AVTCOREInternet Engineering Task Force (IETF) J. LennoxInternet-DraftRequest for Comments: 6904 Vidyo Updates: 3711(if approved) February 8,April 2013Intended status:Category: Standards TrackExpires: August 12, 2013ISSN: 2070-1721 Encryption of Header Extensions in the SecureReal-TimeReal-time Transport Protocol (SRTP)draft-ietf-avtcore-srtp-encrypted-header-ext-05Abstract The SecureReal-TimeReal-time Transport Protocol (SRTP) provides authentication, but not encryption, of the headers ofReal-TimeReal-time Transport Protocol (RTP) packets. However, RTP header extensions may carry sensitive information for which participants in multimedia sessions want confidentiality. This document provides a mechanism, extending the mechanisms of SRTP, to selectively encrypt RTP header extensions in SRTP. This document updates RFC 3711, the SecureReal-TimeReal-time Transport Protocol specification, to require that all future SRTP encryption transforms specify how RTP header extensions are to be encrypted. Status ofthisThis Memo ThisInternet-Draftissubmitted in full conformance with the provisions of BCP 78 and BCP 79. Internet-Drafts are working documentsan Internet Standards Track document. This document is a product of the Internet Engineering Task Force (IETF).Note that other groups may also distribute working documents as Internet-Drafts. The listIt represents the consensus ofcurrent Internet- Drafts is at http://datatracker.ietf.org/drafts/current/. Internet-Drafts are draft documents validthe IETF community. It has received public review and has been approved fora maximumpublication by the Internet Engineering Steering Group (IESG). Further information on Internet Standards is available in Section 2 of RFC 5741. Information about the current status ofsix monthsthis document, any errata, and how to provide feedback on it may beupdated, replaced, or obsoleted by other documentsobtained atany time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." This Internet-Draft will expire on August 12, 2013.http://www.rfc-editor.org/info/rfc6904. Copyright Notice Copyright (c) 2013 IETF Trust and the persons identified as the document authors. All rights reserved. This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License. Table of Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . ..3 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . .34 3. Encryption Mechanism . . . . . . . . . . . . . . . . . . . ..4 3.1. Example Encryption Mask . . . . . . . . . . . . . . . . . 6 3.2. Header Extension Keystream Generation for Existing Encryption Transforms . . . . . . . . . . . . . . . . . . 7 3.3. Header Extension Keystream Generation for Future Encryption Transforms . . . . . . . . . . . . . . . . . .78 4. Signaling (Setup) Information . . . . . . . . . . . . . . . .78 4.1. Backwardcompatibility .Compatibility . . . . . . . . . . . . . . . . . 9 5. Security Considerations . . . . . . . . . . . . . . . . . . .910 6. IANA Considerations . . . . . . . . . . . . . . . . . . . . .1011 7. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 11 8. References . . . . . . . . . . . . . . . . . . . . . . . . ..11 8.1. Normative References . . . . . . . . . . . . . . . . . ..11 8.2. Informative References . . . . . . . . . . . . . . . . ..12 Appendix A. Test Vectors . . . . . . . . . . . . . . . . . . . .1213 A.1. Keyderivation test vectorsDerivation Test Vectors . . . . . . . . . . . . . . .1213 A.2. Header Encryption Test VectorsusingUsing AES-CM . . . . . . .13 Appendix B. Changes From Earlier Versions . . . . . . . . . . . . 14 B.1. Changes from draft-ietf-avtcore -04 . . . . . . . . . . .14B.2. Changes from draft-ietf-avtcore -03 . . . . . . . . . . . 15 B.3. Changes from draft-ietf-avtcore -02 . . . . . . . . . . . 15 B.4. Changes from draft-ietf-avtcore -01 . . . . . . . . . . . 15 B.5. Changes from draft-ietf-avtcore -00 . . . . . . . . . . . 16 B.6. Changes from draft-lennox-avtcore -00 . . . . . . . . . . 16 B.7. Changes from draft-lennox-avt -02 . . . . . . . . . . . . 16 B.8. Changes From Individual Submission Draft -01 . . . . . . . 16 B.9. Changes From Individual Submission Draft -00 . . . . . . . 16 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . . 