EAP Working Group Bernard Aboba INTERNET-DRAFT Dan Simon Category: Informational Microsoft 2 March 2003 EAP Keying Framework Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC 2026. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet- Drafts. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." The list of current Internet-Drafts can be accessed at http://www.ietf.org/ietf/1id-abstracts.txt The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html. Copyright Notice Copyright (C) The Internet Society (2003). All Rights Reserved. Abstract This document describes the framework for EAP key derivation, and provides guidelines for generation and usage of EAP keys by EAP methods and AAA protocols. Algorithms for key derivation or mechanisms for key transport are not specified in this document. Rather, this document provides a framework within which derivation algorithms and transport mechanisms can be discussed and evaluated. Aboba & Simon Informational [Page 1] INTERNET-DRAFT EAP Keying Framework 2 March 2003 Table of Contents 1. Introduction .......................................... 3 1.1 Requirements language ........................... 3 1.2 Terminology ..................................... 4 2. EAP overview .......................................... 5 2.1 Invariants ...................................... 7 3. EAP key hierarchy ..................................... 8 3.1 Exchanges ....................................... 11 4. Security properties ................................... 14 4.1 EAP method requirements ......................... 14 4.2 AAA protocol requirements ....................... 17 4.3 TSK derivation requirements ..................... 18 4.4 Ciphersuite requirements ........................ 19 4.5 Security properties ............................. 19 5. Security considerations ............................... 21 5.1 Assumptions ..................................... 21 5.2 Key binding .................................... 22 5.3 Key strength .................................... 23 5.4 Key wrap ........................................ 23 5.5 Man-in-the-middle attacks ....................... 24 6. Normative references .................................. 24 7. Informative references ................................ 25 Appendix A - Ciphersuite independence ........................ 29 Appendix B - Ciphersuite keying requirements ................. 30 Appendix C - Example TEK hierarchy ........................... 31 Appendix D - Example MSK, EMSK and IV hierarchy .............. 32 Appendix E - Example TSK derivation .......................... 34 Appendix F - Example PMK derivation .......................... 35 Acknowledgments .............................................. 35 Author's Addresses ........................................... 35 Intellectual Property Statement .............................. 36 Full Copyright Statement ..................................... 36 Aboba & Simon Informational [Page 2] INTERNET-DRAFT EAP Keying Framework 2 March 2003 1. Introduction The Extensible Authentication Protocol (EAP), defined in [RFC2284bis], was originally developed to provide extensible authentication for use with PPP [RFC1661]. Since then, it has also been applied to IEEE 802 wired networks [IEEE8021X]. When EAP is used for authentication on PPP or wired IEEE 802 networks, it is typically assumed that the link is physically secure, so that an attacker cannot gain access to the link, or insert a rogue device. EAP methods defined in [RFC2284bis] reflect this usage model. These include EAP MD5, as well as One-Time Password (OTP) and Generic Token Card. These methods support one-way authentication (from EAP peer to authenticator) but not mutual authentication or key derivation. As a result, these methods do not bind the initial authentication and subsequent data traffic, even when the the ciphersuite used to protect data supports per-packet authentication and integrity protection. This leaves these methods vulnerable to hijacking as well as attacks by rogue devices. On wireless networks such as IEEE 802.11 [IEEE80211], these attacks become much easier to mount, since any attacker within range is capable of accessing the wireless medium, or acting as an access point. As a result, new ciphersuites have been proposed for use with wireless LANs [IEEE80211i] which provide per-packet authentication, integrity and replay protection. In addition, mutual authentication and key derivation, provided by methods such as EAP TLS [RFC2716] are required [IEEE80211i], so as to address the threat of rogue devices, and provide keying material to bind the initial authentication to subsequent data traffic. Section 2 provides an overview of EAP. Section 3 describes the EAP key hierarchy. Section 4 describes requirements and the resulting security properties. Section 5 discusses additional security considerations. Appendix A discusses the principle of ciphersuite independence. Appendix B provides a summary of the keying requirements of link layer ciphersuites supported on PPP and IEEE 802.11. Appendix C provides an example EAP Master Key (MK) hierarchy. Appendix D provides an example Master Session Key (MSK) hierarchy. Appendix E provides an example Transient Session Key (TSK) derivation. Appendix F provides an example of PMK derivation in Fast Handoff. 1.1. Requirements language In this document, several words are used to signify the requirements of the specification. These words are often capitalized. The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD Aboba & Simon Informational [Page 3] INTERNET-DRAFT EAP Keying Framework 2 March 2003 NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119]. 1.2. Terminology This document frequently uses the following terms: authenticator The end of the EAP link initiating EAP authentication. Where no backend authentication server is present, the authenticator acts as the EAP server, terminating the EAP conversation with the peer. Where a backend authentication server is present, the authenticator MAY act as a pass-through for one or more authentication methods and for non-local users. This terminology is also used in [IEEE8021X], and has the same meaning in this document. backend authentication server A backend authentication server is an entity that provides an authentication service to an authenticator. When used, this server typically executes EAP Methods for the authenticator. This terminology is also used in [IEEE8021X]. AN-Token The package within which the Master Session Key (MSK) and one or more attributes is transported between the backend authentication server and the authenticator. The attributes provide the authenticator with information on MSK usage. For example, attributes might include the peer layer 2 address, the authenticator layer 2 and IP addresses, the MSK lifetime, etc. The format and wrapping of the AN-Token, which is intended to be accessible only to the backend authentication server and authenticator, is defined by the AAA key distribution specification. Cryptographic binding The demonstration of the EAP peer to the EAP server that a single entity has acted as the EAP peer for all methods executed within a sequence or tunnel. Binding MAY also imply that the EAP server demonstrates to the peer that a single entity has acted as the EAP server for all methods executed within a sequence or tunnel. If executed correctly, binding serves to mitigate man-in-the-middle vulnerabilities. Cryptographic separation Two keys (x and y) are "cryptographically separate" if an adversary that knows all messages exchanged in the protocol cannot compute x from y or y from x without "breaking" some cryptographic assumption. In particular, this definition allows that the Aboba & Simon Informational [Page 4] INTERNET-DRAFT EAP Keying Framework 2 March 2003 adversary has the knowledge of all nonces sent in cleartext as well as all predictable counter values used in the protocol. Breaking a cryptographic assumption would typically require inverting a one- way function or predicting the outcome of a cryptographic pseudo- random number generator without knowledge of the secret state. In other words, if the keys are cryptographically separate, there is no shortcut to compute x from y or y from x, but the work an adversary must do to perform this computation is equivalent to performing exhaustive search for the secret state value. Key derivation This refers to the ability of the EAP method to derive a Master Key (MK) which is not exported, as well as a ciphersuite- independent Master Session Key (MSK), Extended Master Session Key (EMSK) and Initialization Vector (IV). The MSK, EMSK and IV are used only for further key derivation, not directly for protection of the EAP conversation or subsequent data. Key strength If the effective key strength is N bits, the best currently known methods to recover the key (with non-negligible probability) require an effort comparable to 2^N operations of a typical block cipher. EAP server The entity that terminates the EAP authentication with the peer. In the case where there is no backend authentication server, this term refers to the authenticator. Where the authenticator operates in pass-through, it refers to the backend authentication server. Mutual authentication This refers to an EAP method in which, within an interlocked exchange, the authenticator authenticates the peer and the peer authenticates the authenticator. Two one-way conversations, running in opposite directions do not provide mutual authentication as defined here. peer The end of the EAP Link that responds to the authenticator. In [IEEE8021X], this end is known as the Supplicant. 