idnits 2.17.1 draft-ietf-avt-ports-for-ucast-mcast-rtp-07.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 (December 10, 2010) is 4886 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: 'RFCXXXX' is mentioned on line 1010, but not defined ** Obsolete normative reference: RFC 4566 (Obsoleted by RFC 8866) ** Downref: Normative reference to an Informational RFC: RFC 2104 == Outdated reference: A later version (-04) exists of draft-ietf-avt-rtcp-port-for-ssm-03 == Outdated reference: A later version (-10) exists of draft-ietf-avt-app-rtp-keepalive-09 -- Obsolete informational reference (is this intentional?): RFC 5226 (Obsoleted by RFC 8126) -- Obsolete informational reference (is this intentional?): RFC 2818 (Obsoleted by RFC 9110) Summary: 2 errors (**), 0 flaws (~~), 4 warnings (==), 4 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 AVT A. Begen 3 Internet-Draft D. Wing 4 Intended status: Standards Track Cisco 5 Expires: June 13, 2011 T. VanCaenegem 6 Alcatel-Lucent 7 December 10, 2010 9 Port Mapping Between Unicast and Multicast RTP Sessions 10 draft-ietf-avt-ports-for-ucast-mcast-rtp-07 12 Abstract 14 This document presents a port mapping solution that allows RTP 15 receivers to choose their own ports for an auxiliary unicast session 16 in RTP applications using both unicast and multicast services. The 17 solution provides protection against denial-of-service attacks that 18 could be used to cause one or more RTP packets to be sent to a victim 19 client. 21 Status of this Memo 23 This Internet-Draft is submitted in full conformance with the 24 provisions of BCP 78 and BCP 79. 26 Internet-Drafts are working documents of the Internet Engineering 27 Task Force (IETF). Note that other groups may also distribute 28 working documents as Internet-Drafts. The list of current Internet- 29 Drafts is at http://datatracker.ietf.org/drafts/current/. 31 Internet-Drafts are draft documents valid for a maximum of six months 32 and may be updated, replaced, or obsoleted by other documents at any 33 time. It is inappropriate to use Internet-Drafts as reference 34 material or to cite them other than as "work in progress." 36 This Internet-Draft will expire on June 13, 2011. 38 Copyright Notice 40 Copyright (c) 2010 IETF Trust and the persons identified as the 41 document authors. All rights reserved. 43 This document is subject to BCP 78 and the IETF Trust's Legal 44 Provisions Relating to IETF Documents 45 (http://trustee.ietf.org/license-info) in effect on the date of 46 publication of this document. Please review these documents 47 carefully, as they describe your rights and restrictions with respect 48 to this document. Code Components extracted from this document must 49 include Simplified BSD License text as described in Section 4.e of 50 the Trust Legal Provisions and are provided without warranty as 51 described in the Simplified BSD License. 53 Table of Contents 55 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 56 2. Requirements Notation . . . . . . . . . . . . . . . . . . . . 5 57 3. Token-Based Port Mapping . . . . . . . . . . . . . . . . . . . 6 58 3.1. Token Request and Retrieval . . . . . . . . . . . . . . . 6 59 3.2. Unicast Session Establishment . . . . . . . . . . . . . . 6 60 3.2.1. Motivating Scenario . . . . . . . . . . . . . . . . . 6 61 3.2.2. Normative Behavior and Requirements . . . . . . . . . 9 62 4. Message Formats . . . . . . . . . . . . . . . . . . . . . . . 11 63 4.1. Port Mapping Request . . . . . . . . . . . . . . . . . . . 12 64 4.2. Port Mapping Response . . . . . . . . . . . . . . . . . . 12 65 4.3. Token Verification Request . . . . . . . . . . . . . . . . 14 66 4.4. Token Verification Failure . . . . . . . . . . . . . . . . 15 67 5. Procedures for Token Construction . . . . . . . . . . . . . . 17 68 6. Validating Tokens . . . . . . . . . . . . . . . . . . . . . . 19 69 7. SDP Signaling . . . . . . . . . . . . . . . . . . . . . . . . 20 70 7.1. The portmapping-req Attribute . . . . . . . . . . . . . . 20 71 7.2. Requirements . . . . . . . . . . . . . . . . . . . . . . . 21 72 7.3. Example and Discussion . . . . . . . . . . . . . . . . . . 21 73 8. Address Pooling NATs . . . . . . . . . . . . . . . . . . . . . 24 74 9. Security Considerations . . . . . . . . . . . . . . . . . . . 25 75 9.1. Tokens . . . . . . . . . . . . . . . . . . . . . . . . . . 25 76 9.2. The portmapping-req Attribute . . . . . . . . . . . . . . 25 77 10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 27 78 10.1. Registration of SDP Attributes . . . . . . . . . . . . . . 27 79 10.2. Registration of FMT Values . . . . . . . . . . . . . . . . 27 80 10.3. SFMT Values for Port Mapping Messages Registry . . . . . . 27 81 10.4. RAMS Response Code Space Registry . . . . . . . . . . . . 28 82 11. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 29 83 12. References . . . . . . . . . . . . . . . . . . . . . . . . . . 30 84 12.1. Normative References . . . . . . . . . . . . . . . . . . . 30 85 12.2. Informative References . . . . . . . . . . . . . . . . . . 31 86 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 33 88 1. Introduction 90 In (any-source or source-specific) multicast RTP applications, 91 destination ports, i.e., the ports on which the multicast receivers 92 receive the RTP and RTCP packets, are defined declaratively. In 93 other words, the receivers cannot choose their receive ports and the 94 sender(s) use the pre-defined ports. 96 In unicast RTP applications, the receiving end needs to choose its 97 ports for RTP and RTCP since these ports are local resources and only 98 the receiving end can determine which ports are available to use. In 99 addition, Network Address Port Translators (NAPT - hereafter simply 100 called NAT) devices are commonly deployed in networks, thus, static 101 port assignments cannot be used. The receiving may convey its 102 request to the sending end through different ways, one of which is 103 the Offer/Answer Model [RFC3264] for the Session Description Protocol 104 (SDP) [RFC4566]. However, the Offer/Answer Model requires offer/ 105 answer exchange(s) between the endpoints, and the resulting delay may 106 not be desirable in delay-sensitive real-time applications. 107 Furthermore, the Offer/Answer Model may be burdensome for the 108 endpoints that are concurrently running a large number of unicast 109 sessions with other endpoints. 