161. Introduction The SecureReal-TimeReal-time Transport Protocol [RFC3711] specification provides confidentiality, message authentication, and replay protection for multimedia payloads sent using theReal-TimeReal-time Protocol (RTP) [RFC3550]. However, in order to preserve RTP header compression efficiency, SRTP provides only authentication and replay protection for the headers of RTP packets, not confidentiality. For the standard portions of an RTP header,thisproviding only authentication and replay protection does not normally present a problem, as the information carried in an RTP header does not provide much information beyond that which an attacker could infer by observing the size and timing of RTP packets. Thus, there is little need for confidentiality of the header information. However,this is not necessarily truethe security requirements can be different for information carried in RTP header extensions. A number of recent proposals for header extensions using the mechanism described in "A General Mechanism for RTP HeaderExtensionsExtensions" [RFC5285] carry information for which confidentiality could be desired or essential. Notably, two recent specifications ([RFC6464] and [RFC6465])carrycontain information about per-packet sound levels of the media data carried in the RTPpayload,payload and specify that exposing this information to an eavesdropper is unacceptable in many circumstances (as described in therespective RFCs'Security Considerationssections).sections of those RFCs). This document, therefore, defines a mechanism by which encryption can be applied to RTP header extensions when they are transported using SRTP. As an RTP sender may wish some extension information to be sent in the clear (for example, it may be useful for a network monitoring device to be aware of RTP transmission time offsets [RFC5450]), this mechanism can be selectively applied to a subset of the header extension elements carried in an SRTP packet. The mechanism defined by this document encrypts packets' header extensions using the same cryptographic algorithms and parameters as are used to encrypt the packets' RTP payloads. This document defines how this is done for the encryption transforms defined in [RFC3711], [RFC5669], and [RFC6188], which are the SRTP encryption transforms defined bystandards-track IETF documentsStandards Track RFCs at the time of this writing. It also updates[RFC3711],[RFC3711] to indicate that specifications of future SRTP encryption transforms must define how header extension encryption is to be performed. 2. Terminology The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [RFC2119] and indicate requirement levels for compliant implementations. 3. Encryption Mechanism Encrypted header extension elements are carried in the same manner as non-encrypted header extension elements, as defined by [RFC5285]. The(one-one- ortwo-byte)two-byte header of the extension elements is not encrypted, nor is any of the header extension padding. If multiple different header extension elements are being encrypted, they have separate element identifier values, just as they would if they were notencrypted; similarly,encrypted. Similarly, encrypted and non-encrypted header extension elements have separate identifier values. Encrypted header extensionheaderselements areonlycarried only in packets encrypted using the SecureReal-TimeReal-time Transport Protocol [RFC3711]. To encrypt (or decrypt) encrypted header extensionheaders,elements, an SRTP participant first uses the SRTPKey Derivation Algorithm,key derivation algorithm, specified in Section 4.3.1 of [RFC3711], to generate header encryption and header salting keys, using the samepseudo-randompseudorandom function family asareis used for the key derivation for the SRTP session. These keys are derived as follows: o k_he (SRTP header encryption): <label> = 0x06, n=n_e. o k_hs (SRTP header salting key): <label> = 0x07, n=n_s. where n_e and n_s are from the cryptographic context: the same size encryption key and salting key are used as are used for the SRTP payload. Additionally, the same master key, master salt, index, and key_derivation_rate are used asare usedfor the SRTP payload. (Note that since RTP headers, includingextension