idnits 2.17.1 draft-ietf-perc-private-media-framework-04.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 (July 3, 2017) is 2482 days in the past. Is this intentional? -- Found something which looks like a code comment -- if you have code sections in the document, please surround them with '' and '' lines. Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: A later version (-12) exists of draft-ietf-perc-double-04 == Outdated reference: A later version (-12) exists of draft-ietf-perc-dtls-tunnel-01 ** Downref: Normative reference to an Informational draft: draft-ietf-perc-dtls-tunnel (ref. 'I-D.ietf-perc-dtls-tunnel') == Outdated reference: A later version (-13) exists of draft-ietf-perc-srtp-ekt-diet-04 == Outdated reference: A later version (-20) exists of draft-ietf-rtcweb-security-arch-12 -- Obsolete informational reference (is this intentional?): RFC 4474 (Obsoleted by RFC 8224) -- Obsolete informational reference (is this intentional?): RFC 4566 (Obsoleted by RFC 8866) Summary: 1 error (**), 0 flaws (~~), 5 warnings (==), 4 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group P. Jones 3 Internet-Draft D. Benham 4 Intended status: Standards Track Cisco 5 Expires: January 4, 2018 C. Groves 6 Huawei 7 July 3, 2017 9 A Solution Framework for Private Media in Privacy Enhanced RTP 10 Conferencing 11 draft-ietf-perc-private-media-framework-04 13 Abstract 15 This document describes a solution framework for ensuring that media 16 confidentiality and integrity are maintained end-to-end within the 17 context of a switched conferencing environment where media 18 distribution devices are not trusted with the end-to-end media 19 encryption keys. The solution aims to build upon existing security 20 mechanisms defined for the real-time transport protocol (RTP). 22 Status of This Memo 24 This Internet-Draft is submitted in full conformance with the 25 provisions of BCP 78 and BCP 79. 27 Internet-Drafts are working documents of the Internet Engineering 28 Task Force (IETF). Note that other groups may also distribute 29 working documents as Internet-Drafts. The list of current Internet- 30 Drafts is at http://datatracker.ietf.org/drafts/current/. 32 Internet-Drafts are draft documents valid for a maximum of six months 33 and may be updated, replaced, or obsoleted by other documents at any 34 time. It is inappropriate to use Internet-Drafts as reference 35 material or to cite them other than as "work in progress." 37 This Internet-Draft will expire on January 4, 2018. 39 Copyright Notice 41 Copyright (c) 2017 IETF Trust and the persons identified as the 42 document authors. All rights reserved. 44 This document is subject to BCP 78 and the IETF Trust's Legal 45 Provisions Relating to IETF Documents 46 (http://trustee.ietf.org/license-info) in effect on the date of 47 publication of this document. Please review these documents 48 carefully, as they describe your rights and restrictions with respect 49 to this document. Code Components extracted from this document must 50 include Simplified BSD License text as described in Section 4.e of 51 the Trust Legal Provisions and are provided without warranty as 52 described in the Simplified BSD License. 54 Table of Contents 56 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 57 2. Conventions Used in This Document . . . . . . . . . . . . . . 4 58 3. PERC Entities and Trust Model . . . . . . . . . . . . . . . . 5 59 3.1. Untrusted Entities . . . . . . . . . . . . . . . . . . . 5 60 3.1.1. Media Distributor . . . . . . . . . . . . . . . . . . 6 61 3.1.2. Call Processing . . . . . . . . . . . . . . . . . . . 6 62 3.2. Trusted Entities . . . . . . . . . . . . . . . . . . . . 7 63 3.2.1. Endpoint . . . . . . . . . . . . . . . . . . . . . . 7 64 3.2.2. Key Distributor . . . . . . . . . . . . . . . . . . . 7 65 4. Framework for PERC . . . . . . . . . . . . . . . . . . . . . 7 66 4.1. End-to-End and Hop-by-Hop Authenticated Encryption . . . 8 67 4.2. E2E Key Confidentiality . . . . . . . . . . . . . . . . . 9 68 4.3. E2E Keys and Endpoint Operations . . . . . . . . . . . . 9 69 4.4. HBH Keys and Hop Operations . . . . . . . . . . . . . . . 10 70 4.5. Key Exchange . . . . . . . . . . . . . . . . . . . . . . 10 71 4.5.1. Initial Key Exchange and Key Distributor . . . . . . 11 72 4.5.2. Key Exchange during a Conference . . . . . . . . . . 11 73 5. Entity Trust . . . . . . . . . . . . . . . . . . . . . . . . 12 74 5.1. Identity Assertions . . . . . . . . . . . . . . . . . . . 12 75 5.2. Certificate Fingerprints in Session Signaling . . . . . . 13 76 5.3. Conferences Identification . . . . . . . . . . . . . . . 13 77 6. Security Considerations . . . . . . . . . . . . . . . . . . . 13 78 6.1. Third Party Attacks . . . . . . . . . . . . . . . . . . . 14 79 6.2. Media Distributor Attacks . . . . . . . . . . . . . . . . 14 80 6.2.1. Denial of service . . . . . . . . . . . . . . . . . . 14 81 6.2.2. Replay Attack . . . . . . . . . . . . . . . . . . . . 15 82 6.2.3. Delayed Playout Attack . . . . . . . . . . . . . . . 15 83 6.2.4. Splicing Attack . . . . . . . . . . . . . . . . . . . 15 84 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 16 85 8. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 16 86 9. References . . . . . . . . . . . . . . . . . . . . . . . . . 16 87 9.1. Normative References . . . . . . . . . . . . . . . . . . 16 88 9.2. Informative References . . . . . . . . . . . . . . . . . 17 89 Appendix A. PERC Key Inventory . . . . . . . . . . . . . . . . . 18 90 A.1. DTLS-SRTP Exchange Yields HBH Keys . . . . . . . . . . . 19 91 A.2. The Key Distributor Transmits the KEK (EKT Key) . . . . . 20 92 A.3. Endpoints fabricate an SRTP Master Key . . . . . . . . . 20 93 A.4. Who has What Key . . . . . . . . . . . . . . . . . . . . 21 94 Appendix B. PERC Packet Format . . . . . . . . . . . . . . . . . 21 95 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 22 97 1. Introduction 99 Switched conferencing is an increasingly popular model for multimedia 100 conferences with multiple participants using a combination of audio, 101 video, text, and other media types. With this model, real-time media 102 flows from conference participants are not mixed, transcoded, 103 transrated, recomposed, or otherwise manipulated by a Media 104 Distributor, as might be the case with a traditional media server or 105 multipoint control unit (MCU). Instead, media flows transmitted by 106 conference participants are simply forwarded by the Media Distributor 107 to each of the other participants, often forwarding only a subset of 108 flows based on voice activity detection or other criteria. In some 109 instances, the Media Distributors may make limited modifications to 110 RTP [RFC3550] headers, for example, but the actual media content 111 (e.g., voice or video data) is unaltered. 