111 In this specification, we consider an RTP application that uses one 112 or more unicast and multicast RTP sessions together. While the 113 declaration and selection of the ports are well defined and work well 114 for multicast and unicast RTP applications, respectively, the usage 115 of the ports introduces complications when a receiving end mixes 116 unicast and multicast RTP sessions within the same RTP application. 118 An example scenario is where the RTP packets are distributed through 119 source-specific multicast (SSM) and a receiver sends unicast RTCP 120 NACK feedback to a local repair server (also functioning as a unicast 121 RTCP feedback target) [RFC5760] asking for a retransmission of the 122 packets it is missing, and the local repair server sends the 123 retransmission packets over a unicast RTP session [RFC4588]. 125 Another scenario is where a receiver wants to rapidly acquire a new 126 primary multicast RTP session and receives one or more RTP burst 127 packets over a unicast session before joining the SSM session 128 [I-D.ietf-avt-rapid-acquisition-for-rtp]. Similar scenarios exist in 129 applications where some part of the content is distributed through 130 multicast while the receivers get additional and/or auxiliary content 131 through one or more unicast connections, as sketched in Figure 1. 133 In this document, we discuss this problem and introduce a solution 134 that we refer to as Port Mapping. This solution allows receivers to 135 choose their desired UDP ports for RTP and RTCP in every unicast 136 session when they are running RTP applications using both unicast and 137 multicast services, and offer/answer exchange is not available. This 138 solution is not applicable in cases where TCP is used as the 139 transport protocol in the unicast sessions. For such scenarios, 140 refer to [RFC4145]. 142 ----------- 143 | Unicast |................ 144 | Source |............. : 145 | (Server) | : : 146 ----------- : : 147 v v 148 ----------- ---------- ----------- 149 | Multicast |------->| Router |---------->|Client RTP | 150 | Source | | |..........>|Application| 151 ----------- ---------- ----------- 152 | : 153 | : ----------- 154 | :..............>|Client RTP | 155 +---------------->|Application| 156 ----------- 158 -------> Multicast RTP Flow 159 .......> Unicast RTP Flow 161 Figure 1: RTP applications simultaneously using both unicast and 162 multicast services 164 In the remainder of this document, we refer to the RTP endpoints that 165 serve other RTP endpoints over a unicast session as the Servers. The 166 receiving RTP endpoints are referred to as Clients. 168 2. Requirements Notation 170 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 171 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 172 document are to be interpreted as described in [RFC2119]. 174 3. Token-Based Port Mapping 176 Token-based Port Mapping consists of two steps: (i) Token request 177 and retrieval, and (ii) unicast session establishment. These are 178 described below. 180 3.1. Token Request and Retrieval 182 This first step is required to be completed only once. Once a Token 183 is retrieved from a particular server, it can be used for all the 184 unicast sessions the client will be running with this particular 185 server. By default, Tokens are server specific. However, the client 186 can use the same Token to communicate with different servers if these 187 servers are provided with the same secret key and algorithm used to 188 generate the Token and are at least loosely clock-synchronized. The 189 Token becomes invalid if client's public IP address changes or when 190 the server expires the Token. In these cases, the client has to 191 request a new Token. 193 The Token is essentially an opaque encapsulation that is based on 194 client's IP address (as seen by the server). When a request is 195 received, the server creates a Token for this particular client, and 196 sends it back to the client. Later, when the client wants to 197 establish a unicast session, the Token will be validated by the 198 server, making sure that the IP address information matches. This is 199 effective against DoS attacks, e.g., an attacker cannot simply spoof 200 another client's IP address and start a unicast transmission towards 201 random clients. 203 3.2. Unicast Session Establishment 205 The second step is the unicast session establishment. We illustrate 206 this step with an example. First, we describe the motivation 207 scenario and then define the normative behavior and requirements. 209 3.2.1. Motivating Scenario 211 Consider an SSM distribution network where a distribution source 212 multicasts RTP packets to a large number of clients, and one or more 213 retransmission servers function as feedback targets to collect 214 unicast RTCP feedback from these clients [RFC5760]. The 215 retransmission servers also join the multicast session to receive the 216 multicast packets and cache them for a certain time period. When a 217 client detects missing packets in the multicast session, it requests 218 a retransmission from one of the retransmission servers by using an 219 RTCP NACK message [RFC4585]. The retransmission server pulls the 220 requested packet(s) out of the cache and retransmits them to the 221 requesting client [RFC4588]. 223 The RTP and RTCP flows pertaining to the scenario described above are 224 sketched in Figure 2. Between the client and server, there can be 225 one or more NAT devices [RFC4787]. 227 -------------- --- ---------- 228 | |-------------------------------| |-->|P1 | 229 | |-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-| |.->|P2 | 230 | | | | | | 231 | Distribution | ---------------- | | | | 232 | Source | | | | | | | 233 | |---->|P1 | | | | | 234 | |.-.->|P2 | | | | | 235 | | | | | | | | 236 -------------- | P3|<.=.=.=.| |=.=|*c0 | 237 | P3|<~~~~~~~| |~~~|*c1 | 238 MULTICAST RTP | | | | | | 239 SESSION with | | | | | | 240 UNICAST FEEDBACK | | | N | | | 241 | Retransmission | | A | | Client | 242 - - - - - - - - - - -| - - - - - - - -| - - - -| - |- -| - - - - -|- 243 | Server | | T | | | 244 | | | | | | 245 PORT MAPPING | PT|<~~~~~~~| |~~>|*cT | 246 | | | | | | 247 - - - - - - - - - - -| - - - - - - - -| - - - -| - |- -| - - - - -|- 248 | | | | | | 249 AUXILIARY UNICAST | | | | | | 250 RTP SESSION | | | | | | 251 | P3|........| |..>|*c1 | 252 | P3|=.=.=.=.| |=.>|*c1 | 253 | P4|<.=.=.=.| |=.=|*c2 | 254 | | | | | | 255 ---------------- --- ---------- 257 -------> Multicast RTP Flow 258 .-.-.-.> Multicast RTCP Flow 259 .=.=.=.> Unicast RTCP Reports 260 ~~~~~~~> Unicast RTCP Feedback Messages 261 .......