2. EAP overview The EAP authentication process involves a peer, authenticator and (optionally) a backend authentication server. Typically, the peer desires access to the network, and the authenticator is a Network Access Server (NAS) providing that access. However, EAP may also be used in other situations, such as when it is desired for two network devices (e.g. two switches or routers) to authenticate each other. Since EAP is Aboba & Simon Informational [Page 5] INTERNET-DRAFT EAP Keying Framework 2 March 2003 a peer-to-peer protocol, an independent and simultaneous authentication may take place in the reverse direction. Both peers may act as authenticators and authenticatees at the same time. An important goal of EAP is to enable deployment of new methods without requiring development of new code on the authenticator. While the authenticator may implement some EAP methods locally and use those methods to authenticate local users, it may at the same time act as a pass-through for other users and methods, forwarding EAP packets back and forth between the backend authentication server and the peer. EAP presumes that prior to initiation of authentication, the EAP peer has located the authenticator, using an out-of-band mechanism. For example, for use with PPP, the client might be configured with a phone book providing phone numbers for accessing the selected service. For use with IEEE 802.11 wireless LANs, the peer (a Station (STA) in IEEE 802.11 terminology) may locate an authenticator (an Access Point (AP) in IEEE 802.l1 terminology) using the IEEE 802.11 Beacon and Probe Request/Response frames. Since service location is handled out of band, this functionality is not provided within EAP. EAP supports either one-way authentication (in which the peer authenticates to the EAP server), or mutual authentication (in which the peer and EAP server mutually authenticate). In either case, it can be assumed that the parties do not enable the link unless their authentication requirements have been met. For example, a peer completing mutual authentication with an authenticator will not enable its link until the authenticator has authenticated successfully to the peer. As described in Section 3, EAP methods MAY support derivation of keying material used for purposes including protection of the EAP conversation and subsequent data exchanges, man-in-the-middle detection, or fast handoff. EAP methods supporting key derivation must also support mutual authentication. EAP assumes that ciphersuite negotiation, if it occurs, is handled out of band. For example, the peer might be preconfigured with policy indicating the ciphersuite to be used in communicating with a given authenticator, or alternatively, the link layer protocol may support ciphersuite negotiation. Within PPP, the ciphersuite is negotiated within the Encryption Control Protocol (ECP), after EAP authentication is completed. Within [IEEE80211i], the AP capabilities (including ciphersuite) are advertised in the Beacon and Probe Responses, and are securely verified during a 4-way exchange after EAP authentication has completed. The desired ciphersuite is indicated within the Association/Reassociation Request/Response exchange. Aboba & Simon Informational [Page 6] INTERNET-DRAFT EAP Keying Framework 2 March 2003 2.1. Invariants Several basic principles govern the design of the EAP keying framework. These are known as the "EAP Invariants": Media independence As described in [RFC2284bis], EAP authentication is supported on lower layers, including PPP [RFC1661] and IEEE 802 wired networks [IEEE8021X]. Use with IEEE 802.11 wireless LANs is also contemplated [IEEE80211i]. Since EAP methods cannot be assumed to have knowledge of the lower layer on which they are being run, EAP methods MUST be designed to function on any lower layer meeting the criteria outlined in [RFC2284bis], Section 3.1. Ciphersuite independence Since ciphersuite negotiation occurs out-of-band of EAP, and may occur after EAP authentication and key derivation is complete, EAP methods deriving keys MUST provide keying material that is independent of the ciphersuite subsequently negotiated for protection of data. Since it is the peer and authenticator that negotiate and implement the ciphersuite, knowledge of the ciphersuite is restricted to those entities. The backend authentication server is not a party to the ciphersuite negotiation nor is it an intermediary in the data flow between the peer and authenticator. As a result, it cannot be assumed to have knowledge of the ciphersuites implemented by the peer and authenticator, to be aware of the ciphersuite negotiated between them, or to implement ciphersuite-specific code. Since the backend authentication server may not know the ciphersuite negotiated between the peer and authenticator, it cannot make this information available to a resident EAP method. This means that ciphersuite-specific key generation, if implemented within an EAP method, will not function correctly on every EAP implementation. The advantages of ciphersuite independence are discussed in Appendix A. Method independence Supporting pass-through of authentication to the backend authentication server enables the authenticator to support any authentication method implemented on the backend authentication server and peer, not just locally implemented methods. This implies that the authenticator need not implement code for each EAP method required by authenticating peers; in fact the authenticator is not required to implement any EAP methods at all, nor cannot it be assumed to implement code specific to any EAP method. Aboba & Simon Informational [Page 7] INTERNET-DRAFT EAP Keying Framework 2 March 2003 This is useful where there is no single EAP method that is both mandatory-to-implement and offers acceptable security for the media in use. For example, the [RFC2284bis] mandatory-to-implement EAP method (MD5-Challenge) does not provide dictionary attack resistance, mutual authentication or key derivation, and as a result is not appropriate for use with IEEE 802.11 wireless LANs. 3. EAP key hierarchy The EAP keying hierarchy, illustrated in Figure 1, makes use of the following types of keys: EAP Master key (MK) A key derived between the EAP client and server during the EAP authentication process that is purely local to the EAP method. The MK MUST NOT be exported from the EAP method or be made available to a third party. Since derivation of the MK is a residue of the successful completion of the EAP authentication exchange, proof of MK possession may be used to shorten future EAP exchanges between the same EAP client and server, a technique known as "fast resume". Master Session Key (MSK) Keying material (64 octets) that is derived between the EAP client and server. The MSK is used in the derivation of Transient Session Keys (TSKs) for the ciphersuite negotiated between the EAP peer and authenticator. Where a backend authentication server is present, acting as an EAP server, it will typically transport the MSK to the authenticator. The MSK differs from the MK in that it not assumed to remain local to the EAP method, and is known by all parties in the EAP exchange: the peer, authenticator and the authentication server (if present). The MSK MAY be derived from the MK via a one-way function, or it may be an independent quantity. However possession of the MSK MUST NOT provide any information useful in determining the MK. An example, MSK, EMSK and IV key derivation is given in Appendix D. Extended Master Session Key (EMSK) Additional keying material (64 octets) derived between the EAP client and server that is not assumed to remain local to the EAP method. However, unlike the MSK, the EMSK is known only to the EAP client and server and MUST NOT be provided to a third party. The EMSK therefore MUST NOT be transported by the backend authentication server to the authenticator. The EMSK is reserved for future uses that are not defined yet. For example, it could be used to derive additional keying material for purposes such as fast handoff, man-in-the-middle vulnerability protection, etc. An example of fast handoff key derivation is given in Appendix F. Aboba & Simon Informational [Page 8] INTERNET-DRAFT EAP Keying Framework 2 March 2003 Initialization Vector (IV) A 64 octet quantity suitable for use in an initialization vector field, that is derived between the EAP client and server. Since in some EAP methods such as [RFC2716] the IV is a known value, the IV MUST NOT be used in computation of