idnits 2.17.1 draft-ietf-tsvwg-fecframe-ext-02.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (May 1, 2018) is 2185 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) No issues found here. Summary: 0 errors (**), 0 flaws (~~), 1 warning (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 TSVWG V. Roca 3 Internet-Draft INRIA 4 Intended status: Standards Track A. Begen 5 Expires: November 2, 2018 Networked Media 6 May 1, 2018 8 Forward Error Correction (FEC) Framework Extension to Sliding Window 9 Codes 10 draft-ietf-tsvwg-fecframe-ext-02 12 Abstract 14 RFC 6363 describes a framework for using Forward Error Correction 15 (FEC) codes with applications in public and private IP networks to 16 provide protection against packet loss. The framework supports 17 applying FEC to arbitrary packet flows over unreliable transport and 18 is primarily intended for real-time, or streaming, media. However 19 FECFRAME as per RFC 6363 is restricted to block FEC codes. The 20 present document extends FECFRAME to support FEC Codes based on a 21 sliding encoding window, in addition to Block FEC Codes, in a 22 backward compatible way. During multicast/broadcast real-time 23 content delivery, the use of sliding window codes significantly 24 improves robustness in harsh environments, with less repair traffic 25 and lower FEC-related added latency. 27 Status of This Memo 29 This Internet-Draft is submitted in full conformance with the 30 provisions of BCP 78 and BCP 79. 32 Internet-Drafts are working documents of the Internet Engineering 33 Task Force (IETF). Note that other groups may also distribute 34 working documents as Internet-Drafts. The list of current Internet- 35 Drafts is at https://datatracker.ietf.org/drafts/current/. 37 Internet-Drafts are draft documents valid for a maximum of six months 38 and may be updated, replaced, or obsoleted by other documents at any 39 time. It is inappropriate to use Internet-Drafts as reference 40 material or to cite them other than as "work in progress." 42 This Internet-Draft will expire on November 2, 2018. 44 Copyright Notice 46 Copyright (c) 2018 IETF Trust and the persons identified as the 47 document authors. All rights reserved. 49 This document is subject to BCP 78 and the IETF Trust's Legal 50 Provisions Relating to IETF Documents 51 (https://trustee.ietf.org/license-info) in effect on the date of 52 publication of this document. Please review these documents 53 carefully, as they describe your rights and restrictions with respect 54 to this document. Code Components extracted from this document must 55 include Simplified BSD License text as described in Section 4.e of 56 the Trust Legal Provisions and are provided without warranty as 57 described in the Simplified BSD License. 59 Table of Contents 61 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 62 2. Definitions and Abbreviations . . . . . . . . . . . . . . . . 4 63 3. Architecture Overview . . . . . . . . . . . . . . . . . . . . 7 64 4. Procedural Overview . . . . . . . . . . . . . . . . . . . . . 9 65 4.1. General . . . . . . . . . . . . . . . . . . . . . . . . . 9 66 4.2. Sender Operation with Sliding Window FEC Codes . . . . . 10 67 4.3. Receiver Operation with Sliding Window FEC Codes . . . . 12 68 5. Protocol Specification . . . . . . . . . . . . . . . . . . . 14 69 5.1. General . . . . . . . . . . . . . . . . . . . . . . . . . 14 70 5.2. FEC Framework Configuration Information . . . . . . . . . 15 71 5.3. FEC Scheme Requirements . . . . . . . . . . . . . . . . . 15 72 6. Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . 15 73 7. Transport Protocols . . . . . . . . . . . . . . . . . . . . . 16 74 8. Congestion Control . . . . . . . . . . . . . . . . . . . . . 16 75 9. Implementation Status . . . . . . . . . . . . . . . . . . . . 16 76 10. Security Considerations . . . . . . . . . . . . . . . . . . . 16 77 11. Operations and Management Considerations . . . . . . . . . . 17 78 12. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 17 79 13. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 17 80 14. References . . . . . . . . . . . . . . . . . . . . . . . . . 17 81 14.1. Normative References . . . . . . . . . . . . . . . . . . 17 82 14.2. Informative References . . . . . . . . . . . . . . . . . 17 83 Appendix A. About Sliding Encoding Window Management (non 84 Normative) . . . . . . . . . . . . . . . . . . . . . 