headers,header extensions, are authenticated in SRTP, no new authentication key is needed forextension headers.)header extensions.) A header extension keystream is generated for each packet containing encrypted header extension elements. The details of how this header extension keystream is generated depend on the encryption transform that is used for the SRTP packet. For encryption transforms that have been standardized as of the date of publication of this document, see Section 3.2; for requirements for new transforms, see Section 3.3.OnceAfter the header extension keystream is generated, the SRTP participant then computes an encryption mask for the header extension, identifying the portions of the header extension that are, or are to be, encrypted. (For an example of this procedure, see Section3.1 below.)3.1.) This encryption mask corresponds to the entire payload of each header extension element that is encrypted. It does not include any non-encrypted header extension elements, any extension element headers, or any padding octets. The encryption mask has all-bits-1 octets (i.e., hexadecimal 0xff) for header extension octetswhichthat are to beencrypted,encrypted and all-bits-0 octets for header extension octetswhichthat are not tobe.be encrypted. The set of extension elements to be encrypted is communicated between the sender and the receiver using the signaling mechanisms described in Section 4. This encryption mask is computed separately for every packet that carries a header extension. Based on the non-encrypted portions of the headers and the signaled list of encrypted extension elements, a receiver can always determine the correct encryption mask for any encrypted header extension. The SRTP participant bitwise-ANDs the encryption mask with the keystream to produce a masked keystream. It then bitwiseexclusive- orsexclusive-ORs the header extension with this masked keystream to produce the ciphertext version of the header extension. (Thus, octets indicated as all-bits-1 in the encrypted mask are encrypted, whereas those indicated as all-bits-0 are not.) The header extension encryption process does not include the "defined by profile" or "length" fields of the header extension, only the field that[RFC3550]Section 5.3.1 of [RFC3550] calls "header extension" proper, starting with the first [RFC5285] ID and length. Thus, both the encryption mask and the keystream begin at this point. This header extension encryption process could, equivalently, be computed by considering the encryption mask as a mixture of the encrypted and unencrypted headers,i.e.i.e., as EncryptedHeader = (Encrypt(Key, Plaintext) AND MASK) OR (Plaintext AND (NOT MASK)) where Encrypt is the encryption function, MASK is the encryption mask, and AND, OR, and NOT are bitwise operations. This formulation of the encryption process might be preferred by implementations for which encryption is performed by a separatemodule,module and cannoteasilybemodified.modified easily. The SRTP authentication tag is computed across the encrypted header extension, i.e., the data that is actually transmitted on the wire. Thus, header extension encryption MUST be done before the authentication tag is computed, and authentication tag validation MUST be done on the encrypted header extensions. For receivers, header extension decryption SHOULD be done only after the receiver has validated the packet's message authentication tag, and the receiver MUST NOT take any actions based on decryptedheadersheaders, prior to validating the authentication tag, that could affect the security or proper functioning of thesystem, prior to validating the authentication tag.system. 