any quantity that needs to remain secret. Pairwise Master Key (PMK) In [RFC2716], the MSK is divided into two halves, corresponding to the "Peer to Authenticator Encryption Key" (Enc-RECV-Key, 32 octets) and "Authenticator to Peer Encryption Key" (Enc-SEND-Key, 32 octets) (reception is defined from the point of view of the authenticator). Within [IEEE80211i] Octets 0-31 of the MSK (Enc- RECV-Key) are also known as the Pairwise Master Key (PMK). [IEEE80211i] ciphersuites derive their Transient Session Keys (TSKs) solely from the PMK, whereas the WEP ciphersuite, when used with [IEEE8021X], as noted in [Congdon], derives its TSKs from both halves of the MSK, the Enc-RECV-Key and the Enc-SEND-Key. Transient EAP Keys (TEKs) Session keys which are used to establish a protected channel between the EAP peer and server during the EAP authentication exchange. The TEKs are typically derived from the MK, and are appropriate for use with the ciphersuite negotiated between EAP peer and server as part the EAP authentication exchange. Note that the ciphersuite used to set up the protected channel between the EAP peer and server during EAP authentication is unrelated to the ciphersuite used to subsequently protect data sent between the EAP peer and authenticator. In particular, the TEKs used to protect the EAP exchange MUST be cryptographically separate from TSKs used to protect data. An example TEK key hierarchy is described in Appendix C. Transient Session Keys (TSKs) Session keys used to protect data which are appropriate for the ciphersuite negotiated between the EAP peer and authenticator. The TSKs are derived from the MSK by a process which is link layer specific. In the case of IEEE 802.11, TSK derivation is supported via a 4-way handshake that supports mutual authentication between the EAP peer and authenticator. The 4-handshake also confirms mutual possession of the PMK as well as supporting protected ciphersuite negotiation. An example TSK derivation is given in Appendix E. Aboba & Simon Informational [Page 9] INTERNET-DRAFT EAP Keying Framework 2 March 2003 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ---+ | | ^ | EAP Method | | | | | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | | | | | | | | | | | | | | | | | | EAP Master Key (MK) | | | | | Derivation | | | | | | | Local to | | | | | EAP | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Method | | | | | | | | | | | | | MK | | | | | | | | | V | | | | | +-+-+-+-+-+-+ +-+-+-+-+-+-+ +-+-+-+-+-+-+ +-+-+-+-+-+-+ | | | | TEK | | MSK | |EMSK | |IV | | | | |Derivation | |Derivation | |Derivation | |Derivation | | | | +-+-+-+-+-+-+ +-+-+-+-+-+-+ +-+-+-+-+-+-+ +-+-+-+-+-+-+ | | | | | | | | | | | | | V +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ---+ | | | ^ | | | | | MSK (64B) | EMSK (64B) | IV (64B) | | | | Exported| | | | by | | V V EAP | | +-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+ Method| | | Reserved | | Known | | | | | |(Not Secret) | | | +-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+ V | ---+ | Transported | | by AAA | | Protocol | V V +-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ---+ | | ^ | TSK | Ciphersuite | | Derivation | Specific | | | V +-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ---+ Figure 1 - EAP Key Hierarchy Aboba & Simon Informational [Page 10] INTERNET-DRAFT EAP Keying Framework 2 March 2003 3.1. Exchanges Figure 2 illustrates the EAP exchange in the case where no backend authentication server is present. Here EAP is spoken between the peer and authenticator, encapsulated within a lower layer protocol, such as PPP, defined in [RFC1661] and IEEE 802, defined in [IEEE802]. Since the authenticator acts as an endpoint of the EAP conversation rather than a pass-through, EAP methods are implemented on the authenticator as well as the peer. If the EAP method negotiated between the EAP peer and authenticator supports mutual authentication and key derivation, an EAP Master Key (MK) is derived on the EAP peer and authenticator and stored locally within the EAP method. The MK may then be used to derive Transient EAP Keys (TEKs) used to protect some or all of the EAP exchange. The TEKs are also stored locally within the EAP method and are not exported. Once mutual authentication completes and is successful, the peer and authenticator exchange data, which may be protected using a ciphersuite. In order to provide keys for the ciphersuite, Transient Session Keys (TSKs) are required. On completion of a successful authentication, EAP methods on the peer and authenticator export the Master Session Key (MSK), Extended Master Session Key (EMSK) and IV. Of these quantities, only the MSK is used today, for derivation of the TSKs. The mechanism for this is specific to the ciphersuite; for example, in [IEEE80211i], the 4-way handshake is used. Where no backend authentication server is present, the MSK and EMSK are known only to the peer and authenticator and neither is transported to a third party. As demonstrated in [RoamCERT], despite the absence of a backend authentication server, such exchanges can support roaming between providers; it is even possible to support fast handoff [IEEE80211f] without re-authentication. However, this is typically only possible where both the EAP peer and authenticator support certificate- based authentication, or where the user base is sufficiently small that EAP authentication can occur locally. Where these conditions cannot be met, a backend authentication server is typically required. In this exchange, as described in [RFC2869bis], the authenticator acts as a pass-through between the EAP peer and a backend authentication server. In this model, the authenticator delegates the access control decision to the backend authentication server, which acts as a Key Distribution Center (KDC), supplying keying material to both the EAP peer and authenticator. Aboba & Simon Informational [Page 11] INTERNET-DRAFT EAP Keying Framework 2 March 2003 +-+-+-+-+-+ +-+-+-+-+-+ | | | | | | | | | Cipher- | | Cipher- | | Suite | | Suite | | | | | +-+-+-+-+-+ +-+-+-+-+-+ ^ ^ | | | | | | V V +-+-+-+-+-+ +-+-+-+-+-+ | | | | | |===============| | | |EAP, TEK Deriv.|Authenti-| | |<------------->| cator | | | | | | | TSK Deriv. | | | peer |<------------->| (EAP | | |===============| server) | | | Link layer | | | | (PPP,IEEE802) | | | | | | |MSK,EMSK | |MSK,EMSK | | (TSKs) | | (TSKs) | | | | | +-+-+-+-+-+ +-+-+-+-+-+ ^ ^ | | | MSK, EMSK, IV | MSK, EMSK, IV | | | | +-+-+-+-+-+ +-+-+-+-+-+ | | | | | EAP | | EAP | | Method | | Method | | | | | |(MK,TEKs)| |(MK,TEKs)| | | | | +-+-+-+-+-+ +-+-+-+-+-+ Figure 2 - Relationship between EAP peer and authenticator (acting as an EAP server), where no backend authentication server is present. Aboba & Simon Informational [Page 12] INTERNET-DRAFT EAP Keying Framework 2 March 2003 +-+-+-+-+-+ +-+-+-+-+-+ | | | | | | | | | Cipher- | | Cipher- | | Suite | | Suite | | | | | +-+-+-+-+-+ +-+-+-+-+-+ ^ ^ | | | | | | V V +-+-+-+-+-+ +-+-+-+-+-+ +-+-+-+-+-+ | |===============| |========| | | |EAP, TEK Deriv.| | | | | |<-------------------------------->| backend | | | | | | | | | TSK Deriv. | | MSK | | | peer |<------------->|Authenti-|<-------| auth | | |===============| cator |========| server | | | Link Layer | | AAA | (EAP | | | (PPP,IEEE 802)| |Protocol| server) | | | | | | | |MSK,EMSK | | MSK | |MSK,EMSK | | (TSKs) | | (TSKs) | | | | | | | | | +-+-+-+-+-+ +-+-+-+-+-+ +-+-+-+-+-+ ^ ^ | | | MSK, EMSK, IV | MSK, EMSK, IV | | | | +-+-+-+-+-+ +-+-+-+-+-+ | | | | | EAP | | EAP | | Method | | Method | | | | | |(MK,TEKs)| |(MK,TEKs)| | | | | +-+-+-+-+-+ +-+-+-+-+-+ Figure 3 - Pass-through relationship between EAP peer, authenticator and backend authentication server. Aboba & Simon Informational [Page 13] INTERNET-DRAFT EAP Keying Framework 2 March 2003 Figure 3 illustrates the EAP authentication process in the case where the authenticator acts as a pass-through. Here EAP is spoken between the peer and authenticator as before. The authenticator then encapsulates EAP packets within a AAA protocol such as RADIUS [RFC2869bis] or Diameter [DiamEAP], and forwards packets to and from the backend authentication server, which acts as the EAP server. Since the authenticator acts as a pass-through, EAP methods (as well as the derived EAP Master Key, and TEKs) reside only on the peer and backend authentication server. Once mutual authentication completes and is successful, the EAP method present on the peer and authenticator export the MSK, EMSK and IV. The backend authentication server then sends a message to the authenticator indicating that authentication has been successful, providing the MSK within a protected package known as the AN-Token. Along with the MSK, the AN-Token contains attributes indicating the parameters of key usage. The MSK is then used by the authenticator and peer to derive Transient Session Keys (TSKs) required for the negotiated ciphersuite. The TSKs are known only to the peer and authenticator, and as noted earlier, the TSK derivation process varies by ciphersuite. For example, within the 4-way handshake described in [IEEE80211i], the peer and authenticator confirm mutual possession of the MSK, demonstrate liveness, and do a protected ciphersuite and capabilities negotiation. 