19 85 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 20 87 1. Introduction 89 Many applications need to transport a continuous stream of packetized 90 data from a source (sender) to one or more destinations (receivers) 91 over networks that do not provide guaranteed packet delivery. In 92 particular packets may be lost, which is strictly the focus of this 93 document: we assume that transmitted packets are either received 94 without any corruption or totally lost (e.g., because of a congested 95 router, of a poor signal-to-noise ratio in a wireless network, or 96 because the number of bit errors exceeds the correction capabilities 97 of a low-layer error correcting code). 99 For these use-cases, Forward Error Correction (FEC) applied within 100 the transport or application layer, is an efficient technique to 101 improve packet transmission robustness in presence of packet losses 102 (or "erasures"), without going through packet retransmissions that 103 create a delay often incompatible with real-time constraints. The 104 FEC Building Block defined in [RFC5052] provides a framework for the 105 definition of Content Delivery Protocols (CDPs) that make use of 106 separately defined FEC schemes. Any CDP defined according to the 107 requirements of the FEC Building Block can then easily be used with 108 any FEC Scheme that is also defined according to the requirements of 109 the FEC Building Block. 111 Then FECFRAME [RFC6363] provides a framework to define Content 112 Delivery Protocols (CDPs) that provide FEC protection for arbitrary 113 packet flows over unreliable transports such as UDP. It is primarily 114 intended for real-time or streaming media applications, using 115 broadcast, multicast, or on-demand delivery. 117 However [RFC6363] only considers block FEC schemes defined in 118 accordance with the FEC Building Block [RFC5052] (e.g., [RFC6681], 119 [RFC6816] or [RFC6865]). These codes require the input flow(s) to be 120 segmented into a sequence of blocks. Then FEC encoding (at a sender 121 or an encoding middlebox) and decoding (at a receiver or a decoding 122 middlebox) are both performed on a per-block basis. This approach 123 has major impacts on FEC encoding and decoding delays. The data 124 packets of continuous media flow(s) may be passed to the transport 125 layer immediately, without delay. But the block creation time, that 126 depends on the number k of source symbols in this block, impacts the 127 FEC encoding delay since encoding requires that all source symbols be 128 known. This block creation time also impacts the decoding delay a 129 receiver will experience in case of erasures, since no repair symbol 130 for the current block can be received before. Therefore a good value 131 for the block size is necessarily a balance between the maximum 132 decoding latency at the receivers (which must be in line with the 133 most stringent real-time requirement of the protected flow(s), hence 134 an incentive to reduce the block size), and the desired robustness 135 against long loss bursts (which increases with the block size, hence 136 an incentive to increase this size). 138 This document extends [RFC6363] in order to also support FEC codes 139 based on a sliding encoding window (A.K.A. convolutional codes). 140 This encoding window, either of fixed or variable size, slides over 141 the set of source symbols. FEC encoding is launched whenever needed, 142 from the set of source symbols present in the sliding encoding window 143 at that time. This approach significantly reduces FEC-related 144 latency, since repair symbols can be generated and passed to the 145 transport layer on-the-fly, at any time, and can be regularly 146 received by receivers to quickly recover packet losses. Using 147 sliding window FEC codes is therefore highly beneficial to real-time 148 flows, one of the primary targets of FECFRAME. [RLC-ID] provides an 149 example of such FEC Scheme for FECFRAME, built upon the simple 150 sliding window Random Linear Codes (RLC). 152 This document is fully backward compatible with [RFC6363] that it 153 extends but does not replace. Indeed: 155 o this extension does not prevent nor compromize in any way the 156 support of block FEC codes. Both types of codes can nicely co- 157 exist, just like different block FEC schemes can co-exist; 159 o any receiver, for instance a legacy receiver that only supports 160 block FEC schemes, can easily identify the FEC Scheme used in a 161 FECFRAME session thanks to the associated SDP file and its FEC 162 Encoding ID information (i.e., the "encoding-id=" parameter of a 163 "fec-repair-flow" attribute, [RFC6364]). This mechanism is not 164 specific to this extension but is the basic approach for a 165 FECFRAME receiver to determine whether or not it supports the FEC 166 Scheme used in a given FECFRAME session; 168 This document leverages on [RFC6363] and re-uses its structure. It 169 proposes new sections specific to sliding window FEC codes whenever 170 required. The only exception is Section Section 3 that provides a 171 quick summary of FECFRAME in order to facilitate the understanding of 172 this document to readers not familiar with the concepts and 173 terminology. 