| Internet-Draft | KEM-based Authentication for IKEv2 | November 2026 |
| Wang & Smyslov | Expires 19 May 2027 | [Page] |
This draft specifies a new authentication mechanism, called KEM (Key Encapsulation Mechanism) -based authentication, for the Internet Key Exchange Protocol Version 2 (IKEv2). This is motivated by the fact that some post-quantum KEMs (like ML-KEM) are more efficient than post-quantum signature algorithms (like ML-DSA).¶
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.¶
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.¶
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."¶
This Internet-Draft will expire on 19 May 2027.¶
Copyright (c) 2026 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 (https://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 Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶
Version 05: Section 8 was fully updated, with togther a number of editorial changes.¶
Version 04: Leave only one exchange sequence.¶
Version 03: Add identity protection against active attackers.¶
Version 02: Completely rewritten to accommodate ideas expressed in PQuAKE protocol.¶
Version 01: Two changes have been made, as a response to comments received at 122 meeting.¶
Cryptographically-relevant quantum computers (CRQC) pose a threat to cryptographically protected data. In particular, the so-called harvest-now-and-decrypt-later (HNDL) attack is considered an imminent threat. To mitigate this threat against the Internet Key Exchange Protocol Version 2 (IKEv2) [RFC7296], an extension allowing multiple key exchanges in IKEv2 [RFC9370] was introduced to achieve post-quantum (PQ) security for the key exchange.¶
"Signature Authentication in the Internet Key Exchange Version 2 (IKEv2) using PQC" [I-D.ietf-ipsecme-ikev2-pqc-auth] specifies how NIST PQ digital algorithms ML-DSA [FIPS204] and SLH-DSA [FIPS205] can be used in IKEv2. On the other hand, "Post-Quantum Traditional (PQ/T) Hybrid PKI Authentication in the Internet Key Exchange Version 2 (IKEv2)" [I-D.hu-ipsecme-pqt-hybrid-auth] specifies how to run general hybrid PQ/T digital algorithms in IKEv2.¶
Motivated by the fact that ML-KEM [FIPS203] is much more efficient than ML-DSA [FIPS204], "KEM-based Authentication for TLS 1.3" [I-D.celi-wiggers-tls-authkem] [I-D.wiggers-tls-authkem-psk] specifies how a KEM can be used to achieve post-quantum secure authentication for the TLS 1.3 protocol. The basic idea is as follows: when one entity A receives the certified long term public key of another entity B, A can authenticate B by encapsulating a secret key k using B's KEM public key, and confirming that the communicating entity is indeed B if the entity can successfully return the correct k to A, explicitly or implicitly. This idea was previously presented in [SSW20], followed by a number of research papers then. Besides saving communication overhead and computational time, as pointed out in [I-D.celi-wiggers-tls-authkem], KEM-based authentication also reduces implementation code size, as the code for PQ signature may not be needed, and KEM-based authentication can re-use the KEM code for ephemeral key establishment.¶
This draft specifies how to use KEM-based authentication for the Internet Key Exchange Protocol Version 2 (IKEv2) [RFC7296].¶
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here.¶
A key encapsulation mechanism (KEM) is a set of algorithms that enables key exchange to be performed, which allows one entity to encapsulate a secret key under a (longterm or ephemeral) public key of another entity. By following the definition given in [I-D.ietf-ipsecme-ikev2-mlkem], a KEM consists of three algorithms:¶
By following the definition given in [I-D.ietf-pquip-hybrid-signature-spectrums], a signature scheme consists of the following three algorithms:¶
In August of 2024, NIST released ML-KEM [FIPS203] and ML-DSA [FIPS204], which are both module-lattice-based post-quantum algorithms. Each of both algorithms has three sets of parameters corresponding to three NIST security levels. Namely, ML-KEM-512 for Level 1, ML-KEM-768 for Level 3, and ML-KEM-1024 for Level 5, while ML-DSA-44 for Level 2, ML-DSA-65 for Level 3, and ML-DSA-87 for Level 5.¶
This section compares ML-KEM and ML-DSA, with respect to the sizes of key, ciphertext and signature, and also the computational overhead of key generation, encapsulation, signing, decapsulation, and verifying.¶
To compare the communication overhead when using ML-DSA and ML-KEM, data from Table 2 in [FIPS204] and Table 3 in [FIPS204] are used to generate the following Table 1. The results show the comparison of keys and signature over ciphertext between ML-DSA and ML-KEM. For all corresponding security levels, the following statements can be concluded:¶
| Private Key/Decapsulation Key | Public Key/Encapsulation Key | Signature/Ciphertext | |
|---|---|---|---|
| ML-DSA-44/ML-KEM-512 | 2,650/1,632 =157% |
1,312/800 =164% |
2,420/768 =315% |
| ML-DSA-65/ML-KEM-768 | 4,032/2,400 =168% |
1,952/1,184 =165% |
3,309/1,088 =304% |
| ML-DSA-87/ML-KEM-1024 | 4,896/3,168 =155% |
2,592/1,568 =165% |
4,627/1,568 =295% |