113 An advantage of switched conferencing is that Media Distributors can 114 be more easily deployed on general-purpose computing hardware, 115 including virtualized environments in private and public clouds. 116 Deploying conference resources in a public cloud environment might 117 introduce a higher security risk. Whereas traditional conference 118 resources were usually deployed in private networks that were 119 protected, cloud-based conference resources might be viewed as less 120 secure since they are not always physically controlled by those who 121 use them. Additionally, there are usually several ports open to the 122 public in cloud deployments, such as for remote administration, and 123 so on. 125 This document defines a solution framework wherein media privacy is 126 ensured by making it impossible for a media distribution device to 127 gain access to keys needed to decrypt or authenticate the actual 128 media content sent between conference participants. At the same 129 time, the framework allows for the Media Distributors to modify 130 certain RTP headers; add, remove, encrypt, or decrypt RTP header 131 extensions; and encrypt and decrypt RTCP packets. The framework also 132 prevents replay attacks by authenticating each packet transmitted 133 between a given participant and the media distribution device using a 134 unique key per endpoint that is independent from the key for media 135 encryption and authentication. 137 A goal of this document is to define a framework for enhanced privacy 138 in RTP-based conferencing environments while utilizing existing 139 security procedures defined for RTP with minimal enhancements. 141 2. Conventions Used in This Document 143 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 144 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 145 document are to be interpreted as described in [RFC2119] when they 146 appear in ALL CAPS. These words may also appear in this document in 147 lower case as plain English words, absent their normative meanings. 149 Additionally, this solution framework uses the following conventions, 150 terms and acronyms: 152 End-to-End (E2E): Communications from one endpoint through one or 153 more Media Distribution Devices to the endpoint at the other end. 155 Hop-by-Hop (HBH): Communications between an endpoint and a Media 156 Distribution Device or between Media Distribution Devices. 158 Trusted Endpoint: An RTP flow terminating entity that has possession 159 of E2E media encryption keys and terminates E2E encryption. This may 160 include embedded user conferencing equipment or browsers on 161 computers, media gateways, MCUs, media recording device and more that 162 are in the trusted domain for a given deployment. 164 Media Distributor (MD): An RTP middlebox that is not allowed to to 165 have access to E2E encryption keys. It operates according to the 166 Selective Forwarding Middlebox RTP topologies 167 [I-D.ietf-avtcore-rtp-topologies-update] per the constraints defined 168 by the PERC system, which includes, but not limited to, having no 169 access to RTP media unencrypted and having limits on what RTP header 170 field it can alter. 172 Key Distributor: An entity that is a logical function which 173 distributes keying material and related information to trusted 174 endpoints and Media Distributor(s), only that which is appropriate 175 for each. The Key Distributor might be co-resident with another 176 entity trusted with E2E keying material. 178 Conference: Two or more participants communicating via trusted 179 endpoints to exchange RTP flows through one or more Media 180 Distributor. 182 Call Processing: All trusted endpoints in the conference connect to 183 it by a call processing dialog, such as with the Focus defined in the 184 Framework for Conferencing with SIP [RFC4353]. 186 Third Party: Any entity that is not an Endpoint, Media Distributor, 187 Key Distributor or Call Processing entity as described in this 188 document. 190 3. PERC Entities and Trust Model 192 The following figure depicts the trust relationships, direct or 193 indirect, between entities described in the subsequent sub-sections. 194 Note that these entities may be co-located or further divided into 195 multiple, separate physical devices. 197 Please note that some entities classified as untrusted in the simple, 198 general deployment scenario used most commonly in this document might 199 be considered trusted in other deployments. This document does not 200 preclude such scenarios, but will keep the definitions and examples 201 focused by only using the the simple, most general deployment 202 scenario. 204 | 205 +----------+ | +-----------------+ 206 | Endpoint | | | Call Processing | 207 +----------+ | +-----------------+ 208 | 209 | 210 +----------------+ | +--------------------+ 211 | Key Distributor| | | Media Distributor | 212 +----------------+ | +--------------------+ 213 | 214 Trusted | Untrusted 215 Entities | Entities 216 | 218 Figure 1: Trusted and Untrusted Entities in PERC 220 3.1. Untrusted Entities 222 The architecture described in this framework document enables 223 conferencing infrastructure to be hosted in domains, such as in a 224 cloud conferencing provider's facilities, where the trustworthiness 225 is below the level needed to assume the privacy of participant's 226 media will not be compromised. The conferencing infrastructure in 227 such a domain is still trusted with reliably connecting the 228 participants together in a conference, but not trusted with keying 229 material needed to decrypt any of the participant's media. Entities 230 in such lower trustworthiness domains will simply be referred to as 231 untrusted entities from this point forward. This does not mean that 232 they are completely untrusted as they may be trusted with most non- 233 media related aspects of hosting a conference. 