> Unicast RTP Flow 263 Figure 2: Example scenario showing an SSM distribution with support 264 for retransmissions from a server 266 In Figure 2, we have the following multicast and unicast ports: 268 o Ports P1 and P2 denote the destination RTP and RTCP ports in the 269 multicast session, respectively. The clients listen to these 270 ports to receive the multicast RTP and RTCP packets. Ports P1 and 271 P2 are defined declaratively. 273 o Port P3 denotes the RTCP port on the feedback target running on 274 the retransmission server to collect any RTCP packet sent by the 275 clients including feedback messages, and RTCP receiver and 276 extended reports. This is also the port that the retransmission 277 server uses to send the RTP packets and RTCP sender reports in the 278 unicast session. Port P3 is defined declaratively. 280 o Port P4 denotes the RTCP port on the retransmission server used to 281 collect the RTCP receiver and extended reports for the unicast 282 session. Port P4 is defined declaratively. 284 o Ports *c0, *c1 and *c2 are chosen by the client. *c0 denotes the 285 port on the client used to send the RTCP reports for the multicast 286 session. *c1 denotes the port on the client used to send the 287 unicast RTCP feedback messages in the multicast session and to 288 receive the RTP packets and RTCP sender reports in the unicast 289 session. *c2 denotes the port on the client used to send the RTCP 290 receiver and extended reports in the unicast session. Ports c0, 291 c1 and c2 could be the same port or different ports. There are 292 two advantages of using the same port for both c0 and c1: 294 1. Some NATs only keep bindings active when a packet goes from 295 the inside to the outside of the NAT (See REQ-6 of Section 4.3 296 of [RFC4787]). When the gap between the packets sent from the 297 client to the server is long, this can exceed that timeout. 298 If c0=c1, the occasional (periodic) RTCP receiver reports sent 299 from port c0 (for the multicast session's RTCP port P3) will 300 ensure the NAT does not time out the public port associated 301 with the incoming unicast traffic to port c1. 303 2. Having c0=c1 conserves NAT port bindings. 305 o Ports PT and *cT denote the ports through which the Token request 306 and retrieval occur at the server and client sides, respectively. 307 Port PT is declared on a per unicast session basis, although the 308 same port could be used for two or more unicast sessions sourced 309 by the server. A Token once requested and retrieved by a client 310 from port PT remains valid until its expiration time. 312 We assume that the information declaratively defined is available as 313 part of the session description information and is provided to the 314 clients. The Session Description Protocol (SDP) [RFC4566] and other 315 session description methods can be used for this purpose. 317 3.2.2. Normative Behavior and Requirements 319 In this section, we describe the normative behavior and requirements. 320 To simplify the presentation, we refer to the port numbers described 321 in the example presented in Figure 2. However, the behavior and 322 requirements described here are not specific to that particular 323 example. 325 The following steps summarize the Token-based solution: 327 1. The client ascertains server address and port numbers (P3, P4 and 328 PT) from the session description. Port P4 MUST be different from 329 port P3. Port PT MAY be equal to port P3. 331 2. The client selects its local port numbers (*c0, *c1, *c2 and 332 *cT). It is strongly RECOMMENDED that the client uses the same 333 port for c0 and c1. Port cT MAY be equal to ports c0 and c1. 335 A client cannot keep using the same receive port for different 336 unicast sessions since there could be packet leakage when 337 switching from one unicast session to another unless each 338 received unicast stream has its own distinct Synchronization 339 Source (SSRC) identifier to allow the client to filter out the 340 undesired packets. Unless this is guaranteed (which is not often 341 easy), a client SHOULD use separate receive ports for subsequent 342 unicast sessions. After a sufficient time, a previously used 343 receive port could be used again. 345 3. If the client does not have a Token (or the existing Token has 346 expired): 348 A. The client first sends a message to the server via a new RTCP 349 message, called Port Mapping Request to port PT. This 350 message is sent from port *cT on the client side. The server 351 learns client's public IP address from the received message. 352 The client can send this message anytime it wants (e.g., 353 during initialization), and does not normally ever need to 354 re-send this message (See Section 6). 356 B. The server generates an opaque encapsulation (i.e., the 357 Token) based on certain information including client's IP 358 address. 360 C. The server sends the Token back to the client using a new 361 RTCP message, called Port Mapping Response. This message 362 MUST be sent from port PT to port cT. 364 4. The client needs to provide the Token to the server using a new 365 RTCP message, called Token Verification Request, whenever the 366 client sends an RTCP feedback message for triggering or 367 controlling a unicast session (See Section 4.3). Note that the 368 unicast session is only established after the server has received 369 a feedback message (along with a valid Token) from the client for 370 which it needs to react by sending unicast data. Until a unicast 371 session is established, neither the server nor the client needs 372 to send RTCP reports for the unicast session. 374 5. Normal flows ensue as shown in Figure 2. Note that in the 375 unicast session, traffic from the server to the client (i.e., 376 both the RTP and RTCP packets sent from port P3 to port c1) MUST 377 be multiplexed on the (same) port c1. If the client uses the 378 same port for both c0 and c1, the RTCP reports sent for the 379 multicast session keep the P3->c1(=c0) binding alive. If the 380 client uses different ports for c0 and c1, the client needs to 381 periodically send an explicit keep-alive message 382 [I-D.ietf-avt-app-rtp-keepalive] to keep the P3->c1 binding alive 383 during the lifetime of the unicast session if the unicast 384 session's lifetime is likely to exceed the NAT's timeout value. 