6.3.4. 1430 6.3.1. Associating the Source and Repair Packets 1432 Before associating source and repair packets, the receiver must know 1433 in which RTP sessions the source and repair respectively are being 1434 sent. After this is established by the reciever the first step is 1435 associating the source and repair packets. This association can be 1436 via flexible bitmasks, or fixed L and D offsets which can be in the 1437 FEC header or signaled in SDP in optional payload format parameters 1438 when L=D=0 in the FEC header. 1440 6.3.1.1. Using Bitmasks 1442 To use flexible bitmasks, the first two FEC header bits MUST have R=0 1443 and F=0. A 15-bit, 46-bit, or 110-bit mask indicates which source 1444 packets are protected by a FEC repair packet. If the bit i in the 1445 mask is set to 1, the source packet number N + i is protected by this 1446 FEC repair packet, where N is the sequence number base indicated in 1447 the FEC header. The most significant bit of the mask corresponds to 1448 i=0. The least signficant bit of the mask corresponds to i=14 in the 1449 15-bit mask, i=45 in the 46-bit mask, or i=109 in the 110-bit mask. 1451 The bitmasks are able to represent arbitrary protection patterns, for 1452 example, 1-D interleaved, 1-D non-interleaved, 2-D. 1454 6.3.1.2. Using L and D Offsets 1456 Denote the set of the source packets associated with repair packet p* 1457 by set T(p*). Note that in a source block whose size is L columns by 1458 D rows, set T includes D source packets plus one repair packet for 1459 the FEC protection applied over a column, and L source packets plus 1460 one repair packet for the FEC protection applied over a row. Recall 1461 that 1-D interleaved and non-interleaved FEC protection can fully 1462 recover the missing information if there is only one source packet 1463 missing per column or row in set T. If there are more than one 1464 source packets missing per column or row in set T, 1-D FEC protection 1465 may fail to recover all the missing information. 1467 When value of L is non-zero, the 8-bit fields indicate the offset of 1468 packets protected by an interleaved (D>0) or non-interleaved (D=0) 1469 FEC packet. Using a combination of interleaved and non-interleaved 1470 FEC repair packets can form 2-D protection patterns. 1472 Mathematically, for any received repair packet, p*, the sequence 1473 numbers of the source packets that are protected by this repair 1474 packet are determined as follows, where p*_snb is the sequence number 1475 base in the FEC header: 1477 When D = 0: 1479 p*_snb, p*_snb+1,..., p*_snb+L 1481 When D > 0: 1483 p*_snb, p*_snb+(Lx1), p*_snb+(Lx2),..., p*_snb+(LxD) 1485 6.3.1.3. Signaled in SDP 1487 If the endpoint relies entirely on out-of-band signaling (R=0, F=1, 1488 L=0, D=0 in the FEC header), then this information may be inferred 1489 from the media type parameters specified in the SDP description. 1490 Furthermore, the payload type field in the RTP header assists the 1491 receiver to distinguish an interleaved or non-interleaved FEC packet. 1493 Mathematically, for any received repair packet, p*, the sequence 1494 numbers of the source packets that are protected by this repair 1495 packet are determined as follows: 1497 p*_snb + i * X_1 (modulo 65536) 1499 where p*_snb denotes the value in the SN base field of p*'s FEC 1500 header, X_1 is set to L and 1 for the interleaved and non-interleaved 1501 FEC repair packets, respectively, and 1503 0 <= i < X_2 1505 where X_2 is set to D and L for the interleaved and non-interleaved 1506 FEC repair packets, respectively. 