235 3.1.1. Media Distributor 237 A Media Distributor forwards RTP flows between endpoints in the 238 conference while performing per-hop authentication of each RTP 239 packet. The Media Distributor may need access to one or more RTP 240 headers or header extensions, potentially adding or modifying a 241 certain subset. The Media Distributor will also relay secured 242 messaging between the endpoints and the Key Distributor and will 243 acquire per-hop key information from the Key Distributor. The actual 244 media content *MUST NOT* not be decryptable by a Media Distributor, 245 so it is untrusted to have access to the E2E media encryption keys, 246 which this framework's key exchange mechanisms will prevent. 248 An endpoint's ability to join a conference hosted by a Media 249 Distributor MUST NOT alone be interpreted as being authorized to have 250 access to the E2E media encryption keys, as the Media Distributor 251 does not have the ability to determine whether an endpoint is 252 authorized. Trusted endpoint authorization is described in 253 [I-D.roach-perc-webrtc]. 255 A Media Distributor MUST perform its role in properly forwarding 256 media packets while taking measures to mitigate the adverse effects 257 of denial of service attacks (refer to Section 6), etc, to a level 258 equal to or better than traditional conferencing (i.e. non-PERC) 259 deployments. 261 A Media Distributor or associated conferencing infrastructure may 262 also initiate or terminate various conference control related 263 messaging, which is outside the scope of this framework document. 265 3.1.2. Call Processing 267 The call processing function is untrusted in the simple, general 268 deployment scenario. When a physical subset of the call processing 269 function resides in facilities outside the trusted domain, it should 270 not be trusted to have access to E2E key information. 272 The call processing function may include the processing of call 273 signaling messages, as well as the signing of those messages. It may 274 also authenticate the endpoints for the purpose of call signaling and 275 subsequently joining of a conference hosted through one or more Media 276 Distributors. Call processing may optionally ensure the privacy of 277 call signaling messages between itself, the endpoint, and other 278 entities. 280 In any deployment scenario where the call processing function is 281 considered trusted, the call processing function MUST ensure the 282 integrity of received messages before forwarding to other entities. 284 3.2. Trusted Entities 286 From the PERC model system perspective, entities considered trusted 287 (refer to Figure 1) can be in possession of the E2E media encryption 288 key(s) for one or more conferences. 290 3.2.1. Endpoint 292 An endpoint is considered trusted and will have access to E2E key 293 information. While it is possible for an endpoint to be compromised, 294 subsequently performing in undesired ways, defining endpoint 295 resistance to compromise is outside the scope of this document. 296 Endpoints will take measures to mitigate the adverse effects of 297 denial of service attacks (refer to Section 6) from other entities, 298 including from other endpoints, to a level equal to or better than 299 traditional conference (i.e., non-PERC) deployments. 301 3.2.2. Key Distributor 303 The Key Distributor, which may be collocated with an endpoint or 304 exist standalone, is responsible for providing key information to 305 endpoints for both end-to-end and hop-by-hop security and for 306 providing key information to Media Distributors for the hop-by-hop 307 security. 309 Interaction between the Key Distributor and the call processing 310 function is necessary to for proper conference-to-endpoint mappings. 311 This is described in Section 5.3. 313 The Key Distributor needs to be secured and managed in a way to 314 prevent exploitation by an adversary, as any kind of compromise of 315 the Key Distributor puts the security of the conference at risk. 317 4. Framework for PERC 319 The purpose for this framework is to define a means through which 320 media privacy can be ensured when communicating within a conferencing 321 environment consisting of one or more Media Distributors that only 322 switch, hence not terminate, media. It does not otherwise attempt to 323 hide the fact that a conference between endpoints is taking place. 325 This framework reuses several specified RTP security technologies, 326 including SRTP [RFC3711], PERC EKT [I-D.ietf-perc-srtp-ekt-diet], and 327 DTLS-SRTP [RFC5764]. 329 4.1. End-to-End and Hop-by-Hop Authenticated Encryption 331 This solution framework focuses on the end-to-end privacy and 332 integrity of the participant's media by limiting access of the end- 333 to-end key information to trusted entities. However, this framework 334 does give a Media Distributor access to RTP headers and all or most 335 header extensions, as well as the ability to modify a certain subset 336 of those headers and to add header extensions. Packets received by a 337 Media Distributor or an endpoint are authenticated hop-by-hop. 339 To enable all of the above, this framework defines the use of two 340 security contexts and two associated encryption keys: an "inner" key 341 (an E2E key distinct for each transmitted media flow) for 342 authenticated encryption of RTP media between endpoints and an 343 "outer" key (HBH key) known only to media distributor and the 344 adjacent endpoint) for the hop between an endpoint and a Media 345 Distributor or between Media Distributor. Reference the following 346 figure. 