386 4. Message Formats 388 This section defines the formats of the RTCP transport-layer feedback 389 messages that are exchanged between a server and a client for the 390 purpose of Token-based port mapping. Four RTCP messages are defined: 392 1. Port Mapping Request 394 2. Port Mapping Response 396 3. Token Verification Request 398 4. Token Verification Failure 400 These are all payload-independent RTCP feedback messages with a 401 common format defined in Section 6.1 of [RFC4585], also sketched in 402 Figure 3. 404 0 1 2 3 405 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 1 406 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 407 |V=2|P| FMT | PT | length | 408 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 409 | SSRC of packet sender | 410 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 411 | SSRC of media source | 412 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 413 : Feedback Control Information (FCI) : 414 : : 416 Figure 3: The common packet format for the RTCP feedback messages 418 Each feedback message has a fixed-length field for version, padding, 419 feedback message type (FMT), packet type (PT), length, SSRC of packet 420 sender, SSRC of media source as well as a variable-length field for 421 feedback control information (FCI). 423 In the new messages defined in this section, the PT field is set to 424 RTPFB (205) and the FMT field is set to Port Mapping (7). Individual 425 Port Mapping messages are identified by a sub-field called Sub 426 Feedback Message Type (SFMT). Any Reserved field SHALL be set to 427 zero and ignored. 429 Following the rules specified in [RFC3550], all integer fields in the 430 messages defined below are carried in network-byte order, that is, 431 most significant byte (octet) first, also known as big-endian. 432 Unless otherwise stated, numeric constants are in decimal (base 10). 434 Note that RTCP is not a timely or reliable protocol. The RTCP 435 packets might get lost or re-ordered in the network. When a client 436 sends a Port Mapping Request or Token Verification Request message 437 but it does not receive a response back from the server (either a 438 Port Mapping Response or Token Verification Failure message), it MAY 439 resend its request when it is eligible to do so based on the timer 440 rules defined in [RFC4585]. 442 4.1. Port Mapping Request 444 The Port Mapping Request message is identified by SFMT=1. This 445 message is a unicast feedback message transmitted by the client to a 446 dedicated server port to request a Token. In the Port Mapping 447 Request message, the client MUST set both the packet sender SSRC and 448 media source SSRC fields to its own SSRC since the Port Mapping 449 Request message is not necessarily linked to any specific media 450 source. The FCI field has the structure depicted in Figure 4. 452 0 1 2 3 453 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 1 454 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 455 | SFMT=1 | Random Nonce : 456 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 457 : Random Nonce | 458 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 460 Figure 4: The FCI field of Port Mapping Request message 462 o Random Nonce (56 bits): Mandatory field that contains a random 463 nonce value generated by the client following the procedures of 464 [RFC4086]. This nonce is taken into account by the server when 465 generating a Token for the client to enable better security for 466 clients that share the same IP address. If the Port Mapping 467 Request message is transmitted multiple times for redundancy 468 reasons, the random nonce value MUST remain the same in these 469 duplicated messages. However, the client MUST generate a new 470 random nonce for every new Port Mapping Request message. 472 4.2. Port Mapping Response 474 The Port Mapping Response message is identified by SFMT=2. This 475 message is sent by the server and delivers the Token to the client as 476 a response to the Port Mapping Request message. In the Port Mapping 477 Response message, the packet sender SSRC and media sender SSRC fields 478 are both set to the client's SSRC since the Port Mapping Response 479 message is not necessarily linked to any specific media source. The 480 FCI field has the structure depicted in Figure 5. 482 0 1 2 3 483 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 1 484 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 485 | SFMT=2 | Associated Nonce : 486 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 487 : Associated Nonce | 488 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 489 : Token Element : 490 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 491 | Absolute | 492 | Expiration Time | 493 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 494 | Relative Expiration Time | 495 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 497 Figure 5: FCI field syntax for the Port Mapping Response message 499 o Associated Nonce (56 bits): Mandatory field that contains the 500 nonce received in the Port Mapping Request message and used in 501 Token construction. 503 o Token Element (Variable size): Mandatory element that is used to 504 carry the Token generated by the server. This element is a 505 Length-Value element. The Length field, which is 8 bits, 506 indicates the length (in octets) of the Value field that follows 507 the Length field. The Value field carries the Token (or more 508 accurately, the output of the encoding process on the server). 510 o Absolute Expiration Time (64 bits): Mandatory field that contains 511 the absolute expiration time of the Token. The absolute 512 expiration time is expressed as a Network Time Protocol (NTP) 513 timestamp value in seconds since year 1900 [RFC5905]. The client 514 does not need to use this element directly, thus, does not need to 515 synchronize its clock with the server. However, the client needs 516 to send this element back to the server along with the associated 517 nonce in the Token Verification Request message, thus, needs to 518 keep it associated with the Token. 520 o Relative Expiration Time (32 bits): Mandatory field that contains 521 the relative expiration time of the Token. The relative 522 expiration time is expressed in seconds from the time the Token 523 was generated. A relative expiration time of zero indicates that 524 the accompanying Token is not valid. 526 The server conveys the relative expiration time in the clear to 527 the client to allow the client to request a new Token well before 528 the expiration time. 530 4.3. Token Verification Request 532 The Token Verification Request message is identified by SFMT=3. This 533 message contains the Token and accompanies any RTCP message that 534 would trigger a new or control an existing unicast session. 