175 2. Definitions and Abbreviations 177 The following list of definitions and abbreviations is copied from 178 [RFC6363], adding only the Block/sliding window FEC Code and 179 Encoding/Decoding Window definitions (tagged with "ADDED"): 181 Application Data Unit (ADU): The unit of source data provided as 182 payload to the transport layer. 184 ADU Flow: A sequence of ADUs associated with a transport-layer flow 185 identifier (such as the standard 5-tuple {source IP address, 186 source port, destination IP address, destination port, transport 187 protocol}). 189 AL-FEC: Application-layer Forward Error Correction. 191 Application Protocol: Control protocol used to establish and control 192 the source flow being protected, e.g., the Real-Time Streaming 193 Protocol (RTSP). 195 Content Delivery Protocol (CDP): A complete application protocol 196 specification that, through the use of the framework defined in 197 this document, is able to make use of FEC schemes to provide FEC 198 capabilities. 200 FEC Code: An algorithm for encoding data such that the encoded data 201 flow is resilient to data loss. Note that, in general, FEC codes 202 may also be used to make a data flow resilient to corruption, but 203 that is not considered in this document. 205 Block FEC Code: (ADDED) An FEC Code that operates in a block manner, 206 i.e., for which the input flow MUST be segmented into a sequence 207 of blocks, FEC encoding and decoding being performed 208 independently on a per-block basis. 210 Sliding Window (or Convolutional) FEC Code: (ADDED) An FEC Code that 211 can generate repair symbols on-the-fly, at any time, from the set 212 of source symbols present in the sliding encoding window at that 213 time. 215 FEC Framework: A protocol framework for the definition of Content 216 Delivery Protocols using FEC, such as the framework defined in 217 this document. 219 FEC Framework Configuration Information: Information that controls 220 the operation of the FEC Framework. 222 FEC Payload ID: Information that identifies the contents of a packet 223 with respect to the FEC Scheme. 225 FEC Repair Packet: At a sender (respectively, at a receiver), a 226 payload submitted to (respectively, received from) the transport 227 protocol containing one or more repair symbols along with a 228 Repair FEC Payload ID and possibly an RTP header. 230 FEC Scheme: A specification that defines the additional protocol 231 aspects required to use a particular FEC code with the FEC 232 Framework. 234 FEC Source Packet: At a sender (respectively, at a receiver), a 235 payload submitted to (respectively, received from) the transport 236 protocol containing an ADU along with an optional Explicit Source 237 FEC Payload ID. 239 Repair Flow: The packet flow carrying FEC data. 241 Repair FEC Payload ID: A FEC Payload ID specifically for use with 242 repair packets. 244 Source Flow: The packet flow to which FEC protection is to be 245 applied. A source flow consists of ADUs. 247 Source FEC Payload ID: A FEC Payload ID specifically for use with 248 source packets. 250 Source Protocol: A protocol used for the source flow being 251 protected, e.g., RTP. 253 Transport Protocol: The protocol used for the transport of the 254 source and repair flows, e.g., UDP and the Datagram Congestion 255 Control Protocol (DCCP). 257 Encoding Window: (ADDED) Set of Source Symbols available at the 258 sender/coding node that are used to generate a repair symbol, 259 with a Sliding Window FEC Code. 261 Decoding Window: (ADDED) Set of received or decoded source and 262 repair symbols available at a receiver that are used to decode 263 erased source symbols, with a Sliding Window FEC Code. 265 Code Rate: The ratio between the number of source symbols and the 266 number of encoding symbols. By definition, the code rate is such 267 that 0 < code rate <= 1. A code rate close to 1 indicates that a 268 small number of repair symbols have been produced during the 269 encoding process. 271 Encoding Symbol: Unit of data generated by the encoding process. 272 With systematic codes, source symbols are part of the encoding 273 symbols. 