348 +-------------+ +-------------+ 349 | |################################| | 350 | Media |------------------------------->| Media | 351 | Distributor |<-------------------------------| Distributor | 352 | X |################################| Y | 353 | | HBH Key (XY) | | 354 +-------------+ +-------------+ 355 # ^ | # # ^ | # 356 # | | # HBH HBH # | | # 357 # | | # <== Key(AX) Key(YB) ==> # | | # 358 # | | # # | | # 359 # |<+--#---- E2E Key (A) E2E Key (B) ---#->| | # 360 # | | # # | | # 361 # | v # # | v # 362 +-------------+ +-------------+ 363 | Endpoint A | | Endpoint B | 364 +-------------+ +-------------+ 366 E2E and HBH Keys Used for Authenticated Encryption 368 The PERC Double specification [I-D.ietf-perc-double] uses standard 369 SRTP keying material and recommended cryptographic transform(s) to 370 first form the inner, end-to-end SRTP cryptographic context. That 371 end-to-end SRTP cryptographic context MAY be used to encrypt some RTP 372 header extensions along with RTP media content. The output of this 373 is treated like an RTP packet and encrypted again using the outer 374 hop-by-hop cryptographic context. The endpoint executes the entire 375 Double operation while the Media Distributor just performs the outer, 376 hop-by-hop operation. (See Appendix A for a description of the keys 377 used in PERC and Appendix B for an overview of how the packet appears 378 on the wire.) 380 RTCP can only be encrypted hop-by-hop, not end-to-end. This 381 framework introduces no additional step for RTCP authenticated 382 encryption, so the procedures needed are specified in [RFC3711] and 383 use the same outer, hop-by-hop cryptographic context chosen in the 384 Double operation described above. 386 4.2. E2E Key Confidentiality 388 To ensure the confidentiality of E2E keys shared between endpoints, 389 endpoints will make use of a common Key Encryption Key (KEK) that is 390 known only by the trusted entities in a conference. That KEK, 391 defined in the PERC EKT [I-D.ietf-perc-srtp-ekt-diet] as the EKT Key, 392 will be used to subsequently encrypt the SRTP master key used for E2E 393 authenticated encryption of media sent by a given endpoint. Each 394 endpoint in the conference will create a random SRTP master key for 395 E2E authenticated encryption, thus participants in the conference 396 MUST keep track of the E2E keys received via the Full EKT Field for 397 each distinct SSRC in the conference so that it can properly decrypt 398 received media. Note, too, that an endpoint may change its E2E key 399 at any time and advertise that new key to the conference as specified 400 in [I-D.ietf-perc-srtp-ekt-diet]. 402 4.3. E2E Keys and Endpoint Operations 404 Any given RTP media flow can be identified by its SSRC, and endpoints 405 might send more than one at a time and change the mix of media flows 406 transmitted during the life of a conference. 408 Thus, endpoints MUST maintain a list of SSRCs from received RTP flows 409 and each SSRC's associated E2E key information. Following a change 410 in an E2E key, prior E2E keys SHOULD be retained by receivers for a 411 period long enough to ensure that late-arriving or out-of-order 412 packets from the endpoint can be successfully decrypted. Receiving 413 endpoints MUST discard old E2E keys no later than when it leaves the 414 conference. 416 If there is a need to encrypt one or more RTP header extensions end- 417 to-end, an encryption key is derived from the end-to-end SRTP master 418 key to encrypt header extensions as per [RFC6904]. The Media 419 Distributor will not be able use the information contained in those 420 header extensions encrypted with an E2E key. 422 4.4. HBH Keys and Hop Operations 424 To ensure the integrity of transmitted media packets, this framework 425 requires that every packet be authenticated hop-by-hop (HBH) between 426 an endpoint and a Media Distributor, as well between Media 427 Distributors. The authentication key used for hop-by-hop 428 authentication is derived from an SRTP master key shared only on the 429 respective hop. Each HBH key is distinct per hop and no two hops 430 ever intentionally use the same SRTP master key. 432 Using hop-by-hop authentication gives the Media Distributor the 433 ability to change certain RTP header values. Which values the Media 434 Distributor can change in the RTP header are defined in 435 [I-D.ietf-perc-double]. RTCP can only be encrypted, giving the Media 436 Distributor the flexibility to forward RTCP content unchanged, 437 transmit compound RTCP packets or to initiate RTCP packets for 438 reporting statistics or conveying other information. Performing hop- 439 by-hop authentication for all RTP and RTCP packets also helps provide 440 replay protection (see Section 6). 442 If there is a need to encrypt one or more RTP header extensions hop- 443 by-hop, an encryption key is derived from the hop-by-hop SRTP master 444 key to encrypt header extensions as per [RFC6904]. This will still 445 give the Media Distributor visibility into header extensions, such as 446 the one used to determine audio level [RFC6464] of conference 447 participants. Note that when RTP header extensions are encrypted, 448 all hops - in the untrusted domain at least - will need to decrypt 449 and re-encrypt these encrypted header extensions. 451 4.5. Key Exchange 453 To facilitate key exchange required to establish or generate an E2E 454 key and a HBH key for an endpoint and the same HBH key for the Media 455 Distributor, this framework utilizes a DTLS-SRTP [RFC5764] 456 association between an endpoint and the Key Distributor. To 457 establish this association, an endpoint will send DTLS-SRTP messages 458 to the Media Distributor which will then forward them to the Key 459 Distributor as defined in [I-D.ietf-perc-dtls-tunnel]. The Key 460 Encryption Key (KEK) (i.e., EKTKey) is also conveyed by the Key 461 Distributor over the DTLS association to endpoints via procedures 462 defined in PERC EKT [I-D.ietf-perc-srtp-ekt-diet]. 464 Media Distributors use DTLS-SRTP [RFC5764] directly with a peer Media 465 Distributor to establish the HBH key for transmitting RTP and RTCP 466 packets to that peer Media Distributor. The Key Distributor does not 467 facilitate establishing a HBH key for use between Media Distributors. 469 4.5.1. Initial Key Exchange and Key Distributor 471 The procedures defined in DTLS Tunnel for PERC 472 [I-D.ietf-perc-dtls-tunnel] establish one or more TLS tunnels between 473 the Media Distributor and Key Distributor, making it is possible for 474 the Media Distributor to facilitate the establishment of a secure 475 DTLS association between each endpoint and the Key Distributor as 476 shown the following figure. The DTLS association between endpoints 477 and the Key Distributor will enable each endpoint to receive E2E key 478 information, Key Encryption Key (KEK) information (i.e., EKT Key), 479 and HBH key information. At the same time, the Key Distributor can 480 securely provide the HBH key information to the Media Distributor. 481 The key information summarized here may include the SRTP master key, 482 SRTP master salt, and the negotiated cryptographic transform. 