535 Currently, the following RTCP messages are REQUIRED to be accompanied 536 by a Token Verification Request message: 538 o Messages that trigger a new unicast session: 540 * NACK messages [RFC4585] 542 * RAMS-R messages [I-D.ietf-avt-rapid-acquisition-for-rtp] 544 o Messages that control an existing unicast session associated with 545 a multicast session: 547 * BYE messages [RFC3550] 549 * RAMS-T messages [I-D.ietf-avt-rapid-acquisition-for-rtp] 551 * CCM messages [RFC5104] 553 Other RTCP messages defined in the future, which could be abused to 554 cause packet amplification attacks, SHOULD also be authenticated 555 using the mechanism described in this document. The Token 556 Verification Request message might also be bundled with packets 557 carrying RTCP receiver or extended reports. While such packets do 558 not have a strong security impact, a specific application might 559 desire to have a more controlled reporting scheme from the clients. 561 In the Token Verification Request message, the client MUST set both 562 the packet sender SSRC and media source SSRC fields to its own SSRC 563 since the media source SSRC may not be known. The client MUST NOT 564 send a Token Verification Request message with a Token that has 565 expired. The FCI field has the structure depicted in Figure 6. 567 0 1 2 3 568 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 1 569 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 570 | SFMT=3 | Associated Nonce : 571 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 572 : Associated Nonce | 573 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 574 : Token Element : 575 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 576 | Associated Absolute | 577 | Expiration Time | 578 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 580 Figure 6: FCI field syntax for the Token Verification message 582 o Associated Nonce (56 bits): Mandatory field that contains the 583 nonce associated with the Token above. 585 o Token Element (Variable size): Mandatory Token element that was 586 previously received in the Port Mapping Response message. 588 o Associated Absolute Expiration Time (64 bits): Mandatory field 589 that contains the absolute expiration time associated with the 590 Token above. 592 4.4. Token Verification Failure 594 The Token Verification Failure message is identified by SFMT=4. This 595 message is sent by the server and notifies the client that the Token 596 was invalid or that the client did not include a Token Verification 597 Request message in the RTCP packet although it was supposed to. In 598 the Token Verification Failure message, the packet sender SSRC and 599 media sender SSRC fields are both set to the client's SSRC. The FCI 600 field has the structure depicted in Figure 6. 602 0 1 2 3 603 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 1 604 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 605 | SFMT=4 | Associated Nonce : 606 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 607 : Associated Nonce | 608 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 610 Figure 7: FCI field syntax for the Token Failure message 612 o Associated Nonce (56 bits): Mandatory field that contains the 613 nonce received in the Token Verification Request message. If 614 there was no Token Verification Request message included by the 615 client, this field is set to 0. 617 5. Procedures for Token Construction 619 The Token encoding is known to the server but opaque to the client. 620 Implementations MUST encode the following information into the Token 621 as a minimum, in order to provide adequate security: 623 o Client's IP address as seen by the server (32/128 bits for IPv4/ 624 IPv6 addresses) 626 o The nonce generated and inserted in the Port Mapping Request 627 message by the client (56 bits) 629 o The absolute expiration time chosen by the server indicated as an 630 NTP timestamp value in seconds since year 1900 [RFC5905] (64 bits, 631 to protect against replay attacks) 633 An example way for constructing Tokens is to perform HMAC-SHA1 634 [RFC2104] on the concatenated values of the information listed above. 635 The HMAC key should be at least 160 bits long, and generated using a 636 cryptographically secure random source [RFC4086]. However, 637 implementations MAY adopt different approaches and are encouraged to 638 encode whatever additional information is deemed necessary or useful. 639 For example, key rollover is simplified by encoding a key-id into the 640 Token. As another example, a cluster of anycast servers could find 641 advantage by encoding a server identifier into the Token. As another 642 example, if HMAC-SHA1 has been compromised, a replacement HMAC 643 algorithm could be used instead (e.g., HMAC-SHA256). 645 To protect from offline attacks, the server SHOULD occasionally 646 choose a new HMAC key. To ease implementation, a key-id can be 647 assigned to each HMAC key. This can be encoded as simply as one bit 648 (where the new key is X (e.g., 1) and the old key is the inverted 649 value of X (e.g., 0)), or if several keys are supported at once could 650 be encoded into several bits. As the encoding of the Token is 651 entirely private to the server and opaque to the clients, any 652 encoding can be used. By encoding the key-id into the Token element, 653 the server can reject an old key without bothering to do HMAC 654 validation (saving CPU cycles). The key-id can be encoded into the 655 Value field of the Token element by simply concatenating the 656 (plaintext) key-id with the hashed information (i.e., the Token 657 itself). 659 For example, the Value field in the Token element can be computed as: 661 key-id || hash-alg (client-ip | nonce | abs-expiration) 663 During Token construction, the expiration time has to be chosen 664 carefully based on the intended service duration. Tokens that are 665 valid for an unnecessarily long period of time (e.g., several hours) 666 might impose security risks. Depending on the application and use 667 cases, a reasonable value needs to be chosen by the server. Note 668 that using shorter lifetimes requires the clients to acquire Tokens 669 more frequently. However, since a client can acquire a new Token 670 well before it will need to use it, the client will not necessarily 671 be penalized for the acquisition delay. 673 Finally, be aware that NTP timestamps will wrap around in year 2036 674 and implementations might need to handle this eventually. Refer to 675 Section 6 of [RFC5905] for further details. 