275 Packet Erasure Channel: A communication path where packets are 276 either lost (e.g., by a congested router, or because the number 277 of transmission errors exceeds the correction capabilities of the 278 physical-layer codes) or received. When a packet is received, it 279 is assumed that this packet is not corrupted. 281 Repair Symbol: Encoding symbol that is not a source symbol. 283 Source Block: Group of ADUs that are to be FEC protected as a single 284 block. This notion is restricted to Block FEC Codes. 286 Source Symbol: Unit of data used during the encoding process. 288 Systematic Code: FEC code in which the source symbols are part of 289 the encoding symbols. 291 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 292 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 293 document are to be interpreted as described in [RFC2119]. 295 3. Architecture Overview 297 The architecture of [RFC6363], Section 3, equally applies to this 298 FECFRAME extension and is not repeated here. However we provide 299 hereafter a quick summary to facilitate the understanding of this 300 document to readers not familiar with the concepts and terminology. 302 +----------------------+ 303 | Application | 304 +----------------------+ 305 | 306 | (1) Application Data Units (ADUs) 307 | 308 v 309 +----------------------+ +----------------+ 310 | FEC Framework | | | 311 | |-------------------------->| FEC Scheme | 312 |(2) Construct source |(3) Source Block | | 313 | blocks | |(4) FEC Encoding| 314 |(6) Construct FEC |<--------------------------| | 315 | Source and Repair | | | 316 | Packets |(5) Explicit Source FEC | | 317 +----------------------+ Payload IDs +----------------+ 318 | Repair FEC Payload IDs 319 | Repair symbols 320 | 321 |(7) FEC Source and Repair Packets 322 v 323 +----------------------+ 324 | Transport Layer | 325 | (e.g., UDP) | 326 +----------------------+ 328 Figure 1: FECFRAME architecture at a sender. 330 The FECFRAME architecture is illustrated in Figure 1 from the 331 sender's point of view, in case of a block FEC Scheme. It shows an 332 application generating an ADU flow (other flows, from other 333 applications, may co-exist). These ADUs, of variable size, must be 334 somehow mapped to source symbols of fixed size. This is the goal of 335 an ADU to symbols mapping process that is FEC Scheme specific (see 336 below). Once the source block is built, taking into account both the 337 FEC Scheme constraints (e.g., in terms of maximum source block size) 338 and the application's flow constraints (e.g., in terms of real-time 339 constraints), the associated source symbols are handed to the FEC 340 Scheme in order to produce an appropriate number of repair symbols. 341 FEC Source Packets (containing ADUs) and FEC Repair Packets 342 (containing one or more repair symbols each) are then generated and 343 sent using UDP (more precisely [RFC6363], Section 7, requires a 344 transport protocol providing an unreliable datagram service, like UDP 345 or DCCP). In practice FEC Source Packets may be passed to the 346 transport layer as soon as available, without having to wait for FEC 347 encoding to take place. In that case a copy of the associated source 348 symbols needs to be kept within FECFRAME for future FEC encoding 349 purposes. 351 At a receiver (not shown), FECFRAME processing operates in a similar 352 way, taking as input the incoming FEC Source and Repair Packets 353 received. In case of FEC Source Packet losses, the FEC decoding of 354 the associated block may recover all (in case of successful decoding) 355 or a subset potentially empty (otherwise) of the missing source 356 symbols. After source symbol to ADU mapping, when lost ADUs are 357 recovered, they are then assigned to their respective flow (see 358 below). ADUs are returned to the application(s), either in their 359 initial transmission order (in that case ADUs received after an 360 erased one will be delayed until FEC decoding has taken place) or not 361 (in that case each ADU is returned as soon as it is received or 362 recovered), depending on the application requirements. 