484 +-----------+ 485 KEK info | Key | HBH Key info to 486 to Endpoints |Distributor| Endpoints & Media Distributor 487 +-----------+ 488 # ^ ^ # 489 # | | #-TLS Tunnel 490 # | | # 491 +-----------+ +-----------+ +-----------+ 492 | Endpoint | DTLS | Media | DTLS | Endpoint | 493 | KEK |<------------|Distributor|------------>| KEK | 494 | HBH Key | to Key Dist | HBH Keys | to Key Dist | HBH Key | 495 +-----------+ +-----------+ +-----------+ 497 Figure 2: Exchanging Key Information Between Entities 499 Endpoints will establish a DTLS-SRTP association over the RTP 500 session's media ports for the purposes of key information exchange 501 with the Key Distributor. The Media Distributor will not terminate 502 the DTLS signaling, but will instead forward DTLS packets received 503 from an endpoint on to the Key Distributor (and vice versa) via a 504 tunnel established between Media Distributor and the Key Distributor. 505 This tunnel is used to encapsulate the DTLS-SRTP signaling between 506 the Key Distributor and endpoints will also be used to convey HBH key 507 information from the Key Distributor to the Media Distributor, so no 508 additional protocol or interface is required. 510 4.5.2. Key Exchange during a Conference 512 Following the initial key information exchange with the Key 513 Distributor, an endpoints will be able to encrypt media end-to-end 514 with an E2E key, sending that E2E key to other endpoints encrypted 515 with the KEK, and will be able to encrypt and authenticate RTP 516 packets using a HBH key. The procedures defined do not allow the 517 Media Distributor to gain access to the KEK information, preventing 518 it from gaining access to any endpoint's E2E key and subsequently 519 decrypting media. 521 The KEK (i.e., EKT Key) may need to change from time-to-time during 522 the life of a conference, such as when a new participant joins or 523 leaves a conference. Dictating if, when or how often a conference is 524 to be re-keyed is outside the scope of this document, but this 525 framework does accommodate re-keying during the life of a conference. 527 When a Key Distributor decides to re-key a conference, it transmits a 528 specific message defined in PERC EKT [I-D.ietf-perc-srtp-ekt-diet] to 529 each of the conference participants. The endpoint MUST create a new 530 SRTP master key and prepare to send that key inside a Full EKT Field 531 using the new EKTKey. Since it may take some time for all of the 532 endpoints in conference to finish re-keying, senders MUST delay a 533 short period of time before sending media encrypted with the new 534 master key, but it MUST be prepared to make use of the information 535 from a new inbound EKT Key immediately. See Section 2.2.2 of 536 [I-D.ietf-perc-srtp-ekt-diet]. 538 5. Entity Trust 540 It is important to this solution framework that the entities can 541 trust and validate the authenticity of other entities, especially the 542 Key Distributor and endpoints. The details of this are outside the 543 scope of specification but a few possibilities are discussed in the 544 following sections. The key requirements is that endpoints can 545 verify they are connected to the correct Key Distributor for the 546 conference and the Key Distributor can verify the endpoints are the 547 correct endpoints for the conference. 549 Two possible approaches to solve this are Identity Assertions and 550 Certificate Fingerprints. 552 5.1. Identity Assertions 554 WebRTC Identity assertion [I-D.ietf-rtcweb-security-arch] can be used 555 to bind the identity of the user of the endpoint to the fingerprint 556 of the DTLS-SRTP certificate used for the call. This certificate is 557 unique for a given call and a conference. This allows the Key 558 Distributor to ensure that only authorized users participate in the 559 conference. Similarly the Key Distributor can create a WebRTC 560 Identity assertion to bind the fingerprint of the unique certificate 561 used by the Key Distributor for this conference so that the endpoint 562 can validate it is talking to the correct Key Distributor. Such a 563 setup requires an Identity Provider (Idp) trusted by the endpoints 564 and the Key Distributor. 566 5.2. Certificate Fingerprints in Session Signaling 568 Entities managing session signaling are generally assumed to be 569 untrusted in the PERC framework. However, there are some deployment 570 scenarios where parts of the session signaling may be assumed 571 trustworthy for the purposes of exchanging, in a manner that can be 572 authenticated, the fingerprint of an entity's certificate. 574 As a concrete example, SIP [RFC3261] and SDP [RFC4566] can be used to 575 convey the fingerprint information per [RFC5763]. An endpoint's SIP 576 User Agent would send an INVITE message containing SDP for the media 577 session along with the endpoint's certificate fingerprint, which can 578 be signed using the procedures described in [RFC4474] for the benefit 579 of forwarding the message to other entities by the Focus [RFC4353]. 580 Other entities can now verify the fingerprints match the certificates 581 found in the DTLS-SRTP connections to find the identity of the far 582 end of the DTLS-SRTP connection and check that is the authorized 583 entity. 585 Ultimately, if using session signaling, an endpoint's certificate 586 fingerprint would need to be securely mapped to a user and conveyed 587 to the Key Distributor so that it can check that that user is 588 authorized. Similarly, the Key Distributor's certificate fingerprint 589 can be conveyed to endpoint in a manner that can be authenticated as 590 being an authorized Key Distributor for this conference. 592 5.3. Conferences Identification 594 The Key Distributor needs to know what endpoints are being added to a 595 given conference. Thus, the Key Distributor and the Media 596 Distributor will need to know endpoint-to-conference mappings, which 597 is enabled by exchanging a conference-specific unique identifier as 598 defined in [I-D.ietf-perc-dtls-tunnel]. How this unique identifier 599 is assigned is outside the scope of this document. 601 6. Security Considerations 603 This framework, and the individual protocols defined to support it, 604 must take care to not increase the exposure to Denial of Service 605 (DoS) attacks by untrusted or third-party entities and should take 606 measures to mitigate, where possible, more serious DoS attacks from 607 on-path and off-path attackers. 