677 6. Validating Tokens 679 Upon receipt of an RTCP feedback message along with the Token 680 Verification Request message that contains a Token, nonce and 681 absolute expiration time, the server MUST validate the Token. 683 The server first applies the its own procedure for constructing the 684 Tokens by using client's IP address from the received Token 685 Verification Request message, and the nonce and absolute expiration 686 time values reported in the received Token Verification Request 687 message. The server then compares the resulting output with the 688 Token sent by the client in the Token Verification Request message. 689 If they match and the absolute expiration time has not passed yet, 690 the server declares that the Token is valid. 692 Note that if the client's IP address changes, the Token will not 693 validate. Similarly, if the client inserts an incorrect nonce or 694 absolute expiration time value in the Token Verification Request 695 message, validation will fail. It is also possible that the server 696 wants to expire the Token prematurely. In these cases, the server 697 MUST reply back to the client with a Token Verification Failure 698 message (that goes from port P3 on the server to port c1 on the 699 client). 701 In addition to the Token Verification Failure message, it is 702 RECOMMENDED that applications define an application-specific error 703 response to be sent by the server when the server detects that the 704 Token is invalid. For applications using 705 [I-D.ietf-avt-rapid-acquisition-for-rtp], this document defines a new 706 4xx-level response code in the RAMS Response Code Space Registry. A 707 client that received a Token Verification Failure message can request 708 a new Token from the server. 710 7. SDP Signaling 712 7.1. The portmapping-req Attribute 714 This new SDP attribute is used declaratively to indicate the port and 715 optionally the address for obtaining a Token. Its presence indicates 716 that (i) a Token MUST be included in the feedback messages sent to 717 the server triggering or controlling a unicast session (See 718 Section 4.3 for details), and (ii) the client MUST receive the 719 unicast session's RTP and RTCP packets from the server on the port 720 from which it sent the RTCP message triggering or controlling the 721 unicast session. 723 The formal description of the 'portmapping-req' attribute is defined 724 by the following ABNF [RFC5234] syntax: 726 portmapping-req-attribute = "a=portmapping-req:" [port [nettype space 727 addrtype space connection-address]] CRLF 729 Here, 'port', 'nettype', 'addrtype' and 'connection-address' are 730 defined as specified in Section 9 of [RFC4566]. The 'portmapping- 731 req' attribute is used as a session-level or media-level attribute. 732 If used at a media level, the attribute MUST be used in a unicast 733 media block. 735 Note: This does not imply that Token Verification Request 736 messages need to be sent in the unicast session. Token 737 Verification Request messages accompany RTCP messages that trigger 738 or control this unicast session, and are sent either in the 739 multicast session or the unicast session, depending on the RTCP 740 message (See Section 4.3). 742 In the optional address value, only unicast addresses are allowed; 743 multicast addresses SHOULD NOT be used without evaluating the 744 additional security risks such as non-legit servers generating fake 745 Tokens. If the address is not specified, the (source) address in the 746 "c" line corresponding to the unicast media stream is implied. 748 When using this SDP attribute in SDP offer/answer [RFC3264], the 749 following needs to be considered. This attribute is used 750 declaratively. If included at session level, this applies to all 751 media lines that uses RTP. If included at media level, it applies to 752 the RTCP feedback messages declared by this media block. 754 An offerer that desires the answerer to use Tokens in any RTCP 755 message sent to the offerer, i.e., received by the offerer, the 756 attribute is included. In case an offerer desires to declare support 757 for using Tokens as defined in this specification but do not need 758 Tokens to be included for any RTCP messages to be received by the 759 offerer, it can include the 'portmapping-req' attribute without any 760 parameters, neither port nor address, either at a session or media 761 level. 763 An answerer receiving an SDP offer with the "a=portmapping-req" line 764 with a port number SHALL use that port number and the address, either 765 explicitly provided in the attribute or implicitly provided by the 766 "c" line, for any needed Token request. If the "a=portmapping-req" 767 line attribute does not contain a port, the answer SHALL take note of 768 the capability. 770 When sending an answer, if the 'portmapping-req' attribute has been 771 present in the offer including a port number and the answerer 772 supports this specification, then the answerer MUST include the 773 attribute in its answer. The answer may or may not include a port 774 and address. This depends on the application and the desire of the 775 answerer. The answerer includes a port and possibly an address when 776 it requires to receive Tokens in RTCP messages. If it only supports 777 this specification but does not need Tokens to be included, the 778 attribute is included without any port or address. 780 7.2. Requirements 782 The use of SDP for the port mapping solution normatively requires the 783 support for: 785 o The SDP grouping framework and flow identification (FID) semantics 786 [RFC5888] 788 o The RTP/AVPF profile [RFC4585] 790 o The RTCP extensions for SSM sessions with unicast feedback 791 [RFC5760] 793 o The 'multicast-rtcp' attribute [I-D.ietf-avt-rtcp-port-for-ssm] 795 o Multiplexing RTP and RTCP on a single port on both endpoints in 796 the unicast session [RFC5761] 798 7.3. Example and Discussion 800 The declarative SDP describing the scenario given in Figure 2 is 801 written as: 803 v=0 804 o=ali 1122334455 1122334466 IN IP4 nack.example.com 805 s=Local Retransmissions 806 t=0 0 807 a=group:FID 1 2 808 a=rtcp-unicast:rsi 809 m=video 41000 RTP/AVPF 98 810 i=Multicast Stream 811 c=IN IP4 233.252.0.2/255 812 a=source-filter:incl IN IP4 233.252.0.2 198.51.100.1 ; Note 1 813 a=rtpmap:98 MP2T/90000 814 a=multicast-rtcp:41500 ; Note 1 815 a=rtcp:42000 IN IP4 192.0.2.1 ; Note 2 816 a=rtcp-fb:98 nack ; Note 2 817 a=mid:1 818 m=video 42000 RTP/AVPF 99 ; Note 3 819 i=Unicast Retransmission Stream 820 c=IN IP4 192.0.2.1 821 a=sendonly 822 a=rtpmap:99 rtx/90000 823 a=rtcp-mux ; Note 4 824 a=rtcp:42500 ; Note 5 825 a=fmtp:99 apt=98; rtx-time=5000 826 a=portmapping-req:30000 ; Note 6 827 a=mid:2 829 Figure 8: SDP describing an SSM distribution with support for 830 retransmissions from a local server 832 In this description, we highlight the following notes: 834 Note 1: The source stream is multicast from a distribution source 835 with a source IP address of 198.51.100.1 to the multicast destination 836 address of 233.252.0.2 and port 41000 (P1). The associated RTCP 837 packets are multicast in the same group to port 41500 (P2). 