364 FECFRAME features two subtle mechanisms: 366 o ADUs to source symbols mapping: in order to manage variable size 367 ADUs, FECFRAME and FEC Schemes can use small, fixed size symbols 368 and create a mapping between ADUs and symbols. To each ADU this 369 mechanism prepends a length field (plus a flow identifier, see 370 below) and pads the result to a multiple of the symbol size. A 371 small ADU may be mapped to a single source symbol while a large 372 one may be mapped to multiple symbols. The mapping details are 373 FEC Scheme dependant and must be defined in the associated 374 document; 376 o Assignment of decoded ADUs to flows in multi-flow configurations: 377 when multiple flows are multiplexed over the same FECFRAME 378 instance, a problem is to assign a decoded ADU to the right flow 379 (UDP port numbers and IP addresses traditionally used to map 380 incoming ADUs to flows are not recovered during FEC decoding). To 381 make it possible, at the FECFRAME sending instance, each ADU is 382 prepended with a flow identifier (1 byte) during the ADU to source 383 symbols mapping (see above). The flow identifiers are also shared 384 between all FECFRAME instances as part of the FEC Framework 385 Configuration Information. This (flow identifier + length + 386 application payload + padding), called ADUI, is then FEC 387 protected. Therefore a decoded ADUI contains enough information 388 to assign the ADU to the right flow. 390 A few aspects are not covered by FECFRAME, namely: 392 o [RFC6363] section 8 does not detail any congestion control 393 mechanism, but only provides high level normative requirements; 395 o the possibility of having feedbacks from receiver(s) is considered 396 out of scope, although such a mechanism may exist within the 397 application (e.g., through RCTP control messages); 399 o flow adaptation at a FECFRAME sender (e.g., how to set the FEC 400 code rate based on transmission conditions) is not detailed, but 401 it needs to comply with the congestion control normative 402 requirements (see above). 404 4. Procedural Overview 406 4.1. General 408 The general considerations of [RFC6363], Section 4.1, that are 409 specific to block FEC codes are not repeated here. 411 With a Sliding Window FEC Code, the FEC Source Packet MUST contain 412 information to identify the position occupied by the ADU within the 413 source flow, in terms specific to the FEC Scheme. This information 414 is known as the Source FEC Payload ID, and the FEC Scheme is 415 responsible for defining and interpreting it. 417 With a Sliding Window FEC Code, the FEC Repair Packets MUST contain 418 information that identifies the relationship between the contained 419 repair payloads and the original source symbols used during encoding. 420 This information is known as the Repair FEC Payload ID, and the FEC 421 Scheme is responsible for defining and interpreting it. 423 The Sender Operation ([RFC6363], Section 4.2.) and Receiver Operation 424 ([RFC6363], Section 4.3) are both specific to block FEC codes and 425 therefore omitted below. The following two sections detail similar 426 operations for Sliding Window FEC codes. 428 4.2. Sender Operation with Sliding Window FEC Codes 430 With a Sliding Window FEC Scheme, the following operations, 431 illustrated in Figure 2 for the case of UDP repair flows, and in 432 Figure 3 for the case of RTP repair flows, describe a possible way to 433 generate compliant source and repair flows: 435 1. A new ADU is provided by the application. 437 2. The FEC Framework communicates this ADU to the FEC Scheme. 439 3. The sliding encoding window is updated by the FEC Scheme. The 440 ADU to source symbols mapping as well as the encoding window 441 management details are both the responsibility of the FEC Scheme 442 and MUST be detailed there. Appendix A provides some hints on 443 the way it might be performed. 445 4. The Source FEC Payload ID information of the source packet is 446 determined by the FEC Scheme. If required by the FEC Scheme, 447 the Source FEC Payload ID is encoded into the Explicit Source 448 FEC Payload ID field and returned to the FEC Framework. 450 5. The FEC Framework constructs the FEC Source Packet according to 451 [RFC6363] Figure 6, using the Explicit Source FEC Payload ID 452 provided by the FEC Scheme if applicable. 454 6. The FEC Source Packet is sent using normal transport-layer 455 procedures. This packet is sent using the same ADU flow 456 identification information as would have been used for the 457 original source packet if the FEC Framework were not present 458 (for example, in the UDP case, the UDP source and destination 459 addresses and ports on the IP datagram carrying the source 460 packet will be the same whether or not the FEC Framework is 461 applied). 463 7. When the FEC Framework needs to send one or several FEC Repair 464 Packets (e.g., according to the target Code Rate), it asks the 465 FEC Scheme to create one or several repair packet payloads from 466 the current sliding encoding window along with their Repair FEC 467 Payload ID. 469 8. The Repair FEC Payload IDs and repair packet payloads are 470 provided back by the FEC Scheme to the FEC Framework. 