1508 6.3.2. Recovering the RTP Header 1510 For a given set T, the procedure for the recovery of the RTP header 1511 of the missing packet, whose sequence number is denoted by SEQNUM, is 1512 as follows: 1514 1. For each of the source packets that are successfully received in 1515 T, compute the 80-bit string by concatenating the first 64 bits 1516 of their RTP header and the unsigned network-ordered 16-bit 1517 representation of their length in bytes minus 12. 1519 2. For the repair packet in T, compute the FEC bit string from the 1520 first 80 bits of the FEC header. 1522 3. Calculate the recovered bit string as the XOR of the bit strings 1523 generated from all source packets in T and the FEC bit string 1524 generated from the repair packet in T. 1526 4. Create a new packet with the standard 12-byte RTP header and no 1527 payload. 1529 5. Set the version of the new packet to 2. Skip the first 2 bits 1530 in the recovered bit string. 1532 6. Set the Padding bit in the new packet to the next bit in the 1533 recovered bit string. 1535 7. Set the Extension bit in the new packet to the next bit in the 1536 recovered bit string. 1538 8. Set the CC field to the next 4 bits in the recovered bit string. 1540 9. Set the Marker bit in the new packet to the next bit in the 1541 recovered bit string. 1543 10. Set the Payload type in the new packet to the next 7 bits in the 1544 recovered bit string. 1546 11. Set the SN field in the new packet to SEQNUM. Skip the next 16 1547 bits in the recovered bit string. 1549 12. Set the TS field in the new packet to the next 32 bits in the 1550 recovered bit string. 1552 13. Take the next 16 bits of the recovered bit string and set the 1553 new variable Y to whatever unsigned integer this represents 1554 (assuming network order). Convert Y to host order. Y 1555 represents the length of the new packet in bytes minus 12 (for 1556 the fixed RTP header), i.e., the sum of the lengths of all the 1557 following if present: the CSRC list, header extension, RTP 1558 payload and RTP padding. 1560 14. Set the SSRC of the new packet to the SSRC of the missing source 1561 RTP stream. 1563 This procedure recovers the header of an RTP packet up to (and 1564 including) the SSRC field. 1566 6.3.3. Recovering the RTP Payload 1568 Following the recovery of the RTP header, the procedure for the 1569 recovery of the RTP "payload" is as follows, where "payload" refers 1570 to everything following the fixed 12-byte RTP header, including 1571 extensions, CSRC list, true payload and padding. 1573 1. Append Y bytes to the new packet. 1575 2. For each of the source packets that are successfully received in 1576 T, compute the bit string from the Y octets of data starting with 1577 the 13th octet of the packet. If any of the bit strings 1578 generated from the source packets has a length shorter than Y, 1579 pad them to that length. The zero-padding octets MUST be added 1580 at the end of the bit string. Note that the information of the 1581 first 8 octets are protected by the FEC header. 1583 3. For the repair packet in T, compute the FEC bit string from the 1584 repair packet payload, i.e., the Y octets of data following the 1585 FEC header. Note that the FEC header may be different sizes 1586 depending on the variant and bitmask size. 1588 4. Calculate the recovered bit string as the XOR of the bit strings 1589 generated from all source packets in T and the FEC bit string 1590 generated from the repair packet in T. 1592 5. Append the recovered bit string (Y octets) to the new packet 1593 generated in Section 6.3.2. 1595 6.3.4. Iterative Decoding Algorithm for the 2-D Parity FEC Protection 1597 In 2-D parity FEC protection, the sender generates both non- 1598 interleaved and interleaved FEC repair packets to combat with the 1599 mixed loss patterns (random and bursty). At the receiver side, these 1600 FEC packets are used iteratively to overcome the shortcomings of the 1601 1-D non-interleaved/interleaved FEC protection and improve the 1602 chances of full error recovery. 