839 Note 2: A retransmission server including feedback target 840 functionality with an IP address of 192.0.2.1 and port of 42000 (P3) 841 is specified with the 'rtcp' attribute. The feedback functionality 842 is enabled for the RTP stream with payload type 98 through the 843 'rtcp-fb' attribute [RFC4585]. 845 Note 3: The port specified in the second "m" line (for the unicast 846 stream) does not mean anything in this scenario as the client does 847 not send any RTP traffic back to the server. 849 Note 4: The server multiplexes RTP and RTCP packets on the same port 850 (c1 in Figure 2). 852 Note 5: The server uses port 42500 (P4) for the unicast sessions. 854 Note 6: The "a=portmapping-req" line indicates that a Token needs to 855 be retrieved first before a unicast session associated to the 856 multicast session can be established and that the Port Mapping 857 Request message needs to be sent to port 30000 (PT). Since there is 858 no address indiciated in this line, the client needs to retrieve the 859 Token from the address specified in the "c" line. 861 8. Address Pooling NATs 863 Large-scale NAT devices have a pool of public IPv4 addresses and map 864 internal hosts to one of those public IPv4 addresses. As long as an 865 internal host maintains an active mapping in the NAT, the same IPv4 866 address is assigned to new connections. However, once all of the 867 host's mappings have been deleted (e.g., because of timeout), it is 868 possible that a new connection from that same host will be assigned a 869 different IPv4 address from the pool. When that occurs, the Token 870 will be considered invalid by the server, causing an additional round 871 trip for the client to acquire a fresh Token. 873 Any traffic from the host which traverses the NAT will prevent this 874 problem. As the host is sending RTCP receiver reports at least every 875 5 seconds (Section 6.2 of [RFC3550]) for the multicast session it is 876 receiving, those RTCP messages will be sufficient to prevent this 877 problem. 879 9. Security Considerations 881 9.1. Tokens 883 The Token, which is generated based on a client's IP address and 884 expiration date, provides protection against denial-of-service (DoS) 885 attacks. An attacker using a certain IP address cannot cause one or 886 more RTP packets to be sent to a victim client who has a different IP 887 address. However, if the attacker acquires a valid Token for a 888 victim and can spoof the victim's source address, this approach 889 becomes vulnerable to replay attacks. This is especially easy if the 890 attacker and victim are behind a large-scale NAT and share the same 891 IP address. 893 Multicast is deployed on managed networks - not the Internet. These 894 managed networks will choose to enable network ingress filtering 895 [RFC2827] or not. If ingress filtering is enabled on a network, an 896 attacker attacker cannot spoof a victim's IP address to use a Token 897 to initiate an attack against a victim. However, if ingress 898 filtering is not enabled on a network, an attacker could obtain a 899 Token and spoof the victim's address, causing traffic to flood the 900 victim. On such a network, the server can reduce the time period for 901 such an attack by expiring a Token in a short period of time. In the 902 extreme case, the server can expire the Token in such a short period 903 of time, such that the client will have to acquire a new Token 904 immediately before using it in a Token Verification Request message. 906 HMAC-SHA1 provides a level of security that is widely regarded as 907 being more than sufficient for providing message authentication. It 908 is believed that the economic cost of breaking that algorithm is 909 significantly higher than the cost of more direct approaches to 910 violating system security, e.g., theft, bribery, wiretapping, and 911 other forms of malfeasance. HMAC-SHA1 is secure against all known 912 cryptanalytic attacks that use computational resources that are 913 currently economically feasible. 915 9.2. The portmapping-req Attribute 917 The 'portmapping-req' attribute is not believed to introduce any 918 significant security risk to multimedia applications. A malevolent 919 third party could use this attribute to redirect the Port Mapping 920 Request messages by altering the port number or cause the unicast 921 session establishment to fail by removing it from the SDP 922 description. But, this requires intercepting and rewriting the 923 packets carrying the SDP description; and if an interceptor can do 924 that, many more attacks are possible, including a wholesale change of 925 the addresses and port numbers at which the media will be sent. 927 In order to avoid attacks of this sort, the SDP description needs to 928 be integrity protected and provided with source authentication. This 929 can, for example, be achieved on an end-to-end basis using S/MIME 930 [RFC5652] when SDP is used in a signaling packet using MIME types 931 (application/sdp). Alternatively, HTTPS [RFC2818] or the 932 authentication method in the Session Announcement Protocol (SAP) 933 [RFC2974] could be used as well. 935 10. IANA Considerations 937 The following contact information shall be used for all registrations 938 in this document: 940 Ali Begen 941 abegen@cisco.com 943 Note to the RFC Editor: In the following, please replace "XXXX" with 944 the number of this document prior to publication as an RFC. 946 10.1. Registration of SDP Attributes 948 This document registers a new attribute name in SDP. 950 SDP Attribute ("att-field"): 951 Attribute name: portmapping-req 952 Long form: Port for requesting Token 953 Type of name: att-field 954 Type of attribute: Either session or media level 955 Subject to charset: No 956 Purpose: See this document 957 Reference: [RFCXXXX] 958 Values: See this document 960 10.2. Registration of FMT Values 962 Within the RTPFB range, the following format (FMT) value is 963 registered: 965 Name: Port Mapping 966 Long name: Port Mapping Between Unicast and Multicast RTP Sessions 967 Value: 7 968 Reference: [RFCXXXX] 970 10.3. SFMT Values for Port Mapping Messages Registry 972 This document creates a new sub-registry for the sub-feedback message 973 type (SFMT) values to be used with the FMT value registered for Port 974 Mapping messages. The registry is called the SFMT Values for Port 975 Mapping Messages Registry. This registry is to be managed by the 976 IANA according to the Specification Required policy of [RFC5226]. 