472 9. The FEC Framework constructs FEC Repair Packets according to 473 [RFC6363] Figure 7, using the FEC Payload IDs and repair packet 474 payloads provided by the FEC Scheme. 476 10. The FEC Repair Packets are sent using normal transport-layer 477 procedures. The port(s) and multicast group(s) to be used for 478 FEC Repair Packets are defined in the FEC Framework 479 Configuration Information. 481 +----------------------+ 482 | Application | 483 +----------------------+ 484 | 485 | (1) New Application Data Unit (ADU) 486 v 487 +---------------------+ +----------------+ 488 | FEC Framework | | FEC Scheme | 489 | |-------------------------->| | 490 | | (2) New ADU |(3) Update of | 491 | | | encoding | 492 | |<--------------------------| window | 493 |(5) Construct FEC | (4) Explicit Source | | 494 | Source Packet | FEC Payload ID(s) |(7) FEC | 495 | |<--------------------------| encoding | 496 |(9) Construct FEC | (8) Repair FEC Payload ID | | 497 | Repair Packet(s) | + Repair symbol(s) +----------------+ 498 +---------------------+ 499 | 500 | (6) FEC Source Packet 501 | (10) FEC Repair Packets 502 v 503 +----------------------+ 504 | Transport Layer | 505 | (e.g., UDP) | 506 +----------------------+ 508 Figure 2: Sender Operation with Convolutional FEC Codes 510 +----------------------+ 511 | Application | 512 +----------------------+ 513 | 514 | (1) New Application Data Unit (ADU) 515 v 516 +---------------------+ +----------------+ 517 | FEC Framework | | FEC Scheme | 518 | |-------------------------->| | 519 | | (2) New ADU |(3) Update of | 520 | | | encoding | 521 | |<--------------------------| window | 522 |(5) Construct FEC | (4) Explicit Source | | 523 | Source Packet | FEC Payload ID(s) |(7) FEC | 524 | |<--------------------------| encoding | 525 |(9) Construct FEC | (8) Repair FEC Payload ID | | 526 | Repair Packet(s) | + Repair symbol(s) +----------------+ 527 +---------------------+ 528 | | 529 |(6) Source |(10) Repair payloads 530 | packets | 531 | + -- -- -- -- -+ 532 | | RTP | 533 | +-- -- -- -- --+ 534 v v 535 +----------------------+ 536 | Transport Layer | 537 | (e.g., UDP) | 538 +----------------------+ 540 Figure 3: Sender Operation with RTP Repair Flows 542 4.3. Receiver Operation with Sliding Window FEC Codes 544 With a Sliding Window FEC Scheme, the following operations, 545 illustrated in Figure 4 for the case of UDP repair flows, and in 546 Figure 5 for the case of RTP repair flows. The only differences with 547 respect to block FEC codes lie in steps (4) and (5). Therefore this 548 section does not repeat the other steps of [RFC6363], Section 4.3, 549 "Receiver Operation". The new steps (4) and (5) are: 551 4. The FEC Scheme uses the received FEC Payload IDs (and derived FEC 552 Source Payload IDs when the Explicit Source FEC Payload ID field 553 is not used) to insert source and repair packets into the 554 decoding window in the right way. If at least one source packet 555 is missing and at least one repair packet has been received and 556 the rank of the associated linear system permits it, then FEC 557 decoding can be performed in order to recover missing source 558 payloads. The FEC Scheme determines whether source packets have 559 been lost and whether enough repair packets have been received to 560 decode any or all of the missing source payloads. 562 5. The FEC Scheme returns the received and decoded ADUs to the FEC 563 Framework, along with indications of any ADUs that were missing 564 and could not be decoded. 566 +----------------------+ 567 | Application | 568 +----------------------+ 569 ^ 570 |(6) ADUs 571 | 572 +----------------------+ +----------------+ 573 | FEC Framework | | FEC Scheme | 574 | |<--------------------------| | 575 |(2)Extract FEC Payload|(5) ADUs |(4) FEC Decoding 576 | IDs and pass IDs & |-------------------------->| | 577 | payloads to FEC |(3) Explicit Source FEC +----------------+ 578 | scheme | Payload IDs 579 +----------------------+ Repair FEC Payload IDs 580 ^ Source payloads 581 | Repair payloads 582 |(1) FEC Source 583 | and Repair Packets 584 +----------------------+ 585 | Transport Layer | 586 | (e.g., UDP) | 587 +----------------------+ 589 Figure 4: Receiver Operation with Sliding Window FEC Codes 591 +----------------------+ 592 | Application | 593 +----------------------+ 594 ^ 595 |(6) ADUs 596 | 597 +----------------------+ +----------------+ 598 | FEC