1604 The iterative decoding algorithm runs as follows: 1606 1. Set num_recovered_until_this_iteration to zero 1608 2. Set num_recovered_so_far to zero 1610 3. Recover as many source packets as possible by using the non- 1611 interleaved FEC repair packets as outlined in Section 6.3.2 and 1612 Section 6.3.3, and increase the value of num_recovered_so_far by 1613 the number of recovered source packets. 1615 4. Recover as many source packets as possible by using the 1616 interleaved FEC repair packets as outlined in Section 6.3.2 and 1617 Section 6.3.3, and increase the value of num_recovered_so_far by 1618 the number of recovered source packets. 1620 5. If num_recovered_so_far > num_recovered_until_this_iteration 1621 ---num_recovered_until_this_iteration = num_recovered_so_far 1622 ---Go to step 3 1623 Else 1624 ---Terminate 1626 The algorithm terminates either when all missing source packets are 1627 fully recovered or when there are still remaining missing source 1628 packets but the FEC repair packets are not able to recover any more 1629 source packets. For the example scenarios when the 2-D parity FEC 1630 protection fails full recovery, refer to Section 1.1.4. Upon 1631 termination, variable num_recovered_so_far has a value equal to the 1632 total number of recovered source packets. 1634 Example: 1636 Suppose that the receiver experienced the loss pattern sketched in 1637 Figure 16. 1639 +---+ +---+ +===+ 1641 X X | 3 | | 4 | |R_1| 1643 +---+ +---+ +===+ 1645 +---+ +---+ +---+ +---+ +===+ 1647 | 5 | | 6 | | 7 | | 8 | |R_2| 1649 +---+ +---+ +---+ +---+ +===+ 1651 +---+ +---+ +===+ 1653 | 9 | X X | 12| |R_3| 1655 +---+ +---+ +===+ 1657 +===+ +===+ +===+ +===+ 1659 |C_1| |C_2| |C_3| |C_4| 1661 +===+ +===+ +===+ +===+ 1663 Figure 16: Example loss pattern for the iterative decoding algorithm 1665 The receiver executes the iterative decoding algorithm and recovers 1666 source packets #1 and #11 in the first iteration. The resulting 1667 pattern is sketched in Figure 17. 1669 +---+ +---+ +---+ +===+ 1671 | 1 | X | 3 | | 4 | |R_1| 1673 +---+ +---+ +---+ +===+ 1675 +---+ +---+ +---+ +---+ +===+ 1677 | 5 | | 6 | | 7 | | 8 | |R_2| 1679 +---+ +---+ +---+ +---+ +===+ 1681 +---+ +---+ +---+ +===+ 1683 | 9 | X | 11| | 12| |R_3| 1685 +---+ +---+ +---+ +===+ 1687 +===+ +===+ +===+ +===+ 1689 |C_1| |C_2| |C_3| |C_4| 1691 +===+ +===+ +===+ +===+ 1693 Figure 17: The resulting pattern after the first iteration 1695 Since the if condition holds true, the receiver runs a new iteration. 1696 In the second iteration, source packets #2 and #10 are recovered, 1697 resulting in a full recovery as sketched in Figure 18. 1699 +---+ +---+ +---+ +---+ +===+ 1701 | 1 | | 2 | | 3 | | 4 | |R_1| 1703 +---+ +---+ +---+ +---+ +===+ 1705 +---+ +---+ +---+ +---+ +===+ 1707 | 5 | | 6 | | 7 | | 8 | |R_2| 1709 +---+ +---+ +---+ +---+ +===+ 1711 +---+ +---+ +---+ +---+ +===+ 1713 | 9 | | 10| | 11| | 12| |R_3| 1715 +---+ +---+ +---+ +---+ +===+ 1717 +===+ +===+ +===+ +===+ 1719 |C_1| |C_2| |C_3| |C_4| 1721 +===+ +===+ +===+ +===+ 1723 Figure 18: The resulting pattern after the second iteration 1725 7. Signaling Requirements 1727 Out-of-band signaling should be designed to enable the receiver to 1728 identify the RTP streams associated with source packets and repair 1729 packets, respectively. At a minimum, the signaling must be designed 1730 to allow the receiver to 1732 o Determine whether one or more source RTP streams will be sent. 1734 o Determine whether one or more repair RTP streams will be sent. 1736 o Associate the appropriate SSRC's to both source and repair 1737 streams. 1739 o Clearly identify which SSRC's are associated with each source 1740 block. 