978 The length of the SFMT field in the Port Mapping messages is a single 979 octet, allowing 256 values. The registry is initialized with the 980 following entries: 982 Value Name Reference 983 ----- -------------------------------------------------- ------------- 984 0 Reserved [RFCXXXX] 985 1 Port Mapping Request [RFCXXXX] 986 2 Port Mapping Response [RFCXXXX] 987 3 Token Verification Request [RFCXXXX] 988 4 Token Verification Failure [RFCXXXX] 989 5-254 Assignable - Specification Required 990 255 Reserved [RFCXXXX] 992 The SFMT values 0 and 255 are reserved for future use. 994 Any registration for an unassigned SFMT value needs to contain the 995 following information: 997 o Contact information of the one doing the registration, including 998 at least name, address, and email. 1000 o A detailed description of what the new SFMT represents and how it 1001 shall be interpreted. 1003 10.4. RAMS Response Code Space Registry 1005 This document adds the following entry to the RAMS Response Code 1006 Space Registry. 1008 Code Description Reference 1009 ----- -------------------------------------------------- ------------- 1010 405 Invalid Token [RFCXXXX] 1012 This response code is used when the Token included by the RTP_Rx in 1013 the RAMS-R message is invalid. 1015 11. Acknowledgments 1017 The approach presented in this document came out after discussions 1018 with various individuals in the AVT and MMUSIC WGs, and the breakout 1019 session held in the Anaheim meeting. We thank each of these 1020 individuals, in particular to Magnus Westerlund and Colin Perkins. 1022 12. References 1024 12.1. Normative References 1026 [RFC3550] Schulzrinne, H., Casner, S., Frederick, R., and V. 1027 Jacobson, "RTP: A Transport Protocol for Real-Time 1028 Applications", STD 64, RFC 3550, July 2003. 1030 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1031 Requirement Levels", BCP 14, RFC 2119, March 1997. 1033 [RFC4566] Handley, M., Jacobson, V., and C. Perkins, "SDP: Session 1034 Description Protocol", RFC 4566, July 2006. 1036 [RFC4585] Ott, J., Wenger, S., Sato, N., Burmeister, C., and J. Rey, 1037 "Extended RTP Profile for Real-time Transport Control 1038 Protocol (RTCP)-Based Feedback (RTP/AVPF)", RFC 4585, 1039 July 2006. 1041 [RFC5760] Ott, J., Chesterfield, J., and E. Schooler, "RTP Control 1042 Protocol (RTCP) Extensions for Single-Source Multicast 1043 Sessions with Unicast Feedback", RFC 5760, February 2010. 1045 [RFC5234] Crocker, D. and P. Overell, "Augmented BNF for Syntax 1046 Specifications: ABNF", STD 68, RFC 5234, January 2008. 1048 [RFC4086] Eastlake, D., Schiller, J., and S. Crocker, "Randomness 1049 Requirements for Security", BCP 106, RFC 4086, June 2005. 1051 [RFC5905] Mills, D., Martin, J., Burbank, J., and W. Kasch, "Network 1052 Time Protocol Version 4: Protocol and Algorithms 1053 Specification", RFC 5905, June 2010. 1055 [RFC2104] Krawczyk, H., Bellare, M., and R. Canetti, "HMAC: Keyed- 1056 Hashing for Message Authentication", RFC 2104, 1057 February 1997. 1059 [I-D.ietf-avt-rtcp-port-for-ssm] 1060 Begen, A., "RTP Control Protocol (RTCP) Port for Source- 1061 Specific Multicast (SSM) Sessions", 1062 draft-ietf-avt-rtcp-port-for-ssm-03 (work in progress), 1063 October 2010. 1065 [RFC5888] Camarillo, G. and H. Schulzrinne, "The Session Description 1066 Protocol (SDP) Grouping Framework", RFC 5888, June 2010. 1068 [RFC5761] Perkins, C. and M. Westerlund, "Multiplexing RTP Data and 1069 Control Packets on a Single Port", RFC 5761, April 2010. 1071 12.2. Informative References 1073 [RFC3264] Rosenberg, J. and H. Schulzrinne, "An Offer/Answer Model 1074 with Session Description Protocol (SDP)", RFC 3264, 1075 June 2002. 1077 [RFC4145] Yon, D. and G. Camarillo, "TCP-Based Media Transport in 1078 the Session Description Protocol (SDP)", RFC 4145, 1079 September 2005. 1081 [I-D.ietf-avt-rapid-acquisition-for-rtp] 1082 Steeg, B., Begen, A., Caenegem, T., and Z. Vax, "Unicast- 1083 Based Rapid Acquisition of Multicast RTP Sessions", 1084 draft-ietf-avt-rapid-acquisition-for-rtp-17 (work in 1085 progress), November 2010. 1087 [RFC4787] Audet, F. and C. Jennings, "Network Address Translation 1088 (NAT) Behavioral Requirements for Unicast UDP", BCP 127, 1089 RFC 4787, January 2007. 1091 [RFC4588] Rey, J., Leon, D., Miyazaki, A., Varsa, V., and R. 1092 Hakenberg, "RTP Retransmission Payload Format", RFC 4588, 1093 July 2006. 1095 [I-D.ietf-avt-app-rtp-keepalive] 1096 Marjou, X. and A. Sollaud, "Application Mechanism for 1097 keeping alive the Network Address Translator (NAT) 1098 mappings associated to RTP flows.", 1099 draft-ietf-avt-app-rtp-keepalive-09 (work in progress), 1100 September 2010. 1102 [RFC2827] Ferguson, P. and D. Senie, "Network Ingress Filtering: 1103 Defeating Denial of Service Attacks which employ IP Source 1104 Address Spoofing", BCP 38, RFC 2827, May 2000. 1106 [RFC5226] Narten, T. and H. Alvestrand, "Guidelines for Writing an 1107 IANA Considerations Section in RFCs", BCP 26, RFC 5226, 1108 May 2008. 1110 [RFC5104] Wenger, S., Chandra, U., Westerlund, M., and B. Burman, 1111 "Codec Control Messages in the RTP Audio-Visual Profile 1112 with Feedback (AVPF)", RFC 5104, February 2008. 1114 [RFC5652] Housley, R., "Cryptographic Message Syntax (CMS)", STD 70, 1115 RFC 5652, September 2009. 1117 [RFC2818] Rescorla, E., "HTTP Over TLS", RFC 2818, May 2000. 1119 [RFC2974] Handley, M., Perkins, C., and E. Whelan, "Session 1120 Announcement Protocol", RFC 2974, October 2000. 1122 Authors' Addresses 1124 Ali Begen 1125 Cisco 1126 181 Bay Street 1127 Toronto, ON M5J 2T3 1128 Canada 1130 Email: abegen@cisco.com 1132 Dan Wing 1133 Cisco Systems, Inc. 1134 170 West Tasman Dr. 1135 San Jose, CA 95134 1136 USA 1138 Email: dwing@cisco.com 1140 Tom VanCaenegem 1141 Alcatel-Lucent 1142 Copernicuslaan 50 1143 Antwerpen, 2018 1144 Belgium 1146 Email: Tom.Van_Caenegem@alcatel-lucent.com