Framework | | FEC Scheme | 599 | |<--------------------------| | 600 |(2)Extract FEC Payload|(5) ADUs |(4) FEC Decoding| 601 | IDs and pass IDs & |-------------------------->| | 602 | payloads to FEC |(3) Explicit Source FEC +----------------+ 603 | scheme | Payload IDs 604 +----------------------+ Repair FEC Payload IDs 605 ^ ^ Source payloads 606 | | Repair payloads 607 |Source pkts |Repair payloads 608 | | 609 +-- |- -- -- -- -- -- -+ 610 |RTP| | RTP Processing | 611 | | +-- -- -- --|-- -+ 612 | +-- -- -- -- -- |--+ | 613 | | RTP Demux | | 614 +-- -- -- -- -- -- -- -+ 615 ^ 616 |(1) FEC Source and Repair Packets 617 | 618 +----------------------+ 619 | Transport Layer | 620 | (e.g., UDP) | 621 +----------------------+ 623 Figure 5: Receiver Operation with RTP Repair Flows 625 5. Protocol Specification 627 5.1. General 629 This section discusses the protocol elements for the FEC Framework 630 specific to Sliding Window FEC schemes. The global formats of source 631 data packets (i.e., [RFC6363], Figure 6) and repair data packets 632 (i.e., [RFC6363], Figures 7 and 8) remain the same with Sliding 633 Window FEC codes. They are not repeated here. 635 5.2. FEC Framework Configuration Information 637 The FEC Framework Configuration Information considerations of 638 [RFC6363], Section 5.5, equally applies to this FECFRAME extension 639 and is not repeated here. 641 5.3. FEC Scheme Requirements 643 The FEC Scheme requirements of [RFC6363], Section 5.6, mostly apply 644 to this FECFRAME extension and are not repeated here. An exception 645 though is the "full specification of the FEC code", item (4), that is 646 specific to block FEC codes. The following item (4) applies instead: 648 4. A full specification of the Sliding Window FEC code 650 This specification MUST precisely define the valid FEC-Scheme- 651 Specific Information values, the valid FEC Payload ID values, and 652 the valid packet payload sizes (where packet payload refers to 653 the space within a packet dedicated to carrying encoding 654 symbols). 656 Furthermore, given valid values of the FEC-Scheme-Specific 657 Information, a valid Repair FEC Payload ID value, a valid packet 658 payload size, and a valid encoding window (i.e., a set of source 659 symbols), the specification MUST uniquely define the values of 660 the encoding symbols to be included in the repair packet payload 661 with the given Repair FEC Payload ID value. 663 Additionally, the FEC Scheme associated to a Sliding Window FEC Code: 665 o MUST define the relationships between ADUs and the associated 666 source symbols (mapping); 668 o MUST define the management of the encoding window that slides over 669 the set of ADUs. Appendix A provides a non normative example; 671 o MUST define the management of the decoding window, consisting of a 672 system of linear equations (in case of a linear FEC code); 674 6. Feedback 676 The discussion of [RFC6363], Section 6, equally applies to this 677 FECFRAME extension and is not repeated here. 679 7. Transport Protocols 681 The discussion of [RFC6363], Section 7, equally applies to this 682 FECFRAME extension and is not repeated here. 684 8. Congestion Control 686 The discussion of [RFC6363], Section 8, equally applies to this 687 FECFRAME extension and is not repeated here. 689 9. Implementation Status 691 Editor's notes: RFC Editor, please remove this section motivated by 692 RFC 7942 before publishing the RFC. Thanks! 694 An implementation of FECFRAME extended to Sliding Window codes 695 exists: 697 o Organisation: Inria 699 o Description: This is an implementation of FECFRAME extended to 700 Sliding Window codes and supporting the RLC FEC Scheme [RLC-ID]. 701 It is based on: (1) a proprietary implementation of FECFRAME, made 702 by Inria and Expway for which interoperability tests have been 703 conducted; and (2) a proprietary implementation of RLC Sliding 704 Window FEC Codes. 706 o Maturity: the basic FECFRAME maturity is "production", the 707 FECFRAME extension maturity is "under progress". 709 o Coverage: the software implements a subset of [RFC6363], as 710 specialized by the 3GPP eMBMS standard [MBMSTS]. This software 711 also covers the additional features of FECFRAME extended to 712 Sliding Window codes, in particular the RLC FEC Scheme. 714 o Lincensing: proprietary. 716 o Implementation experience: maximum. 718 o Information update date: March 2018. 720 o Contact: vincent.roca@inria.fr 722 10. Security Considerations 724 This FECFRAME extension does not add any new security consideration. 