4. Security properties This section describes the security requirements for EAP methods, AAA protocols, TSK derivation mechanisms and Ciphersuites. These requirements MUST be met by specifications requesting publication as an RFC. Based on these requirements, the security properties of EAP exchanges are analyzed. 4.1. EAP method requirements Mutual authentication Methods deriving keys MUST support mutual authentication. Master Key Methods deriving keys MUST support derivation of the EAP Master Key (MK), as well as specifying how Transient EAP Keys (TEKs) are derived from the MK. The MK is the root of the EAP key hierarchy. As a result, compromise of MK must be avoided if at all possible, since an attacker in possession of the MK will be able to derive all the other keys in the hierarchy. This would provide an attacker with the ability not only to decrypt and insert data sent between a Aboba & Simon Informational [Page 14] INTERNET-DRAFT EAP Keying Framework 2 March 2003 particular EAP peer and authenticator, but also potentially to decrypt future data as well and to subsequently access the network using authentication mechanisms such as fast resume or fast handoff. Since the MK is known only to the EAP peer and server, and only mutually authenticating EAP methods may distribute keys, possession of the MK is proof of a completed mutual authentication. In order to protect against compromise of the MK, which could be used to impersonate the EAP peer or server the MK and TEKs MUST remain local to the EAP method and MUST NOT be provided to third parties. In addition, the MK MUST NOT be derivable from material exported from the EAP method, such as the MSK, EMSK or IV. The MK MUST NOT be directly used to protect data; rather the TEKs and TSKs are used for this purpose. MSK, EMSK and IV EAP methods supporting key derivation MUST export two 64 octet quantities, known as the Master Session Key (MSK), and the Extended Master Session Key (EMSK) and MAY export a 64 octet quantity known as the IV. It must be demonstrated that possession of the MSK, EMSK or IV does not provide information useful in determining the MK. Cryptographic separation Methods supporting key derivation MUST demonstrate cryptographic separation between the TEK, MSK, EMSK and IV branches of the EAP key hierarchy. Without violating a fundamental cryptographic assumption (such as the non-invertibility of a one-way function) an attacker recovering the TEK, MSK, EMSK or IV MUST NOT be able to recover the other quantities with a level of effort less than brute force. Since Transient Session Keys (TSKs) are derived from the MSK, if branch independence holds, then it is also true that the TSKs are cryptographically separate from the EMSK, IV and TEKs. EMSK reservation While the EMSK is exported by the EAP method, its use is reserved, and as a result it MUST remain known only to the EAP peer and server and MUST NOT be provided to third parties. Since the EMSK is the only keying material exported by an EAP method that is neither provided to a third party nor a known quantity, it is attractive for use in future applications such as fast handoff or man-in-the- middle detection. Given its potential future uses, damage due to EMSK compromise is second only in effect to compromise of the MK, yielding an attacker the ability to access the network at will, and to decrypt past and future data traffic. Ciphersuite Independence The MK, MSK, EMSK and IV derivations MUST be independent of the Aboba & Simon Informational [Page 15] INTERNET-DRAFT EAP Keying Framework 2 March 2003 selected ciphersuite. Key Strength The strength of Transient Session Keys (TSKs) and Transient EAP Keys (TEKs) used to protect data is ultimately dependent on the strength of the MK, MSK and EMSK generated by the EAP method. If EAP method does not produce an MK, MSK and EMSK of sufficient strength, then the TSKs and TEKs may be subject to brute force attack. EAP methods supporting key derivation MUST be capable of generating a MK, MSK and EMSK, each with an effective key strength of at least 128 bits. More details on key strength are provided in Section 5.3. Perfect Forward Secrecy An EAP peer and server may simultaneously derive MSKs suitable for use with several authenticators, so as to enable fast handoff between them. Similarly an EAP server may transport MSKs to multiple authenticators as the result of a single authentication. Where no backend authentication server is present, transport typically occurs via an Inter-Access Point Protocol (IAPP), such as [IEEE80211f]. Where a backend authentication server is present, key transport is provided by the AAA protocol, such as Diameter [DiamEAP] or RADIUS [RFC2869bis]. Key wrap mechanisms for Diameter are specified in [DIAMCMS], and for RADIUS in [RFC2548]. In order to protect against compromise of an individual MSK, Perfect Forward Secrecy (PFS) SHOULD be supported, so that compromise of one MSK does not enable compromise of subsequent or prior MSKs. Uniqueness In order to assure non-repetition of TSKs even in cases where one party may not have a high quality random number generator, the MSK derivation SHOULD include a two-way nonce exchange, using nonces of at least 128-bits. Note although the [IEEE80211i] 4-way handshake includes a nonce exchange, this is not the case for all ciphersuites and media, so that to provide media independence, an EAP method cannot assume that a nonce exchange is guaranteed to occur as part of TSK derivation. A nonce exchange SHOULD also be included in the derivation of the TEKs from the MK. Known-good algorithms The development and validation of key derivation algorithms is difficult, and as a result EAP methods SHOULD reuse existing key derivation algorithms, rather than inventing new ones. EAP methods requesting publication as an RFC MUST provide citations to literature justifying the security of the chosen algorithms. EAP methods SHOULD utilize well established and analyzed mechanisms for Aboba & Simon Informational [Page 16] INTERNET-DRAFT EAP Keying Framework 2 March 2003 MK, MSK, EMSK and IV derivation. 4.2. AAA protocol requirements AAA protocols suitable for use in transporting EAP MUST provide the following facilities: Security services AAA protocols used for transport of EAP MUST support per-packet integrity and authentication and SHOULD support replay protection and confidentiality. These requirements are met by Diameter EAP [DiamEAP], as well as RADIUS over IPsec [RFC2869bis]. Session Keys AAA protocols used for transport of EAP SHOULD provide per-packet security services using session keys, as in Diameter EAP [DiamEAP] and RADIUS over IPsec [RFC3162], rather than using a static key, as in RADIUS [RFC2865]. Mutual authentication AAA protocols used for transport of EAP MUST support mutual authentication between the authenticator and backend authentication server. These requirements are met by Diameter EAP [DiamEAP] as well as by RADIUS [RFC2865]. Forgery protection AAA protocols used for transport of EAP SHOULD provide protection against rogue authenticators masquerading as other authenticators. This can be accomplished, for example, by requiring that AAA agents to check the source address of packets against the origin attributes (Origin-Host AVP in Diameter, NAS-IP-Address, NAS- IPv6-Address, NAS-Identifier in RADIUS). MSKs vs. TSKs Since EAP methods do not export Transient Session Keys (TSKs) in order to maintain media and ciphersuite independence, the AAA protocol MUST NOT transport TSKs from the backend authentication server to authenticator. Key transport specification In order to enable backend authentication servers to provide keying material to the authenticator in a well defined format, AAA protocols suitable for use with EAP MUST define the format and wrapping of the package within which the MSK is transported, known as the AN-Token. The definition of the AN-Token MUST include the definition of attributes binding the key to the appropriate session, and providing limitations on key usage, such as the indicated key lifetime. Aboba & Simon Informational [Page 17] INTERNET-DRAFT EAP Keying Framework 2 March 2003 EMSK exposure Since the EMSK is a secret known only to the backend authentication server and peer, the AAA protocol MUST NOT transport the EMSK from the backend authentication server to the authenticator. AN-Token protection To ensure against compromise, the AN-Token MUST be integrity protected, authenticated, replay protected and encrypted in transit, using well-established cryptographic algorithms. For example, the AN-Token SHOULD be protected with session keys as in Diameter CMS Security [DiamCMS] (a work in progress) or RADIUS over IPsec [RFC2869bis] rather than static keys, as in [RFC2548]. Where untrusted intermediaries are present, the AN-Token SHOULD be protected by data object security mechanisms, such as Diameter CMS Security [DiamCMS] (a work in progress). 