609 The following section enumerates the kind of attacks that will be 610 considered in the development of this framework's solution. 612 6.1. Third Party Attacks 614 On-path attacks are mitigated by HBH integrity protection and 615 encryption. The integrity protection mitigates packet modification 616 and encryption makes selective blocking of packets harder, but not 617 impossible. 619 Off-path attackers may try connecting to different PERC entities and 620 send specifically crafted packets. A successful attacker might be 621 able to get the Media Distributor to forward such packets. If not 622 making use of HBH authentication on the Media Distributor, such an 623 attack could only be detected in the receiving endpoints where the 624 forged packets would finally be dropped. 626 Another potential attack is a third party claiming to be a Media 627 Distributor, fooling endpoints in to sending packets to the false 628 Media Distributor instead of the correct one. The deceived sending 629 endpoints could incorrectly assuming their packets have been 630 delivered to endpoints when they in fact have not. Further, the 631 false Media Distributor may cascade to another legitimate Media 632 Distributor creating a false version of the real conference. 634 This attack can be mitigated by the false Media Distributor not being 635 authenticated by the Key Distributor during PERC Tunnel 636 establishment. Without the tunnel in place, endpoints will not 637 establish secure associations with the Key Distributor and receive 638 the KEK, causing the conference to not proceed. 640 6.2. Media Distributor Attacks 642 The Media Distributor can attack the session in a number of possible 643 ways. 645 6.2.1. Denial of service 647 Any modification of the end-to-end authenticated data will result in 648 the receiving endpoint getting an integrity failure when performing 649 authentication on the received packet. 651 The Media Distributor can also attempt to perform resource 652 consumption attacks on the receiving endpoint. One such attack would 653 be to insert random SSRC/CSRC values in any RTP packet with an inband 654 key-distribution message attached (i.e., Full EKT Field). Since such 655 a message would trigger the receiver to form a new cryptographic 656 context, the Media Distributor can attempt to consume the receiving 657 endpoints resources. 659 Another denial of service attack is where the Media Distributor 660 rewrites the PT field to indicate a different codec. The effect of 661 this attack is that any payload packetized and encoded according to 662 one RTP payload format is then processed using another payload format 663 and codec. Assuming that the implementation is robust to random 664 input, it is unlikely to cause crashes in the receiving software/ 665 hardware. However, it is not unlikely that such rewriting will cause 666 severe media degradation. 668 For audio formats, this attack is likely to cause highly disturbing 669 audio and/or can be damaging to hearing and playout equipment. 671 6.2.2. Replay Attack 673 Replay attack is when an already received packets from a previous 674 point in the RTP stream is replayed as new packet. This could, for 675 example, allow a Media Distributor to transmit a sequence of packets 676 identified as a user saying "yes", instead of the "no" the user 677 actually said. 679 The mitigation for a replay attack is to prevent old packets beyond a 680 small-to-modest jitter and network re-ordering sized window to be 681 rejected. End-to-end replay protection MUST be provided for the 682 whole duration of the conference. 684 6.2.3. Delayed Playout Attack 686 The delayed playout attack is a variant of the replay attack. This 687 attack is possible even if E2E replay protection is in place. 688 However, due to fact that the Media Distributor is allowed to select 689 a sub-set of streams and not forward the rest to a receiver, such as 690 in forwarding only the most active speakers, the receiver has to 691 accept gaps in the E2E packet sequence. The issue with this is that 692 a Media Distributor can select to not deliver a particular stream for 693 a while. 695 Within the window from last packet forwarded to the receiver and the 696 latest received by the Media Distributor, the Media Distributor can 697 select an arbitrary starting point when resuming forwarding packets. 698 Thus what the media source said can be substantially delayed at the 699 receiver with the receiver believing that it is what was said just 700 now, and only delayed due to transport delay. 702 6.2.4. Splicing Attack 704 The splicing attack is an attack where a Media Distributor receiving 705 multiple media sources splices one media stream into the other. If 706 the Media Distributor is able to change the SSRC without the receiver 707 having any method for verifying the original source ID, then the 708 Media Distributor could first deliver stream A and then later forward 709 stream B under the same SSRC as stream A was previously using. Not 710 allowing the Media Distributor to change the SSRC mitigates this 711 attack. 713 7. IANA Considerations 715 There are no IANA considerations for this document. 717 8. Acknowledgments 719 The authors would like to thank Mo Zanaty and Christian Oien for 720 invaluable input on this document. Also, we would like to 721 acknowledge Nermeen Ismail for serving on the initial versions of 722 this document as a co-author. 724 9. References 726 9.1. Normative References 728 [I-D.ietf-perc-double] 729 Jennings, C., Jones, P., and A. Roach, "SRTP Double 730 Encryption Procedures", draft-ietf-perc-double-04 (work in 731 progress), April 2017. 733 [I-D.ietf-perc-dtls-tunnel] 734 Jones, P., Ellenbogen, P., and N. Ohlmeier, "DTLS Tunnel 735 between a Media Distributor and Key Distributor to 736 Facilitate Key Exchange", draft-ietf-perc-dtls-tunnel-01 737 (work in progress), April 2017. 739 [I-D.ietf-perc-srtp-ekt-diet] 740 Jennings, C., Mattsson, J., McGrew, D., and D. Wing, 741 "Encrypted Key Transport for DTLS and Secure RTP", draft- 742 ietf-perc-srtp-ekt-diet-04 (work in progress), April 2017. 744 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 745 Requirement Levels", BCP 14, RFC 2119, 746 DOI 10.17487/RFC2119, March 1997, 747 . 