3.1. Example Encryption Mask If a sender wished to send a header extension containing an encrypted SMPTE timecode [RFC5484] with ID 1, a plaintext transmission time offset [RFC5450] with ID 2, an encrypted audio level indication [RFC6464] with ID 3, and an encrypted NTPTimestamptimestamp [RFC6051] with ID 4, the plaintext RTP header extension might look like this: 0 1 2 3 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 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | ID=1 | len=7 | SMTPE timecode (long form) | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | SMTPE timecode (continued) | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | SMTPE (cont'd)| ID=2 | len=2 | toffset | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | toffset (ct'd)| ID=3 | len=0 | audio level | ID=4 | len=6 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | NTPTimestamptimestamp (Variant B) | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | NTPTimestamptimestamp (Variant B,cont.)cont'd) | padding = 0 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Figure11: Structure of Plaintext Example Header Extension The corresponding encryption mask would then be: 0 1 2 3 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 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |0 0 0 0 0 0 0 0|1 1 1 1 1 1 1 1|1 1 1 1 1 1 1 1|1 1 1 1 1 1 1 1| +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |1 1 1 1 1 1 1 1|1 1 1 1 1 1 1 1|1 1 1 1 1 1 1 1|1 1 1 1 1 1 1 1| +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |1 1 1 1 1 1 1 1|0 0 0 0 0 0 0 0|0 0 0 0 0 0 0 0|0 0 0 0 0 0 0 0| +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |0 0 0 0 0 0 0 0|0 0 0 0 0 0 0 0|1 1 1 1 1 1 1 1|0 0 0 0 0 0 0 0| +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |1 1 1 1 1 1 1 1|1 1 1 1 1 1 1 1|1 1 1 1 1 1 1 1|1 1 1 1 1 1 1 1| +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |1 1 1 1 1 1 1 1|1 1 1 1 1 1 1 1|1 1 1 1 1 1 1 1|0 0 0 0 0 0 0 0| +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Figure22: Encryption Mask for Example Header Extension In the mask, the octets corresponding to the payloads of the encrypted header extension elements are set to all-1 values, and the octets corresponding to non-encrypted header extension elements, element headers, and header extension padding are set toall-0all-zero values. 3.2. Header Extension Keystream Generation for Existing Encryption Transforms For the AES-CM and AES-f8 transforms [RFC3711], the SEED-CTR transform [RFC5669], and the AES_192_CM and AES_256_CM transforms [RFC6188], the header extension keystream SHALL be generated for each packet containing encrypted header extensionelements,elements using the same encryption transform and Initialization Vector (IV) asisare used for that packet's SRTP payload, except that the SRTP encryption and salting keys k_e and k_s are replaced by the SRTP header encryption and header salting keys k_he and k_hs, respectively, as definedabove, respectively.above. For the SEED-CCM and SEED-GCM transforms [RFC5669], the header extension keystream SHALL be generated using the algorithm specified above for the SEED-CTR algorithm. (Because theAEADAuthenticated Encryption with Associated Data (AEAD) transform used on the payload in these algorithms includes the RTP header, including the RTP header extension, in its Associated Authenticated Data (AAD), counter-mode encryption for the header extension is believed to be of equivalent cryptographic strength to the CCM and GCM transforms.) For the NULL encryption transform [RFC3711], the header extension keystream SHALL be all-zero. 3.3. Header Extension Keystream Generation for Future Encryption Transforms When new SRTP encryption transforms are defined, this document updates [RFC3711] as follows: in addition to the rules specified in Section 6 of RFC 3711, thestandard trackStandards Track RFC defining the new transform MUST specify how the encryption transform is to be used with header extension encryption. It is RECOMMENDED that new transformations follow the same mechanisms as are defined in Section3.2,3.2 of this document ifthesethey are applicable and are believed to be cryptographically adequate for the transform in question. 