4.3. TSK derivation requirements The Transient Session Key (TSK) derivation process is assumed to provide for the following: Direct operation The TSK derivation process MUST operate directly between the peer and authenticator, and MUST NOT be passed-through to the backend authentication server. Mutual authentication Where EAP is used on link layers which cannot be assumed to be physically secure (e.g. wireless, the Internet), the TSK derivation process MUST provide for mutual authentication between the authenticator and peer. Protected negotiation Where EAP is used on link layers which cannot be assumed to be physically secure (e.g. wireless, the Internet), the TSK derivation process SHOULD support protected ciphersuite and capabilities negotiation. Uniqueness Where MSKs may be cached on the authenticator and peer, the TSK derivation process MUST provide unique TSKs for each session, even where the MSK is unchanged. Aboba & Simon Informational [Page 18] INTERNET-DRAFT EAP Keying Framework 2 March 2003 4.4. Ciphersuite requirements Ciphersuites suitable for keying by EAP methods MUST provide the following facilities: TSK derivation In order to key a ciphersuite with EAP, it is necessary to specify how the TSKs required by the ciphersuite are derived from the MSK. Derivation of the TSKs from the MSK requires knowledge of the negotiated ciphersuite. TEK derivation In order to establish a protected channel between the EAP peer and server as part of the EAP exchange, a ciphersuite needs to be negotiated and keyed, using TEKs derived from the MK. The ciphersuite used to protect the EAP exchange between the peer and server is distinct from the ciphersuite negotiated between the peer and authenticator, used to protect data. Where a protected channel is established within the EAP method, the method specification MUST specify the mechanism by which the EAP ciphersuite is negotiated, as well as the algorithms for derivation of TEKs from the MK during the EAP authentication exchange. EAP method independence Algorithms for deriving TSKs from the MSK MUST NOT depend on the EAP method. However, algorithms for deriving TEKs from the MK MAY be specific to the EAP method. Cryptographic separation The TSKs derived from the MSK MUST be cryptographically separate from each other. Similarly, TEKs MUST be cryptographically separate from each other. In addition, the TSKs MUST be cryptographically separate from the TEKs. 4.5. Security properties Given the requirements described in the previous sections, Figure 4 illustrates the relationship between the peer, authenticator and backend authentication server. As noted in the figure, each party in the exchange mutually authenticates with each of the other parties, and derives a unique key. All parties in the diagram have access to the MSK. Aboba & Simon Informational [Page 19] INTERNET-DRAFT EAP Keying Framework 2 March 2003 EAP peer /\ / \ Protocol: EAP / \ Protocol: TSK derivation Auth: Mutual / \ Auth: Mutual Unique keys: MK, / \ Unique keys: TSKs TEKs,EMSK / \ / \ Auth. server +--------------+ Authenticator Protocol: AAA Auth: Mutual Unique key: AAA session key Figure 4: Three-party EAP key distribution The EAP peer and backend authentication server mutually authenticate via the EAP method, and derive the MK, TEKs and EMSK which are known only to them. The TEKs are used to protect some or all of the EAP conversation between the peer and authenticator, so as to guard against modification or insertion of EAP packets by an attacker. The degree of protection afforded by the TEKs is determined by the EAP method; some methods may protect the entire EAP packet, including the EAP header, while other methods may only protect the contents of the Type-Data field, defined in [RFC2284bis]. Since EAP is spoken only between the peer and server, if a backend authentication server is present then the EAP conversation does not provide mutual authentication between the peer and authenticator, only between the peer and backend authentication server. As a result, mutual authentication between the peer and authenticator only occurs where a separate TSK derivation step is carried out, such as in [IEEE80211i]. This means that absent the TSK derivation step, from the point of view of the peer, EAP mutual authentication only proves that the authenticator is trusted by the backend authentication server; the identity of the authenticator is not confirmed. Utilizing the AAA protocol, the authenticator and backend authentication server mutually authenticate and derive session keys known only to them, used to provide per-packet integrity and replay protection, authentication and confidentiality. The MSK is distributed by the backend authentication server to the authenticator over this channel, bound to attributes constraining its usage, as part of the AN-Token. The binding of attributes to the MSK within a protected package is important so the authenticator receiving the AN-Token can determine that it has not been compromised, and that the keying material has not been replayed, or mis-directed in some way. Aboba & Simon Informational [Page 20] INTERNET-DRAFT EAP Keying Framework 2 March 2003 Assuming that the AAA protocol provides protection against rogue authenticators forging their identity, then the AN-Token can be assumed to be sent to the correct authenticator, and where it is wrapped appropriately, it can be assumed to be immune to compromise by a snooping attacker. Where an untrusted AAA intermediary is present, data object security SHOULD be used to encrypt, authenticate, integrity and replay protect the AN-Token, so that it cannot be compromised or modified by the intermediary. The TSK derivation step varies by ciphersuite. On link layers that cannot be assumed to be physically secure, the peer and authenticator SHOULD mutually authenticate by proving mutual possession of all or a portion of the MSK. It is also advisable for the TSK derivation step to support protected ciphersuite and capabilities negotiation, and derive TSKs which are guaranteed to be unique for each session. This provides assurance to the peer that it is connecting to the correct authenticator, that the capabilities and offered ciphersuites have not been forged, and that the TSKs are fresh. 5. Security considerations 5.1. Assumptions The security properties of the EAP exchange are dependent on each leg of the triangle: the selected EAP method, AAA protocol and TSK derivation mechanism. If the selected EAP method does not support mutual authentication, then the peer will be vulnerable to attack by rogue authenticators and backend authentication servers. If the EAP method does not derive keys, then TSKs will not be available for use with a negotiated ciphersuite, and there will be no binding between the initial EAP authentication and subsequent data traffic, leaving the session vulnerable to hijack. If the authenticator and backend authentication server do not mutually authenticate, then the peer will be vulnerable to rogue backend authentication servers, authenticators, or both. If there is no per- packet authentication, integrity and replay protection between the authenticator and backend authentication server, then an attacker can spoof or modify packets in transit. If the backend authentication server does not protect against authenticator masquerade, or provide the proper binding of the MSK to the session within the AN-Token, then one or more MSKs may be sent to an unauthorized party, and an attacker may be able to gain access to the network. If the AN-Token is not opaque to an untrusted AAA intermediary, then that intermediary may be able to modify the MSK, or the attributes associated with it, as described in Aboba & Simon Informational [Page 21] INTERNET-DRAFT EAP Keying Framework 2 March 2003 [RFC2607]. If the TSK derivation algorithm does not support mutual authentication, then the peer will not have assurance that it is connected to the correct authenticator, only that the authenticator and backend authentication server share a trust relationship (assuming that the AAA protocol supports mutual authentication). This distinction can become important when multiple authenticators receive MSKs from the backend authentication server, such as where fast handoff is supported. If the TSK derivation does not provide for protected ciphersuite and capabilities negotiation, then downgrade attacks are possible. 