725 All the considerations of [RFC6363], Section 9, apply to this 726 document as well. 728 11. Operations and Management Considerations 730 This FECFRAME extension does not add any new Operations and 731 Management Consideration. All the considerations of [RFC6363], 732 Section 10, apply to this document as well. 734 12. IANA Considerations 736 A FEC Scheme for use with this FEC Framework is identified via its 737 FEC Encoding ID. It is subject to IANA registration in the "FEC 738 Framework (FECFRAME) FEC Encoding IDs" registry. All the rules of 739 [RFC6363], Section 11, apply and are not repeated here. 741 13. Acknowledgments 743 TBD 745 14. References 747 14.1. Normative References 749 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 750 Requirement Levels", BCP 14, RFC 2119, 751 DOI 10.17487/RFC2119, March 1997, 752 . 754 [RFC6363] Watson, M., Begen, A., and V. Roca, "Forward Error 755 Correction (FEC) Framework", RFC 6363, 756 DOI 10.17487/RFC6363, October 2011, 757 . 759 14.2. Informative References 761 [MBMSTS] 3GPP, "Multimedia Broadcast/Multicast Service (MBMS); 762 Protocols and codecs", 3GPP TS 26.346, March 2009, 763 . 765 [RFC5052] Watson, M., Luby, M., and L. Vicisano, "Forward Error 766 Correction (FEC) Building Block", RFC 5052, 767 DOI 10.17487/RFC5052, August 2007, 768 . 770 [RFC6364] Begen, A., "Session Description Protocol Elements for the 771 Forward Error Correction (FEC) Framework", RFC 6364, 772 DOI 10.17487/RFC6364, October 2011, 773 . 775 [RFC6681] Watson, M., Stockhammer, T., and M. Luby, "Raptor Forward 776 Error Correction (FEC) Schemes for FECFRAME", RFC 6681, 777 DOI 10.17487/RFC6681, August 2012, 778 . 780 [RFC6816] Roca, V., Cunche, M., and J. Lacan, "Simple Low-Density 781 Parity Check (LDPC) Staircase Forward Error Correction 782 (FEC) Scheme for FECFRAME", RFC 6816, 783 DOI 10.17487/RFC6816, December 2012, 784 . 786 [RFC6865] Roca, V., Cunche, M., Lacan, J., Bouabdallah, A., and K. 787 Matsuzono, "Simple Reed-Solomon Forward Error Correction 788 (FEC) Scheme for FECFRAME", RFC 6865, 789 DOI 10.17487/RFC6865, February 2013, 790 . 792 [RLC-ID] Roca, V., "Sliding Window Random Linear Code (RLC) Forward 793 Erasure Correction (FEC) Scheme for FECFRAME", Work 794 in Progress, Transport Area Working Group (TSVWG) draft- 795 ietf-tsvwg-rlc-fec-scheme (Work in Progress), March 2018, 796 . 799 Appendix A. About Sliding Encoding Window Management (non Normative) 801 The FEC Framework does not specify the management of the sliding 802 encoding window which is the responsibility of the FEC Scheme. This 803 annex only provides a few non normative hints. 805 Source symbols are added to the sliding encoding window each time a 806 new ADU is available at the sender, after the ADU to source symbol 807 mapping specific to the FEC Scheme. 809 Source symbols are removed from the sliding encoding window, for 810 instance: 812 o after a certain delay, when an "old" ADU of a real-time flow times 813 out. The source symbol retention delay in the sliding encoding 814 window should therefore be initialized according to the real-time 815 features of incoming flow(s); 817 o once the sliding encoding window has reached its maximum size 818 (there is usually an upper limit to the sliding encoding window 819 size). In that case the oldest symbol is removed each time a new 820 source symbol is added. 822 Several considerations can impact the management of this sliding 823 encoding: 825 o at the source flows level: real-time constraints can limit the 826 total time source symbols can remain in the encoding window; 828 o at the FEC code level: theoretical or practical limitations (e.g., 829 because of computational complexity) can limit the number of 830 source symbols in the encoding window; 832 o at the FEC Scheme level: signaling and window management are 833 intrinsically related. For instance, an encoding window composed 834 of a non sequential set of source symbols requires an appropriate 835 signaling to inform a receiver of the composition of the encoding 836 window, and the associated transmission overhead can limit the 837 maximum encoding window size. On the opposite, an encoding window 838 always composed of a sequential set of source symbols simplifies 839 signaling: providing the identity of the first source symbol plus 840 their number is sufficient, which creates a fixed and relatively 841 small transmission overhead. 843 Authors' Addresses 845 Vincent Roca 846 INRIA 847 Grenoble 848 France 850 EMail: vincent.roca@inria.fr 852 Ali Begen 853 Networked Media 854 Konya 855 Turkey 857 EMail: ali.begen@networked.media