4. Signaling (Setup) Information Encrypted header extension elements are signaled in theSDPSession Description Protocol (SDP) extmapattribute,attribute using the URI"urn:ietf:params:rtp-hdrext:encrypt","urn:ietf:params:rtp-hdrext:encrypt" followed by the URI of the header extension element beingencryptedencrypted, as well as any extensionattributes that extension normally takes. Figure 3 gives a formal Augmented Backus-Naur Form (ABNF) [RFC5234] showing this grammar extension, extending the grammar defined in [RFC5285]. enc-extensionname = %x75.72.6e.3a.69.65.74.66.3a.70.61.72.61.6d.73.3a %x72.74.70.2d.68.64.72.65.78.74.3a.65.6e.63.72.79.70.74 ; "urn:ietf:params:rtp-hdrext:encrypt" in lower case extmap/==/ mapentry SP enc-extensionname SP extensionname [SP extensionattributes] ; extmap, mapentry,extensionnameextensionname, and extensionattributes ; are defined in [RFC5285] Figure 3: Syntax of the "encrypt" extmap Thus, for example, to signal an SRTP session using encrypted SMPTE timecodes [RFC5484], while simultaneously signaling plaintext transmission time offsets [RFC5450], an SDP document could contain the text shown in Figure 4 (line breaks have been added forformatting):formatting). m=audio 49170 RTP/SAVP 0 a=crypto:1 AES_CM_128_HMAC_SHA1_32 \ inline:NzB4d1BINUAvLEw6UzF3WSJ+PSdFcGdUJShpX1Zj|2^20|1:32 a=extmap:1 urn:ietf:params:rtp-hdrext:encrypt \ urn:ietf:params:rtp-hdrext:smpte-tc 25@600/24 a=extmap:2 urn:ietf:params:rtp-hdrext:toffset Figure44: Sample SDP Document Offering Encrypted Headers This example uses SDPSecurity Descriptionssecurity descriptions [RFC4568] for SRTP keying, but this is merely forillustration; anyillustration. Any SRTP keying mechanism to establish session keys will work. The extmap SDP attribute is defined in [RFC5285] as being either a session or media attribute. If the extmap for an encrypted header extension is specified as a media attribute, it MUSTonlybe specified only for mediawhichthat use SRTP-based RTP profiles. If such an extmap is specified as a session attribute, there MUST be at least one media in the SDP sessionwhichthat uses an SRTP-based RTPprofile; the session-levelprofile. The session- level extmap applies to all the SRTP-based media in thesession,session and MUST be ignored for all other (non-SRTP or non-RTP) media. The "urn:ietf:params:rtp-hdrext:encrypt" extension MUST NOT be recursively applied to itself. 4.1. BackwardcompatibilityCompatibility Following the procedures in [RFC5285], an SDP endpointwhichthat does not understand the "urn:ietf:params:rtp-hdrext:encrypt" extension URI will ignore theextension, and (forextension and, for SDPoffer/answer)offer/answer, will negotiate not to use it. For backward compatibility with endpointswhichthat do not implement this specification, in a negotiated session (whether using offer/answer or some other means), best-effort encryption of a header extension element is possible: an endpoint MAY offer the same header extension element both encrypted and unencrypted. An offerer MUSTonlyoffer only best-effort negotiation when lack of confidentiality would be acceptable in the backward-compatible case. Answerers (or equivalent peers in a negotiation)whichthat understand header extension encryption SHOULD choose the encrypted form of the offered header extensionelement,element and mark the unencrypted form "inactive", unless they have an explicit reason to prefer the unencrypted form. In all cases, answerers MUST NOT negotiate the use of, and senders MUST NOT send, both encrypted and unencrypted forms of the same header extension. Note that, as always, users of best-effort encryption MUST be cautious of bid-down attacks, where a man-in-the-middle attacker removes a higher-security option, forcing endpoints to negotiate a lower-security one. Appropriate countermeasures depend on the signaling protocol in use, but users can ensure, for example, that signaling is integrity-protected. 5. Security Considerations The security properties of header extension elements protected by the mechanism in this document are equivalent to those for SRTP payloads. The mechanism defined in this document does not provide confidentiality about which header extension elements are used for a given SRTP packet, only for the content of those header extension elements. This appears to be in the spirit of SRTP itself, which does not encrypt RTP headers. If this is a concern, an alternate mechanism would be needed to provide confidentiality. For the two-byte-header form of header extension elements(0x100x),(0x100N, where "N" is the appbits field), this mechanism does not provide any protection to zero-length header extension elements (for which their presence or absence is the only information they carry). It also does not provide any protection for thetwo-byte-headers' app bitsappbits (field 256, the lowest four bits of the "defined by profile"field).field) of the two-byte headers. Neither of these featuresareis present inforthe one-byte-header form of header extension elements (0xBEDE), so these limitations do not apply in that case. This mechanism cannot protect RTP header extensionswhichthat do not use the mechanism defined in [RFC5285]. This document does not specify the circumstances in which extension header encryption should be used. Documents defining specific header extension elements should provide guidance on when encryption is appropriate for these elements. If a middlebox does not have access to the SRTP authentication keys, it has no way to verify the authenticity of unencrypted RTP header extension elements (or the unencrypted RTP header), even though it can monitor them. Therefore, such middleboxes MUST treat such headers as untrusted and potentially generated by an attacker, in the same way as they treat unauthenticated traffic. (This does not mean that middleboxes cannot view and interpret such traffic, of course, only that appropriate skepticism needs to be maintained about the results of suchinterpretation.).interpretation.) There is no mechanism defined to protect header extensions with different algorithms or encryption keys than are used to protect the RTP payloads. In particular, it is not possible to provide confidentiality for a header extension while leaving the payload in cleartext. The dangers of using weak or NULL authentication with SRTP, described in[RFC3711]Section9.5,9.5 of [RFC3711], apply to encrypted header extensions as well. In particular, since some header extension elements will have someeasily-guessedeasily guessed plaintext bits, strong authentication is REQUIRED if an attacker setting such bits could have a meaningful effect on the behavior of the system. The technique defined in this document canonlybe applied only to encryption transforms that work by generating a pseudorandom keystream and bitwiseexclusive-oringexclusive-ORing it with the plaintext, such as CTR or f8. It will not work with ECB, CBC, or any other encryption method that does not use a keystream. 6. IANA Considerations This document defines a new extension URI to the RTP Compact Header Extensions subregistry of the Real-Time Transport Protocol (RTP) Parameters registry, according to the following data: Extension URI: urn:ietf:params:rtp-hdrext:encrypt Description: Encryptedextensionheader extension element Contact: jonathan@vidyo.com Reference: RFCXXXX (Note to the RFC-Editor: please replace "XXXX" with the number of this document prior to publication as an RFC.)6904 7. Acknowledgments Thanks to Benoit Claise, Roni Even, Stephen Farrell, Kevin Igoe, Joel Jaeggli, David McGrew,Magnus Westerlund,David Singer, Robert Sparks, Magnus Westerlund, Qin Wu, and Felix Wyss for their comments and suggestions in the development of this specification. 8. References 8.1. Normative References [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. [RFC3550] Schulzrinne, H., Casner, S., Frederick, R., and V. Jacobson, "RTP: A Transport Protocol for Real-Time Applications", STD 64, RFC 3550, July 2003. [RFC3711] Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K. Norrman, "The Secure Real-time Transport Protocol (SRTP)", RFC 3711, March 2004. [RFC5234] Crocker, D. and P. Overell, "Augmented BNF for Syntax Specifications: ABNF", STD 68, RFC 5234, January 2008. [RFC5285] Singer, D. and H. Desineni, "A General Mechanism for RTP Header Extensions", RFC 5285, July 2008. [RFC5669] Yoon, S., Kim, J., Park, H., Jeong, H., and Y. Won, "The SEED Cipher Algorithm and Its Use with the Secure Real- Time Transport Protocol (SRTP)", RFC 5669, August 2010. [RFC6188] McGrew, D., "The Use of AES-192 and AES-256 in Secure RTP", RFC 6188, March 2011. 