5.2. Key binding Both the RADIUS and Diameter protocols are potentially vulnerable to impersonation by a rogue authenticator. When RADIUS requests are forwarded by a proxy, the NAS-IP-Address or NAS-IPv6-Address attributes may not correspond to the source address. Since the NAS-Identifier attribute need not contain an FQDN, it also may not correspond to the source address, even indirectly. [RFC2865] Section 3 states: A RADIUS server MUST use the source IP address of the RADIUS UDP packet to decide which shared secret to use, so that RADIUS requests can be proxied. This implies that it is possible for a rogue authenticator to forge NAS- IP-Address, NAS-IPv6-Address or NAS-Identifier attributes within a RADIUS Access-Request in order to impersonate another authenticator. Among other things, this can result in messages (and MSKs) being sent to the wrong authenticator. Since the rogue authenticator is authenticated by the RADIUS proxy or server purely based on the source address, other mechanisms are required to detect the forgery. In addition, it is possible for attributes such as the Called-Station-Id and Calling- Station-Id to be forged as well. As recommended in [RFC2869bis], this vulnerability can be mitigated by having RADIUS proxies check authenticator identification attributes against the source address. To allow verification of session parameters such as the Called-Station- Id and Calling-Station-Id, they can be sent by the EAP peer to the server, and protected by TEKs. The RADIUS server can then check the parameters sent by the EAP peer against those claimed by the authenticator. If a discrepancy is found, an error can be logged. While [DiamBASE] requires use of the Route-Record AVP, this utilizes Aboba & Simon Informational [Page 22] INTERNET-DRAFT EAP Keying Framework 2 March 2003 FQDNs, so that impersonation detection requires DNS A/AAAA and PTR RRs to be properly configured. As a result, it appears that Diameter is as vulnerable to this attack as RADIUS, if not more so. To address this vulnerability, it is necessary to utilize data object security to protect the AN-Token, and allow the backend authentication server to authenticate the authenticator directly. This requires the authenticator to provide proof of its identity, ensuring that the MSK is being provided to the correct entity. 5.3. Key strength In order to guard against brute force attacks, EAP methods deriving keys need to be capable of generating an MK, MSK and EMSK with an appropriate effective symmetric key strength. In order to ensure that key generation is not the weakest link, it is necessary for EAP methods utilizing public key cryptography to choose a public key that has a cryptographic strength meeting the symmetric key strength requirement. As noted in Section 5 of [KeyLen], this results in the following required RSA or DH module and DSA subgroup size in bits, for a given level of attack resistance in bits: Attack Resistance RSA or DH Modulus DSA subgroup (bits) size (bits) size (bits) ----------------- ----------------- ------------ 70 947 128 80 1228 145 90 1553 153 100 1926 184 150 4575 279 200 8719 373 250 14596 475 5.4. Key wrap As described in [RFC2869bis], Section 4.3, known problems exist in the key wrap specified in [RFC2548]. Where the same RADIUS shared secret is used by a PAP authenticator and an EAP authenticator, there is a vulnerability to known plaintext attack. Since RADIUS uses the shared secret for multiple purposes, including per-packet authentication, attribute hiding, considerable information is exposed about the shared secret with each packet. This exposes the shared secret to dictionary attacks. MD5 is used both to compute the RADIUS Response Authenticator and the Message-Authenticator attribute, and some concerns exist relating to the security of this hash [MD5Attack]. As discussed in [RFC2869bis], Section 4.2, these and other RADIUS vulnerabilities may be addressed by running RADIUS over IPsec. Aboba & Simon Informational [Page 23] INTERNET-DRAFT EAP Keying Framework 2 March 2003 Where an untrusted AAA intermediary is present (such as a RADIUS proxy or a Diameter agent), and data object security is not used, the MSK may be recovered by an attacker in control of the untrusted intermediary. Possession of the MSK enables decryption of data traffic sent between the peer and a specific authenticator; however where Perfect Forward Secrecy (PFS) is implemented, compromise of the MSK does enable an attacker to impersonate the peer to another authenticator, since that requires possession of the MK or EMSK, which are not transported by the AAA protocol. This vulnerability may be mitigated by implementation of data object security techniques such as [DiamCMS], a work in progress. 5.5. Man-in-the-middle attacks As described in [MiTM], EAP method sequences and compound authentication mechanisms may be subject to man-in-the-middle attacks. When such attacks are successfully carried out, the attacker acts as an intermediary between a victim and a legitimate authenticator. This allows the attacker to authenticate successfully to the authenticator, as well as to obtain access to the network. In order to prevent these attacks, [MiTM] recommends derivation of a compound key by which the EAP peer and authenticator can prove that they have participated in the entire EAP exchange. Since the compound key must not be known to an attacker posing as an authenticator, and yet must be derived from quantities that are exported by EAP methods, it may be desirable to derive the compound key from a portion of the EMSK. In order to provide proper key hygiene, it is recommended that the compound key used for man-in-the-middle protection be cryptographically separate from other keys derived from the EMSK, such as fast handoff keys, discussed in Appendix F. 6. Normative References [RFC1661] Simpson, W., Editor, "The Point-to-Point Protocol (PPP)", STD 51, RFC 1661, July 1994. [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. [RFC2434] Alvestrand, H. and T. Narten, "Guidelines for Writing an IANA Considerations Section in RFCs", BCP 26, RFC 2434, October 1998. [RFC2284bis] Blunk, L., Vollbrecht, J., Aboba, B., "Extensible Authentication Protocol (EAP)", Internet draft (work in progress), draft-ietf-pppext-rfc2284bis-08.txt, December 2002. Aboba & Simon Informational [Page 24] INTERNET-DRAFT EAP Keying Framework 2 March 2003 [IEEE802] IEEE Standards for Local and Metropolitan Area Networks: Overview and Architecture, ANSI/IEEE Std 802, 1990. [IEEE80211] Information technology - Telecommunications and information exchange between systems - Local and metropolitan area networks - Specific Requirements Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications, IEEE Std. 802.11-1997, 1997. [IEEE8021X] IEEE Standards for Local and Metropolitan Area Networks: Port based Network Access Control, IEEE Std 802.1X-2001, June 2002. 7. Informative References [RFC1321] Rivest, R., "The MD5 Message-Digest Algorithm", RFC 1321, April 1992. [RFC1968] Meyer, G., "The PPP Encryption Protocol (ECP)", RFC 1968, June 1996. [RFC2104] Krawczyk, et al, "HMAC: Keyed-Hashing for Message Authentication", RFC 2104, February 1997. [RFC2246] Dierks, T. and Allen, C. "The TLS Protocol Version 1.0", RFC 2246, November 1998. [RFC2409] Harkins, D., Carrel, D., "The Internet Key Exchange (IKE)", RFC 2409, November 1998. [RFC2419] Sklower, K., Meyer, G., "The PPP DES Encryption Protocol, Version 2 (DESE-bis)", RFC 2419, September 1998. [RFC2420] Hummert, K., "The PPP Triple-DES Encryption Protocol (3DESE)", RFC 2420, September 1998. [RFC2434] Alvestrand, H. and T. Narten, "Guidelines for Writing an IANA Considerations Section in RFCs", BCP 26, RFC 2434, October 1998. [RFC2548] Zorn, G., "Microsoft Vendor-Specific RADIUS Attributes", RFC 2548, March 1999. [RFC2607] Aboba, B., Vollbrecht, J., "Proxy Chaining and Policy Implementation in Roaming", RFC 2607, June 1999. Aboba & Simon Informational [Page 25] INTERNET-DRAFT EAP Keying Framework 2 March 2003 [RFC2716] Aboba, B., Simon, D.,"PPP EAP TLS Authentication Protocol", RFC 2716, October 1999. [RFC2865] Rigney, C., Willens, S., Rubens, A., Simpson, W., "Remote Authentication Dial In User Service (RADIUS)", RFC 2865, June 2000. [RFC3078] Pall, G. and Zorn, G. "Microsoft Point-to-Point Encryption (MPPE) RFC 3078, March 2001. [RFC3079] Zorn, G. "Deriving Keys for use with Microsoft Point-to- Point Encryption (MPPE)," RFC 3079, March 2001. [RFC3394] R. Housley, "Advance Encryption Standard (AES) Key Wrap Algorithm", RFC 3394, September 2002. [Congdon] Congdon, P., et al., "IEEE 802.1X RADIUS Usage Guidelines", Internet draft (work in progress), draft- congdon-radius-8021x-23.txt, February 2003. [FIPSDES] National Bureau of Standards, "Data Encryption Standard", FIPS PUB 46 (January 1977). [DESMODES] National Bureau of Standards, "DES Modes of Operation", FIPS PUB 81 (December 1980). [FIPS197] FIPS PUB 197, Advanced Encryption Standard (AES), 2001 November 26H. [SHA] National Institute of Standards and Technology (NIST), "Announcing the Secure Hash Standard," FIPS 180-1, U.S. Department of Commerce, 04/1995 [IEEE80211f] IEEE Draft 802.11F/D5, "Draft Recommended Practice for Multi-Vendor Access Point Interoperability via an Inter- Access Point Protocol Across Distribution Systems Supporting IEEE 802.11 Operation", January 2003. [IEEE80211i] IEEE Draft 802.11I/D3.1, "Draft Supplement to STANDARD FOR Telecommunications and Information Exchange between Systems - LAN/MAN Specific Requirements - Part 11: Wireless Medium Access Control (MAC) and physical layer (PHY) specifications: Specification for Enhanced Security", February 2003. [EAPAPI] Microsoft Developer Network, "Windows 2000 EAP API", August 2000, http://msdn.microsoft.com/library/ default.asp?url=/library/en-us/eap/eapport_0fj9.asp Aboba & Simon Informational [Page 26] INTERNET-DRAFT EAP Keying Framework 2 March 2003 [RFC2869bis] Aboba, B., Calhoun, P., "RADIUS Support For Extensible Authentication Protocol (EAP)", Internet draft (work in progress), draft-aboba-radius-rfc2869bis-09.txt, February 2003. [RoamCERT] Aboba, B., "Certificate-Based Roaming", Internet draft (work in progress), draft-ietf-roamops-cert-02.txt, April 1999. [DiamBASE] Calhoun, P., et al., "Diameter Base Protocol", Internet draft (work in progress), draft-ietf-aaa-diameter-17.txt, December 2002. [DiamCMS] Calhoun, P., Farrell, S., Bulley, W., "Diameter CMS Security Application", Internet draft (work in progress), draft-ietf-aaa-diameter-cms-sec-04.txt, March 2002. [DiamEAP] Hiller, T., Zorn, G., "Diameter Extensible Authentication Protocol (EAP) Application", Internet draft (work in progress), draft-ietf-aaa-eap-00.txt, June 2002. [Handoff] Arbaugh, B., "Experimental Handoff Extension to RADIUS", Internet draft (work in progress), draft-irtf-aaaarch- handoff-00.txt, February 2003. [IEEE-02-758] Mishra, A., Shin, M., Arbaugh, W., Lee, I., Jang, K., "Proactive Caching Strategies for IAPP Latency Improvement during 802.11 Handoff", IEEE 802.11 Working Group, IEEE-02-758r1-F, November 2002. [IEEE-03-084] Mishra, A., Shin, M., Arbaugh, W., Lee, I., Jang, K., "Proactive Key Distribution to support fast and secure roaming", IEEE 802.11 Working Group, IEEE-03-084r1-I, http://www.ieee802.org/11/Documents/DocumentHolder/3-084.zip, January 2003. [IEEE-03-155] Aboba, B., "Fast Handoff Issues", IEEE 802.11 Working Group, IEEE-03-155r0-I, http://www.ieee802.org/11/Documents/DocumentHolder/3-155.zip, March 2003. [KeyLen] Orman, H., Hoffman, P., "Determining Strengths For Public Keys Used For Exchanging Symmetric Keys", Internet draft (work in progress), draft-orman-public-key- lengths-05.txt, December 2001. [8021XHandoff] Pack, S., Choi, Y., "Pre-Authenticated Fast Handoff in a Public Wireless LAN Based on IEEE 802.1X Model", School Aboba & Simon Informational [Page 27] INTERNET-DRAFT EAP Keying Framework 2 March 2003 of Computer Science and Engineering, Seoul National University, Seoul, Korea, 2002. [MD5Attack] Dobbertin, H., "The Status of MD5 After a Recent Attack", CryptoBytes Vol.2 No.2, Summer 1996. [MiTM] Puthenkulam, J., et al, "The Compound Authentication Binding Problem", Internet draft (work in progress), draft-puthekulam-eap-binding-02.txt, March 2003. Aboba & Simon Informational [Page 28] INTERNET-DRAFT EAP Keying Framework 2 March 2003 Appendix A - Ciphersuite independence The Master Session Key (MSK), Extended Master Session Key (EMSK) and IV exported by EAP methods MUST be ciphersuite independent. This confers several advantages: Ciphersuite negotiation Enabling derivation of the TSK(s) in a separate step provides improved security. For example, the TSK derivation algorithm supported within [IEEE80211i] enables the EAP peer and authenticator to mutually authenticate and conduct a protected ciphersuite and capabilities negotiation. If the MSK is used directly as a TSK, then the EAP peer and authenticator may not mutually authenticate each other, and a protected ciphersuite negotiation, if it occurs at all, would typically need to be supported within EAP itself. Since the ciphersuite negotiation mechanisms are link-layer specific, this would introduce media and ciphersuite dependencies into EAP. Document Revision If an EAP method specifies how to derive transient session keys for each ciphersuite, the specification will need to be revised each time a new ciphersuite is developed. This also implies that a backend authentication server supporting an EAP method would not be usable with all EAP-capable authenticators, if the backend authentication server were not upgraded to support a new ciphersuite implemented on the authenticator. EAP method complexity Requiring EAP methods to include ciphersuite-specific code for TSK derivation increases the complexity of the EAP method. Knowledge asymmetry In practice, an EAP method may not have knowledge of the ciphersuite that has been negotiated between the peer and authenticator. In PPP, ciphersuite negotiation occurs via the Encryption Control Protocol (ECP), described in [RFC1968]. Since ECP negotiation occurs after authentication, unless an EAP method is utilized that supports ciphersuite negotiation, the peer, authenticator and backend authentication server may not be able to anticipate the negotiated ciphersuite and therefore this information cannot be provided to the EAP method. Since ciphersuite negotiation is assumed to occur out-of-band, there is no need for ciphersuite negotiation within EAP. Aboba & Simon Informational [Page 29] INTERNET-DRAFT EAP Keying Framework 2 March 2003 Appendix B - Ciphersuite keying requirements To date, PPP and IEEE 802.11 ciphersuites are suitable for keying by EAP. This Appendix describes the keying requirements of common PPP and 802.11 ciphersuites. PPP ciphersuites include DESEbis [RFC2419], 3DES [RFC2420], and MPPE [RFC3078]. The DES algorithm is described in [FIPSDES], and DES modes (such as CBC, used in [RFC2419] and DES-EDE3-CBC, used in [RFC2420]) are described in [DESMODES]. For PPP DESEbis, a single 56-bit encryption key is required, used in both directions. For PPP 3DES, a 168-bit encryption key is needed, used in both directions. As described in [RFC2419] for DESEbis and [RFC2420] for 3DES, the IV, which is different in each direction, is "deduced from an explicit 64-bit nonce, which is exchanged in the clear during the [ECP] negotiation phase." There is therefore no need for the IV to be provided by EAP. For MPPE, 40-bit, 56-bit or 128-bit encryption keys are required in each direction, as described in [RFC3078]. No initialization vector is required. While these PPP ciphersuites provide encryption, they do not provide per-packet authentication or integrity protection, so an authentication key is not required in either direction. Within [IEEE80211], Transient Session Keys (TSKs) are required both for unicast traffic as well as for multicast traffic, and therefore separate key hierarchies are required for unicast keys and multicast keys. IEEE 802.11 ciphersuites include WEP-40, described in [IEEE80211], which requires a 40-bit encryption key, the same in either direction; and WEP-128, which requires a 104-bit encryption key, the same in either direction. These ciphersuites also do not support per-packet authentication and integrity protection. In addition to these unicast keys, authentication and encryption keys are required to wrap the multicast encryption key. Recently, new ciphersuites have been proposed for use with IEEE 802.11 that provide per-packet authentication and integrity protection as well as encryption [IEEE80211i]. These include TKIP, which requires a single 128-bit encryption key and a 128-bit authentication key (used in both directions); AES CCMP, which requires a single 128-bit key (used in both directions) in order to authenticate and encrypt data; and WRAP, which requires a single 128-bit key (used in both directions). As with WEP, authentication and encryption keys are also required to wrap the multicast encryption (and possibly, authentication) keys. Aboba & Simon Informational [Page 30] INTERNET-DRAFT EAP Keying Framework 2 March 2003 Appendix C - Example TEK Hierarchy Figure C-1 illustrates the TEK key hierarchy for EAP-TLS [RFC2716], which is based on the TLS key hierarchy described in [RFC2246]. The TLS-negotiated ciphersuite is used to set up a protected channel for use in protecting the EAP conversation, keyed by the derived TEKs. The TEK derivation proceeds as follows: master_secret = TLS-PRF-48(pre_master_secret, "master secret", client.random || server.random) TEK = TLS-PRF-X(master_secret, "key expansion", server.random || client.random) Where: TLS-PRF-X = TLS pseudo-random function defined in [RFC2246], computed to X octets. master_secret = TLS term for the MK. | | | | | pre_master_secret | server| | | client Random| V | Random | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | | | | | | | | +---->| master_secret |<------+ | | (MK) | | | | | | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | | | | | | | | | V V V +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | | | | Key Block | | (TEKs) | | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | | | | | | client | server | client | server | client | server | MAC | MAC | write | write | IV | IV | | | | | | V V V V V V Figure C-1 - TLS [RFC2246] Key Hierarchy Aboba & Simon Informational [Page 31] INTERNET-DRAFT EAP Keying Framework 2 March 2003 Appendix D - Example MSK, EMSK and IV Hierarchy In EAP-TLS [RFC2716], the MSK is divided into two halves, corresponding to the "Peer to Authenticator Encryption Key" (Enc-RECV-Key, 32 octets, also known as the PMK) and "Authenticator to Peer Encryption Key" (Enc- SEND-Key, 32 octets). In [RFC2548], the Enc-RECV-Key (the PMK) is transported in the MS-MPPE-Recv-Key attribute, and the Enc-SEND-Key is transported in the MS-MPPE-Send-Key