1742 o Clearly identify which repair packets correspond to which source 1743 blocks. 1745 o Make use of repair packets to recover source data associated with 1746 specific SSRC's. 1748 This section provides several Sesssion Description Protocol (SDP) 1749 examples to demonstrate how these requirements can be met. 1751 7.1. SDP Examples 1753 This section provides two SDP [RFC4566] examples. The examples use 1754 the FEC grouping semantics defined in [RFC5956]. 1756 7.1.1. Example SDP for Flexible FEC Protection with in-band SSRC 1757 mapping 1759 In this example, we have one source video stream and one FEC repair 1760 stream. The source and repair streams are multiplexed on different 1761 SSRCs. The repair window is set to 200 ms. 1763 v=0 1765 o=mo 1122334455 1122334466 IN IP4 fec.example.com 1767 s=FlexFEC minimal SDP signalling Example 1769 t=0 0 1771 m=video 30000 RTP/AVP 96 98 1773 c=IN IP4 233.252.0.1/127 1775 a=rtpmap:96 VP8/90000 1777 a=rtpmap:98 flexfec/90000 1779 a=fmtp:98; repair-window=200ms 1781 7.1.2. Example SDP for Flexible FEC Protection with explicit signalling 1782 in the SDP 1784 This example shows one source video stream (ssrc:1234) and one FEC 1785 repair streams (ssrc:2345). One FEC group is formed with the 1786 "a=ssrc-group:FEC-FR 1234 2345" line. The source and repair streams 1787 are multiplexed on different SSRCs. The repair window is set to 200 1788 ms. 1790 v=0 1792 o=ali 1122334455 1122334466 IN IP4 fec.example.com 1794 s=2-D Parity FEC with no in band signalling Example 1796 t=0 0 1798 m=video 30000 RTP/AVP 100 110 1800 c=IN IP4 192.0.2.0/24 1802 a=rtpmap:100 MP2T/90000 1804 a=rtpmap:110 flexfec/90000 1806 a=fmtp:110 L:5; D:10; ToP:2; repair-window:200000 1808 a=ssrc:1234 1810 a=ssrc:2345 1812 a=ssrc-group:FEC-FR 1234 2345 1814 7.2. On the Use of the RTP Stream Identifier Source Description 1816 The RTP Stream Identifier Source Description [I-D.ietf-avtext-rid] is 1817 a format that can be used to identify a single RTP source stream 1818 along with an associated repair stream. However, this specification 1819 already defines a method of source and repair stream identification 1820 that can enable protection of multiple source streams with a single 1821 repair stream. Therefore the RTP Stream Idenfifer Source Description 1822 SHOULD NOT be used for the Flexible FEC payload format 1824 8. Congestion Control Considerations 1826 FEC is an effective approach to provide applications resiliency 1827 against packet losses. However, in networks where the congestion is 1828 a major contributor to the packet loss, the potential impacts of 1829 using FEC should be considered carefully before injecting the repair 1830 streams into the network. In particular, in bandwidth-limited 1831 networks, FEC repair streams may consume a significant part of the 1832 available bandwidth and consequently may congest the network. In 1833 such cases, the applications MUST NOT arbitrarily increase the amount 1834 of FEC protection since doing so may lead to a congestion collapse. 1835 If desired, stronger FEC protection MAY be applied only after the 1836 source rate has been reduced. 1838 In a network-friendly implementation, an application should avoid 1839 sending/receiving FEC repair streams if it knows that sending/ 1840 receiving those FEC repair streams would not help at all in 1841 recovering the missing packets. It is RECOMMENDED that the amount 1842 and type (row, column, or both) of FEC protection is adjusted 1843 dynamically based on the packet loss rate and burst loss length 1844 observed by the applications. 