8.2. Informative References [RFC4568] Andreasen, F., Baugher, M., and D. Wing, "Session Description Protocol (SDP) Security Descriptions for Media Streams", RFC 4568, July 2006. [RFC5450] Singer, D. and H. Desineni, "Transmission Time Offsets in RTP Streams", RFC 5450, March 2009. [RFC5484] Singer, D., "Associating Time-Codes with RTP Streams", RFC 5484, March 2009. [RFC6051] Perkins, C. and T. Schierl, "Rapid Synchronisation of RTP Flows", RFC 6051, November 2010. [RFC6464] Lennox, J., Ivov, E., and E. Marocco, "A Real-time Transport Protocol (RTP) Header Extension for Client-to- Mixer Audio Level Indication", RFC 6464, December 2011. [RFC6465] Ivov, E., Marocco, E., and J. Lennox, "A Real-time Transport Protocol (RTP) Header Extension for Mixer-to- Client Audio Level Indication", RFC 6465, December 2011. Appendix A. Test Vectors A.1. Keyderivation test vectorsDerivation Test Vectors This section provides test data for the header extension key derivation function, using AES-128 in Counter Mode. (The algorithms and keys used are the same as those for thethetest vectors in Appendix B.3 of [RFC3711].) The inputs to the key derivation function are the16 octet16-octet master key and the14 octet14-octet master salt: master key: E1F97A0D3E018BE0D64FA32C06DE4139 master salt: 0EC675AD498AFEEBB6960B3AABE6 Following [RFC3711], the input block for AES-CM is generated byexclusive-oringexclusive-ORing the master salt with the concatenation of the encryption key label 0x06 with (index DIV kdr), then padding on the right with two nulloctets (whichoctets, which implements the multiply-by-2^16operation, seeoperation (see Section 4.3.3 of [RFC3711]). The resulting value is thenAES-CM- encryptedAES-CM-encrypted using the master key to get the cipher key. index DIV kdr: 000000000000 label: 06 master salt: 0EC675AD498AFEEBB6960B3AABE6 --------------------------------------------------xor:XOR: 0EC675AD498AFEEDB6960B3AABE6 (x, PRF input) x*2^16: 0EC675AD498AFEEDB6960B3AABE60000 (AES-CM input) hdr. cipher key: 549752054D6FB708622C4A2E596A1B93 (AES-CM output) Next, we show how the cipher salt is generated. The input block for AES-CM is generated byexclusive-oringexclusive-ORing the master salt with the concatenation of the encryption salt label. That value is padded and encrypted as above. index DIV kdr: 000000000000 label: 07 master salt: 0EC675AD498AFEEBB6960B3AABE6 --------------------------------------------------xor:XOR: 0EC675AD498AFEECB6960B3AABE6 (x, PRF input) x*2^16: 0EC675AD498AFEECB6960B3AABE60000 (AES-CM input) AB01818174C40D39A3781F7C2D270733 (AES-CM ouptut) hdr. cipher salt: AB01818174C40D39A3781F7C2D27 A.2. Header Encryption Test VectorsusingUsing AES-CM This section provides test vectors for the encryption of a headerextension,extension using the AES_CM cryptographic transform. The header extension is encrypted using the header cipher key and header cipher salt computed in Appendix A.1. The header extension is carried in an SRTP-encrypted RTP packet with SSRC 0xCAFEBABE, sequence number 0x1234, and an all-zero rollover counter. Session Key: 549752054D6FB708622C4A2E596A1B93 Session Salt: AB01818174C40D39A3781F7C2D27 SSRC: CAFEBABE Rollover Counter: 00000000 Sequence Number: 1234 ---------------------------------------------- Init. Counter: AB018181BE3AB787A3781F7C3F130000 The SRTP session was negotiated to indicate that header extension ID values 1,33, and 4 are encrypted. In hexadecimal, the header extension being encrypted is as follows (spaces have been added to show the internal structure of the header extension): 17 414273A475262748 22 0000C8 30 8E 46 55996386B395FB 00 This header extension is 24 bytes long. (Its values are intended to represent plausible values of the header extension elements shown in Section 3.1, but their specific meaning is not important for the example.) The header extension "defined by profile" and "length" fields, which in this case are BEDE 0006 in hexadecimal, are not included in the encryption process. In hexadecimal, the corresponding encryption mask selecting the bodies of header extensions 1, 2, and 4 (corresponding to the mask in Figure 2) is: 00 FFFFFFFFFFFFFFFF 00 000000 00 FF 00 FFFFFFFFFFFFFF 00 Finally, we compute the keystream from the session key and the initial counter, apply the mask to the keystream, and thenxorexclusive- OR the keystream with the plaintext: Initial keystream: 1E19C8E1D481C779549ED1617AAA1B7A FC0D933AE7ED6CC8 Mask(Hex):(hex): 00FFFFFFFFFFFFFFFF0000000000FF00 FFFFFFFFFFFFFF00 Masked keystream: 0019C8E1D481C7795400000000001B00 FC0D933AE7ED6C00 Plaintext: 17414273A475262748220000C8308E46 55996386B395FB00 Ciphertext: 