attribute. The EMSK is also divided into two halves, corresponding to the "Peer to Authenticator Authentication Key" (Auth-RECV-Key, 32 octets) and "Authenticator to Peer Authentication Key" (Auth-SEND-Key, 32 octets). The IV is a 64 octet quantity that is a known value; octets 0-31 are known as the "Peer to Authenticator IV" or RECV-IV, and Octets 32-63 are known as the "Authenticator to Peer IV", or SEND-IV. In EAP-TLS, the MSK, EMSK and IV are derived from the MK via a one-way function. This ensures that the MK cannot be derived from the MSK, EMSK or IV unless the one-way function (TLS PRF) is broken. Since the MSK is derived from the MK, if the MK is compromised then the MSK is also compromised. As described in [RFC2716], the formula for the derivation of the MSK, EMSK and IV from the MK is as follows: MSK = TLS-PRF-64(MK, "client EAP encryption", client.random || server.random) EMSK = second 64 octets of: TLS-PRF-128(MK, "client EAP encryption", client.random || server.random) IV = TLS-PRF-64("", "client EAP encryption", client.random || server.random) MSK(0,31) = Peer to Authenticator Encryption Key (Enc-RECV-Key) (MS-MPPE-Recv-Key in [RFC2548]) MSK(32,63) = Authenticator to Peer Encryption Key (Enc-SEND-Key) (MS-MPPE-Send-Key in [RFC2548]) EMSK(0,31) = Peer to Authenticator Authentication Key (Auth-RECV-Key) EMSK(32,63) = Authenticator to Peer Authentication Key (Auth-Send-Key) IV(0,31) = Peer to Authenticator Initialization Vector (RECV-IV) IV(32,63) = Authenticator to Peer Initialization vector (SEND-IV) Where: IV(W,Z) = Octets W through Z inclusive of the IV. MSK(W,Z) = Octets W through Z inclusive of the MSK. EMSK(W,Z) = Octets W through Z inclusive of the EMSK. MK = TLS master_secret TLS-PRF-X = TLS PRF function defined in [RFC2246] computed to X octets client.random = Nonce generated by the TLS client. server.random = Nonce generated by the TLS server. Aboba & Simon Informational [Page 32] INTERNET-DRAFT EAP Keying Framework 2 March 2003 Figure D-1 describes the process by which the MSK,EMSK,IV and ultimately the TSKs, are derived from the MK. Note that in [RFC2716], the MK is referred to as the "TLS Master Secret". ---+ | ^ | TLS Master Secret (MK) | | | V | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | | EAP | | Master Session Key (MSK) | Method | | Derivation | | | | V +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ EAP ---+ | | | API ^ | MSK | EMSK | IV | | | | | V V V v +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ---+ | | | | | | | AAA server | | | | | | | V +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ---+ | | ^ | MSK(0,31) | MSK(32,63) | | (PMK) | Transported | | | via AAA | | | | V V V +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ---+ | | ^ | Ciphersuite-Specific Transient Session | Auth.| | Key Derivation | | | | V +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ---+ Figure D-1 - EAP TLS [RFC2716] MSK, EMSK and IV hierarchy Aboba & Simon Informational [Page 33] INTERNET-DRAFT EAP Keying Framework 2 March 2003 Appendix E - Example Transient Session Key (TSK) Derivation Within IEEE 802.11 RSN, the Pairwise Transient Key (PTK), a transient session key used to protect unicast traffic, is derived from the PMK (octets 0-31 of the MSK), known in [RFC2716] as the Peer to Authenticator Encryption Key. In [IEEE80211i], the PTK is derived from the PMK via the following formula: PTK = EAPOL-PRF-X(PMK, "Pairwise key expansion", Min(AA,SA) || Max(AA, SA) || Min(ANonce,SNonce) || Max(ANonce,SNonce)) Where: PMK = MSK(0,31) SA = Station MAC address AA = Access Point MAC address ANonce = Access Point Nonce SNonce = Station Nonce EAPOL-PRF-X = Pseudo-Random Function based on HMAC-SHA1, generating a PTK of size X octets. TKIP uses X = 64, while CCMP, WRAP, and WEP use X = 48. The EAPOL-Key Confirmation Key (KCK) is used to provide data origin authenticity in the TSK derivation. It utilizes the first 128 bits (bits 0-127) of the PTK. The EAPOL-Key Encryption Key (KEK) provides confidentiality in the TSK derivation. It utilizes bits 128-255 of the PTK. Bits 256-383 of the PTK are used by Temporal Key 1, and Bits 384-511 are used by Temporal Key 2. Usage of TK1 and TK2 is ciphersuite specific. Details are available in [IEEE80211i]. Aboba & Simon Informational [Page 34] INTERNET-DRAFT EAP Keying Framework 2 March 2003 Appendix F - Example PMK Derivation As discussed in [Handoff], [IEEE-02-758], [IEEE-03-084], and [8021XHandoff], keying material may be required for use in fast handoff between IEEE 802.11 authenticators. Where the backend authentication server provides keying material to multiple authenticators in order to fascilitate fast handoff, it is highly desirable for the keying material used on different authenticators to be cryptographically separate, so that if one authenticator is compromised, it does not lead to the compromise of other authenticators. Where keying material is provided by the backend authentication server, a key hierarchy derived from the EMSK, as suggested in [IEEE-03-155] can be used to provide cryptographically separate keying material for use in fast handoff: PMK0-A = MSK(0,31) PMK1-B = PRF(EMSK(0,31),PMK0-A,APB-MAC-Addr,STA-MAC-Addr) PMK1-E = PRF(EMSK(0,31),PMK0-A,APE-MAC-Addr,STA-MAC-Addr) Here PMK0-A is the Pairwise Master Key derived during the initial EAP authentication between the peer and authenticator A. Based on this initial EAP authentication, the EMSK is also derived, the first 32 octets of which can be used to derive PMKs for fast authentication between the EAP peer and authenticators B and E. Since the EMSK is cryptographically separate from the MSK, each of these PMKs is cryptographically separate from each other, and are guaranteed to be unique between the EAP peer (also known as the STA) and the authenticator (also known as the AP). Acknowledgments Thanks to Arun Ayyagari, Ashwin Palekar, and Tim Moore of Microsoft, Dorothy Stanley of Agere, Dave Halasz of Cisco Systems, and Russ Housley of RSA Security for useful feedback. Author Addresses Bernard Aboba Microsoft Corporation One Microsoft Way Redmond, WA 98052 EMail: bernarda@microsoft.com Phone: +1 425 706 6605 Fax: +1 425 936 7329 Dan Simon Microsoft Research Microsoft Corporation Aboba & Simon Informational [Page 35] INTERNET-DRAFT EAP Keying Framework 2 March 2003 One Microsoft Way Redmond, WA 98052 EMail: dansimon@microsoft.com Phone: +1 425 706 6711 Fax: +1 425 936 7329 Intellectual Property Statement The IETF takes no position regarding the validity or scope of any intellectual property or other rights that might be claimed to pertain to the implementation or use of the technology described in this document or the extent to which any license under such rights might or might not be available; neither does it represent that it has made any effort to identify any such rights. Information on the IETF's procedures with respect to rights in standards-track and standards- related documentation can be found in BCP-11. Copies of claims of rights made available for publication and any assurances of licenses to be made available, or the result of an attempt made to obtain a general license or permission for the use of such proprietary rights by implementors or users of this specification can be obtained from the IETF Secretariat. The IETF invites any interested party to bring to its attention any copyrights, patents or patent applications, or other proprietary rights which may cover technology that may be required to practice this standard. Please address the information to the IETF Executive Director. Full Copyright Statement Copyright (C) The Internet Society (2003). All Rights Reserved. This document and translations of it may be copied and furnished to others, and derivative works that comment on or otherwise explain it or assist in its implementation may be prepared, copied, published and distributed, in whole or in part, without restriction of any kind, provided that the above copyright notice and this paragraph are included on all such copies and derivative works. However, this document itself may not be modified in any way, such as by removing the copyright notice or references to the Internet Society or other Internet organizations, except as needed for the purpose of developing Internet standards in which case the procedures for copyrights defined in the Internet Standards process must be followed, or as required to translate it into languages other than English. The limited permissions granted above are perpetual and will not be revoked by the Internet Society or its successors or assigns. This document and the information contained herein is provided on an "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR Aboba & Simon Informational [Page 36] INTERNET-DRAFT EAP Keying Framework 2 March 2003 IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE." Open issues Open issues relating to this specification are tracked on the following web site: http://www.drizzle.com/~aboba/EAP/eapissues.html Expiration Date This memo is filed as , and expires August 22, 2003. Aboba & Simon Informational [Page 37]