1846 In multicast scenarios, it may be difficult to optimize the FEC 1847 protection per receiver. If there is a large variation among the 1848 levels of FEC protection needed by different receivers, it is 1849 RECOMMENDED that the sender offers multiple repair streams with 1850 different levels of FEC protection and the receivers join the 1851 corresponding multicast sessions to receive the repair stream(s) that 1852 is best for them. 1854 9. Security Considerations 1856 RTP packets using the payload format defined in this specification 1857 are subject to the security considerations discussed in the RTP 1858 specification [RFC3550] and in any applicable RTP profile. The main 1859 security considerations for the RTP packet carrying the RTP payload 1860 format defined within this memo are confidentiality, integrity and 1861 source authenticity. Confidentiality is achieved by encrypting the 1862 RTP payload. Integrity of the RTP packets is achieved through a 1863 suitable cryptographic integrity protection mechanism. Such a 1864 cryptographic system may also allow the authentication of the source 1865 of the payload. A suitable security mechanism for this RTP payload 1866 format should provide confidentiality, integrity protection, and at 1867 least source authentication capable of determining if an RTP packet 1868 is from a member of the RTP session. 1870 Note that the appropriate mechanism to provide security to RTP and 1871 payloads following this memo may vary. It is dependent on the 1872 application, transport and signaling protocol employed. Therefore, a 1873 single mechanism is not sufficient, although if suitable, using the 1874 Secure Real-time Transport Protocol (SRTP) [RFC3711] is recommended. 1875 Other mechanisms that may be used are IPsec [RFC4301] and Transport 1876 Layer Security (TLS, see [RFC8446]) (RTP over TCP); other 1877 alternatives may exist. 1879 Given that FLEX FEC enables the protection of multiple source 1880 streams, there exists the possibility that multiple source buffers 1881 may be created that may not be used. An attacker could leverage 1882 unused source buffers to as a means of occupying memory in a FLEX FEC 1883 endpoint. Moreover the application source data may not be perfectly 1884 matched with FLEX FEC source partitioning. If this is the case, 1885 there is a possibility for unprotected source data if, for instance, 1886 the FLEX FEC implementation discards data that does not fit perfectly 1887 into its source processing requirements. 1889 10. IANA Considerations 1891 New media subtypes are subject to IANA registration. For the 1892 registration of the payload formats and their parameters introduced 1893 in this document, refer to Section 5. 1895 11. Acknowledgments 1897 Some parts of this document are borrowed from [RFC5109]. Thus, the 1898 author would like to thank the editor of [RFC5109] and those who 1899 contributed to [RFC5109]. 1901 Thanks to Stephen Botzko , Bernard Aboba , Rasmus Brandt , Brian 1902 Baldino , Roni Even , Stefan Holmer , Jonathan Lennox , and Magnus 1903 Westerlund for providing valuable feedback on earlier versions of 1904 this draft. 1906 12. References 1908 12.1. Normative References 1910 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1911 Requirement Levels", BCP 14, RFC 2119, 1912 DOI 10.17487/RFC2119, March 1997, 1913 . 1915 [RFC3264] Rosenberg, J. and H. Schulzrinne, "An Offer/Answer Model 1916 with Session Description Protocol (SDP)", RFC 3264, 1917 DOI 10.17487/RFC3264, June 2002, 1918 . 1920 [RFC3550] Schulzrinne, H., Casner, S., Frederick, R., and V. 1921 Jacobson, "RTP: A Transport Protocol for Real-Time 1922 Applications", STD 64, RFC 3550, DOI 10.17487/RFC3550, 1923 July 2003, . 1925 [RFC4566] Handley, M., Jacobson, V., and C. Perkins, "SDP: Session 1926 Description Protocol", RFC 4566, DOI 10.17487/RFC4566, 1927 July 2006, . 1929 [RFC4855] Casner, S., "Media Type Registration of RTP Payload 1930 Formats", RFC 4855, DOI 10.17487/RFC4855, February 2007, 1931 . 