749 [RFC3550] Schulzrinne, H., Casner, S., Frederick, R., and V. 750 Jacobson, "RTP: A Transport Protocol for Real-Time 751 Applications", STD 64, RFC 3550, DOI 10.17487/RFC3550, 752 July 2003, . 754 [RFC3711] Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K. 755 Norrman, "The Secure Real-time Transport Protocol (SRTP)", 756 RFC 3711, DOI 10.17487/RFC3711, March 2004, 757 . 759 [RFC5764] McGrew, D. and E. Rescorla, "Datagram Transport Layer 760 Security (DTLS) Extension to Establish Keys for the Secure 761 Real-time Transport Protocol (SRTP)", RFC 5764, 762 DOI 10.17487/RFC5764, May 2010, 763 . 765 [RFC6904] Lennox, J., "Encryption of Header Extensions in the Secure 766 Real-time Transport Protocol (SRTP)", RFC 6904, 767 DOI 10.17487/RFC6904, April 2013, 768 . 770 9.2. Informative References 772 [I-D.ietf-avtcore-rtp-topologies-update] 773 Westerlund, M. and S. Wenger, "RTP Topologies", draft- 774 ietf-avtcore-rtp-topologies-update-10 (work in progress), 775 July 2015. 777 [I-D.ietf-rtcweb-security-arch] 778 Rescorla, E., "WebRTC Security Architecture", draft-ietf- 779 rtcweb-security-arch-12 (work in progress), June 2016. 781 [I-D.roach-perc-webrtc] 782 Roach, A., "Using Privacy Enhanced Real-time Conferencing 783 (PERC) in a WebRTC Context", draft-roach-perc-webrtc-00 784 (work in progress), March 2017. 786 [RFC3261] Rosenberg, J., Schulzrinne, H., Camarillo, G., Johnston, 787 A., Peterson, J., Sparks, R., Handley, M., and E. 788 Schooler, "SIP: Session Initiation Protocol", RFC 3261, 789 DOI 10.17487/RFC3261, June 2002, 790 . 792 [RFC4353] Rosenberg, J., "A Framework for Conferencing with the 793 Session Initiation Protocol (SIP)", RFC 4353, 794 DOI 10.17487/RFC4353, February 2006, 795 . 797 [RFC4474] Peterson, J. and C. Jennings, "Enhancements for 798 Authenticated Identity Management in the Session 799 Initiation Protocol (SIP)", RFC 4474, 800 DOI 10.17487/RFC4474, August 2006, 801 . 803 [RFC4566] Handley, M., Jacobson, V., and C. Perkins, "SDP: Session 804 Description Protocol", RFC 4566, DOI 10.17487/RFC4566, 805 July 2006, . 807 [RFC5763] Fischl, J., Tschofenig, H., and E. Rescorla, "Framework 808 for Establishing a Secure Real-time Transport Protocol 809 (SRTP) Security Context Using Datagram Transport Layer 810 Security (DTLS)", RFC 5763, DOI 10.17487/RFC5763, May 811 2010, . 813 [RFC6464] Lennox, J., Ed., Ivov, E., and E. Marocco, "A Real-time 814 Transport Protocol (RTP) Header Extension for Client-to- 815 Mixer Audio Level Indication", RFC 6464, 816 DOI 10.17487/RFC6464, December 2011, 817 . 819 Appendix A. PERC Key Inventory 821 PERC specifies the use of a number of different keys and, 822 understandably, it looks complicated or confusing on the surface. 823 This section summarizes the various keys used in the system, how they 824 are generated, and what purpose they serve. 826 The keys are described in the order in which they would typically be 827 acquired. 829 The various keys used in PERC are shown in Figure 3 below. 831 +-----------+----------------------------------------------------+ 832 | Key | Description | 833 +-----------+----------------------------------------------------+ 834 | KEK | Key shared by all endpoints and used to encrypt | 835 | (EKT Key) | each endpoint's SRTP master key so receiving | 836 | | endpoints can decrypt media. | 837 +-----------+----------------------------------------------------+ 838 | HBH Key | Key used to encrypt media hop-by-hop. | 839 +-----------+----------------------------------------------------+ 840 | E2E Key | Key used to encrypt media end-to-end. | 841 +-----------+----------------------------------------------------+ 843 Figure 3: Key Inventory 845 As you can see, the number key types is very small. However, what 846 can be challenging is keeping track of all of the distinct E2E keys 847 as the conference grows in size. With 1,000 participants in a 848 conference, there will be 1,000 distinct SRTP master keys, all of 849 which share the same master salt. Each of those keys are passed 850 through the KDF defined in [RFC3711] to produce the actual encryption 851 and authentication keys. Complicating key management is the fact 852 that the KEK can change and, when it does, the endpoints generate new 853 SRTP master keys. And, of course, there is a new SRTP master salt to 854 go with those keys. Endpoints have to retain old keys for a period 855 of time to ensure they can properly decrypt late-arriving or out-of- 856 order packets. 858 The time required to retain old keys (either EKT Keys or SRTP master 859 keys) is not specified, but they should be retained at least for the 860 period of time required to re-key the conference or handle late- 861 arriving or out-of-order packets. A period of 60s should be 862 considered a generous retention period, but endpoints may keep old 863 keys on hand until the end of the conference. 865 Or more detailed explanation of each of the keys follows. 867 A.1. DTLS-SRTP Exchange Yields HBH Keys 869 The first set of keys acquired are for hop-by-hop encryption and 870 decryption. Assuming the use of Double [I-D.ietf-perc-double], the 871 endpoint would perform DTLS-SRTP exchange with the key distributor 872 and receive a key that is, in fact, "double" the size that is needed. 873 Per the Double specification, the E2E part is the first half of the 874 key, so the endpoint will just discard that information in PERC. It 875 is not used. The second half of the key material is for HBH 876 operations, so that half of the key (corresponding to the least 877 significant bits) is assigned internally as the HBH key. 879 The media distributor doesn't perform DTLS-SRTP, but it is at this 880 point that the key distributor will inform the media distributor of 881 the HBH key value via the tunnel protocol 882 ([I-D.ietf-perc-dtls-tunnel]). The key distributor will send the 883 least significant bits corresponding to the half of the keying 884 material determined through DTLS-SRTP with the endpoint to the media 885 distributor via the tunnel protocol. There is a salt generated along 886 with the HBH key. The salt is also longer than needed for HBH 887 operations, thus only the least significant bits of the required 888 length (i.e., half of the generated salt material) are sent to the 889 media distributor via the tunnel protocol. 