17588A9270F4E15E1C220000C8309546 A994F0BC54789700Appendix B. Changes From Earlier Versions Note to the RFC-Editor: please remove this section prior to publication as an RFC. B.1. Changes from draft-ietf-avtcore -04 o Clarified that simultaneous offer of encrypted and unencrypted headers is only to be used for backward compatibility, and that endpoints must never actually negotiate or send encrypted and unencrypted versions of the same header extension simultaneously. o Clarified that the same master key, master salt, index, and key derivation rate are to be used for the header keys and salt as for the payload keys. o Added a paragraph to the security consideration emphasizing the dangers of weak or NULL authentication. o Editorial changes. o Added Benoit Claise, Stephen Farrell, and Joel Jaeggli to the Acknowledgments. B.2. Changes from draft-ietf-avtcore -03 o Modified the ABNF syntax to avoid rule recursion. o Added Robert Sparks to the Acknowledgments. B.3. Changes from draft-ietf-avtcore -02 o Clarified that the header extension encryption mask must be calculated separately for each packet, and can always be derived from the plaintext portions of the encrypted header extension. o Presented an alternate formulation of the header extension encryption process, so implementations can use their existing encryption algorithms unmodified. o Added a security consideration emphasizing that this mechanism must only be used with keystream-based encryption algorithms. B.4. Changes from draft-ietf-avtcore -01 o Made the draft update RFC 3711, and added a section specifying that all future SRTP encryption transforms must specify how header extension encryption is to be done. o Explicitly distinguished the processing of existing encryption transforms from future ones. o Clarified description of the process by which the encryption mask is applied, and that encryption does not apply to the header extension "defined by profile" or "length" fields. o Defined how header extension encryption is to be done with the SEED algorithms defined in RFC 5669, and with the NULL algorithm. o Added ABNF grammar for the SDP syntax. o Clarified that header extension encryption must not be applied to itself. o Expanded discussion of bid-down attacks. o Pointed out that this mechanism can't protect non-RFC5285 header extensions, and that there's no way to give different protection to header extensions than to payloads. o Updated references to now-published RFCs. o Editorial clarifications. o Added Magnus Westerlund to the Acknowledgments. B.5. Changes from draft-ietf-avtcore -00 o Clarified usage of Key Derivation Algorithm o Provided non-normative guidance for how to use this mechanism with Authenticated Encryption with Associated Data (AEAD) transforms. o Corrected SMPTE Timecode header extension element in example header extension (it's eight bytes, not sixteen). Added an NTP timestamp to the example to fill it back out to original size. o Specified applicability of the extmap attribute if it's specified as a session-level attribute. o Added description of backward compatibility, including a description of how you can negotiate best-effort encryption. o Added a note to the security considerations about the dangers for middleboxes observing unencrypted headers (both header extension elements and RTP headers) without being able to verify the authentication keys. o Added test vectors. o Added acknowledgments section. B.6. Changes from draft-lennox-avtcore -00 o Published as working group item. o Added discussion of limitations when used with the two-byte-header form of header extension elements. o Added open issue about how to use this mechanism with Authenticated Encryption with Associated Data (AEAD) transforms. o Updated references. B.7. Changes from draft-lennox-avt -02 o Retargeted at AVTCORE working group. o Updated references. B.8. Changes From Individual Submission Draft -01 o Minor editorial changes. B.9. Changes From Individual Submission Draft -00 o Clarified description of encryption mask creation. o Added example encryption mask. o Editorial changes.Author's Address Jonathan Lennox Vidyo, Inc. 433 Hackensack Avenue Seventh Floor Hackensack, NJ 07601 USEmail:EMail: jonathan@vidyo.com