1933 [RFC4856] Casner, S., "Media Type Registration of Payload Formats in 1934 the RTP Profile for Audio and Video Conferences", 1935 RFC 4856, DOI 10.17487/RFC4856, February 2007, 1936 . 1938 [RFC5956] Begen, A., "Forward Error Correction Grouping Semantics in 1939 the Session Description Protocol", RFC 5956, 1940 DOI 10.17487/RFC5956, September 2010, 1941 . 1943 [RFC6363] Watson, M., Begen, A., and V. Roca, "Forward Error 1944 Correction (FEC) Framework", RFC 6363, 1945 DOI 10.17487/RFC6363, October 2011, 1946 . 1948 [RFC6838] Freed, N., Klensin, J., and T. Hansen, "Media Type 1949 Specifications and Registration Procedures", BCP 13, 1950 RFC 6838, DOI 10.17487/RFC6838, January 2013, 1951 . 1953 [RFC7022] Begen, A., Perkins, C., Wing, D., and E. Rescorla, 1954 "Guidelines for Choosing RTP Control Protocol (RTCP) 1955 Canonical Names (CNAMEs)", RFC 7022, DOI 10.17487/RFC7022, 1956 September 2013, . 1958 [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 1959 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 1960 May 2017, . 1962 12.2. Informative References 1964 [I-D.ietf-avtext-rid] 1965 Roach, A., Nandakumar, S., and P. Thatcher, "RTP Stream 1966 Identifier Source Description (SDES)", draft-ietf-avtext- 1967 rid-09 (work in progress), October 2016. 1969 [RFC2326] Schulzrinne, H., Rao, A., and R. Lanphier, "Real Time 1970 Streaming Protocol (RTSP)", RFC 2326, 1971 DOI 10.17487/RFC2326, April 1998, 1972 . 1974 [RFC2733] Rosenberg, J. and H. Schulzrinne, "An RTP Payload Format 1975 for Generic Forward Error Correction", RFC 2733, 1976 DOI 10.17487/RFC2733, December 1999, 1977 . 1979 [RFC2974] Handley, M., Perkins, C., and E. Whelan, "Session 1980 Announcement Protocol", RFC 2974, DOI 10.17487/RFC2974, 1981 October 2000, . 1983 [RFC3711] Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K. 1984 Norrman, "The Secure Real-time Transport Protocol (SRTP)", 1985 RFC 3711, DOI 10.17487/RFC3711, March 2004, 1986 . 1988 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 1989 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, 1990 December 2005, . 1992 [RFC4585] Ott, J., Wenger, S., Sato, N., Burmeister, C., and J. Rey, 1993 "Extended RTP Profile for Real-time Transport Control 1994 Protocol (RTCP)-Based Feedback (RTP/AVPF)", RFC 4585, 1995 DOI 10.17487/RFC4585, July 2006, 1996 . 1998 [RFC5109] Li, A., Ed., "RTP Payload Format for Generic Forward Error 1999 Correction", RFC 5109, DOI 10.17487/RFC5109, December 2000 2007, . 2002 [RFC7656] Lennox, J., Gross, K., Nandakumar, S., Salgueiro, G., and 2003 B. Burman, Ed., "A Taxonomy of Semantics and Mechanisms 2004 for Real-Time Transport Protocol (RTP) Sources", RFC 7656, 2005 DOI 10.17487/RFC7656, November 2015, 2006 . 2008 [RFC7826] Schulzrinne, H., Rao, A., Lanphier, R., Westerlund, M., 2009 and M. Stiemerling, Ed., "Real-Time Streaming Protocol 2010 Version 2.0", RFC 7826, DOI 10.17487/RFC7826, December 2011 2016, . 2013 [RFC8446] Rescorla, E., "The Transport Layer Security (TLS) Protocol 2014 Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018, 2015 . 2017 [SMPTE2022-1] 2018 SMPTE 2022-1-2007, "Forward Error Correction for Real-Time 2019 Video/Audio Transport over IP Networks", 2007. 2021 Authors' Addresses 2023 Mo Zanaty 2024 Cisco 2025 Raleigh, NC 2026 USA 2028 Email: mzanaty@cisco.com 2030 Varun Singh 2031 CALLSTATS I/O Oy 2032 Runeberginkatu 4c A 4 2033 Helsinki 00100 2034 Finland 2036 Email: varun.singh@iki.fi 2037 URI: http://www.callstats.io/ 2039 Ali Begen 2040 Networked Media 2041 Konya 2042 Turkey 2044 Email: ali.begen@networked.media 2046 Giridhar Mandyam 2047 Qualcomm Inc. 2048 5775 Morehouse Drive 2049 San Diego, CA 92121 2050 USA 2052 Phone: +1 858 651 7200 2053 Email: mandyam@qti.qualcomm.com