891 No two endpoints will have the same HBH key, thus the media 892 distributor must keep track each distinct HBH key (and the 893 corresponding salt) and use it only for the specified hop. 895 This key is also used for HBH encryption of RTCP. RTCP is not end- 896 to-end encrypted in PERC. 898 A.2. The Key Distributor Transmits the KEK (EKT Key) 900 Via the aforementioned DTLS-SRTP association, the key distributor 901 will send the endpoint the KEK (i.e., EKT Key per 902 [I-D.ietf-perc-srtp-ekt-diet]). This key is known only to the key 903 distributor and endpoints. This key is the most important to protect 904 since having knowledge of this key (and the SRTP master salt 905 transmitted as a part of the same message) will allow an entity to 906 decrypt any media packet in the conference. 908 Note that the key distributor can send any number of EKT Keys to 909 endpoints. This can be used to re-key the entire conference. Each 910 key is identified by a "Security Parameter Index" (SPI) value. 911 Endpoints should expect that a conference might be re-keyed when a 912 new participant joins a conference or when a participant leaves a 913 conference in order to protect the confidentiality of the 914 conversation before and after such events. 916 The SRTP master salt to be used by the endpoint is transmitted along 917 with the EKT Key. All endpoints in the conference utilize the same 918 SRTP master salt that corresponds with a given EKT Key. 920 The EKT Field in media packets is encrypted using a cipher specified 921 via the EKTKey message (e.g., AES Key Wrap with a 128-bit key). This 922 cipher is different than the cipher used to protect media and is only 923 used to encrypt the endpoint's SRTP master key (and other EKT Field 924 data as per [I-D.ietf-perc-srtp-ekt-diet]). 926 The media distributor is not given the KEK (i.e., EKT Key). 928 A.3. Endpoints fabricate an SRTP Master Key 930 As stated earlier, the E2E key determined via DTLS-SRTP is discarded. 931 While it could have been used, the fact that endpoints may need to 932 change the SRTP master key periodically or are forced to change the 933 SRTP master key as a result of the EKT key changing means using it 934 has only limited utility. To reduce complexity, PERC prescribes that 935 endpoints manufacturer random SRTP master keys locally to be used for 936 E2E encryption. 938 This locally-generated SRTP master key is used along with the master 939 salt transmitted to the endpoint from the key distributor via the 940 EKTKey message to encrypt media end-to-end. 942 Since the media distributor is not involved in E2E functions, it will 943 not create this key nor have access to any endpoint's E2E key. Note, 944 too, that even the key distributor is unaware of the locally- 945 generated E2E keys used by each endpoint. 947 The endpoint will transmit its E2E key to other endpoints in the 948 conference by periodically including it in SRTP packets in a Full EKT 949 Field. When placed in the Full EKT Field, it is encrypted using the 950 EKT Key provided by the key distributor. The master salt is not 951 transmitted, though, since all endpoints will have received the same 952 master salt via the EKTKey message. The recommended frequency with 953 which an endpoint transmits its SRTP master key is specified in 954 [I-D.ietf-perc-srtp-ekt-diet]. 956 A.4. Who has What Key 958 All endpoints have knowledge of the KEK. 960 Every HBH key is distinct for a given endpoint, thus Endpoint A and 961 endpoint B do not have knowledge of the other's HBH key. 963 Each endpoint generates its own E2E Key (SRTP master key), thus the 964 key distinct per endpoint. This key is transmitted (encrypted) via 965 the EKT Field to other endpoints. Endpoints that receive media from 966 a given transmitting endpoint will therefore have knowledge of the 967 transmitter's E2E key. 969 To summarize the various keys and which entity is in possession of a 970 given key, refer to Figure 4. 972 +----------------------+------------+-------+-------+------------+ 973 | Key / Entity | Endpoint A | MD X | MD Y | Endpoint B | 974 +----------------------+------------+-------+-------+------------+ 975 | KEK | Yes | No | No | Yes | 976 +----------------------+------------+-------+-------+------------+ 977 | E2E Key (A and B) | Yes | No | No | Yes | 978 +----------------------+------------+-------+-------+------------+ 979 | HBH Key (A<=>MD X) | Yes | Yes | No | No | 980 +----------------------+------------+-------+-------+------------+ 981 | HBH Key (B<=>MD Y) | No | No | Yes | Yes | 982 +----------------------+------------+---------------+------------+ 983 | HBH Key (MD X<=>MD Y)| No | Yes | Yes | No | 984 +----------------------+------------+---------------+------------+ 986 Figure 4: Keys per Entity 988 Appendix B. PERC Packet Format 990 Figure 5 presents a complete picture of what a PERC packet looks like 991 when transmitted over the wire. 993 0 1 2 3 994 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 1 995 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 996 A |V=2|P|X| CC |M| PT | sequence number | 997 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 998 A | timestamp | 999 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1000 A | synchronization source (SSRC) identifier | 1001 +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+ 1002 A | contributing source (CSRC) identifiers | 1003 A | .... | 1004 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1005 A | RTP extension (OPTIONAL) | 1006 A | (including the OHB) | 1007 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1008 C : : 1009 C : Ciphertext Payload : 1010 C : : 1011 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1012 R : : 1013 R : EKT Field : 1014 R : : 1015 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1017 C = Ciphertext (encrypted and authenticated) 1018 A = Associated Data (authenticated only) 1019 R = neither encrypted nor authenticated, added 1020 after Authenticated Encryption completed 1022 Figure 5: PERC Packet Format 1024 Authors' Addresses 1026 Paul E. Jones 1027 Cisco 1028 7025 Kit Creek Rd. 1029 Research Triangle Park, North Carolina 27709 1030 USA 1032 Phone: +1 919 476 2048 1033 Email: paulej@packetizer.com 1034 David Benham 1035 Cisco 1036 170 West Tasman Drive 1037 San Jose, California 95134 1038 USA 1040 Email: dbenham@cisco.com 1042 Christian Groves 1043 Huawei 1044 Melbourne 1045 Australia 1047 Email: Christian.Groves@nteczone.com