For the three security levels, when ML-KEM is used to replace ML-DSA for authentication, the saved communication overhead, namely public key+signature for ML-DSA over encapsulation key+ciphertext for ML-KEM, is 2,164, 2,989, and 4,083 bytes, respectively. Those savings are helpful to improve the performance of IKEv2. In fact, as shown in the experiment in simulated environment, the average time delay for IKEv2 can increase a few times due to large size of PQ public key, ciphertext and signature, especially when the IP packet loss rate reaches 1% or more.¶
About the computation overhead comparison, [HAZ24] present implementation results of Dilithium and Kyber, via various optimization techniques for Keccak and the two PQ algorithms. Concretely, Table 6 in [HAZ24] shows the performance comparison of Dilithium and Kyber on ARMv7 Cortex-M4 processor, presented by clock cycles. By ignoring the small difference between ML-KEM and its informal version Kyber, also ML-DSA and its informal version Dilithium, the following Table 2 is obtained.¶
| SKeyGen Time/KeyGen Time | Sign Time/Encap Time | Verify Time/Decap Time | |
|---|---|---|---|
| ML-DSA-44/ML-KEM-512 | 1,357k/369k =368% |
3,487k/448k =778% |
1,350k/409k =330% |
| ML-DSA-65/ML-KEM-768 | 2,394k/604k =396% |
5,574k/732k =761% |
2,302k/674k =342% |
| ML-DSA-87/ML-KEM-1024 | 4,069k/962k =423% |
7,730k/1,119k =691% |
3,998k/1,043k =383% |
By assuming that the computational comparison shown in Table 2 reasonably presents a performance difference of ML-KEM and ML-DSA on the same platform with optimal implementations, for all three corresponding security levels, the following statements can be derived:¶
In the scenario of KEM-based authentication, both the private keys for ML-DSA and ML-KEM will be long term keys, so the private key generation time can be ignored, as it is just one time overhead. Therefore, by combining signature signing and verifying together, and also combining encapsulation and decapsulation together, we may simply say that the computational time of ML-DSA is about 5 times of that of ML-KEM.¶
With a KEM-based authentication in IKEv2 peers first perform an ephemeral key exchange that allows them to derive session keys for protecting subsequent exchanges. This ephemeral key exchange may take place entirely in the IKE_SA_INIT exchange or be accompanied by a series of subsequent IKE_INTERMEDIATE exchanges [RFC9370], in which case it would be a hybrid key exchange.¶
To run KEM-based Authentication in IKEv2, the two peers first exchange their long-term KEM certificates, and then exchange random values encapsulated with each other's long-term KEM public key, certified by the corresponding KEM certificate. Here, both long-term KEM certificates and the KEM ciphertexts encapsulating the random values are sent via the IKE_INTERMEDIATE exchanges. After decapsulating these random values the two peers can demonstrate their knowledge of the random values sent by the other side in the subsequent IKE_AUTH exchange, thus confirming that they do possess the corresponding KEM private keys linked to their long-term KEM certificates.¶
More details about the KEM-based authentication are provided in [I-D.uri-cfrg-pquake] and the corresonding paper.¶
The KEM-based authentication in IKEv2 relies on the IKE_INTERMEDIATE exchange [RFC9242] that MUST be supported. In addition, the following extensions can contribute to the security and the robustness of the protocol.¶
The initiator wishing to use the KEM-based authentication indicates this to the responder by including a new status type notification USE_AUTHKEM (with type <TBA2> by IANA) into the IKE_SA_INIT request message. This notification has the Protocol ID and SPI Size both set to 0 (meaning no SPI is present), and its notification data is empty. To be able to use the KEM-based authentication the initiator MUST also include the INTERMEDIATE_EXCHANGE_SUPPORTED notification.¶
If the responder receives the IKE_SA_INIT request containing the USE_AUTHKEM notification and it supports and is configured to use the KEM-based authentication, then it sends this notification back in the response. The responder can optionally include the Certificate Request payload (as defined in [RFC7296]) to help the initiator to select the right KEM certificate. The responder can also indicate which KEMs it is willing to use for authentication by including the SUPPORTED_AUTH_METHODS notification [RFC9593] in the response. In this case the notification SHOULD NOT contain announcements with the Auth Method other than a newly defined "KEM-based Authentication" (<TBA1> by IANA). The format of announcement is defined in Section 6.2.2.¶
Peers can also negotiate other IKEv2 extensions during IKE_SA_INIT. Some of them are listed in Section 6. In particular, the figures below show the optional use of PPK [RFC9867] during KEM-based authentication.¶
Initiator Responder
---------------------------------------------------------------------
HDR(IKE_SA_INIT), SAi1, KEi, Ni, --->
N(INTERMEDIATE_EXCHANGE_SUPPORTED),
[N(USE_PPK_INT),]
N(USE_AUTHKEM)
<--- HDR(IKE_SA_INIT), SAr1, KEr, Nr, [CERTREQ,]
N(INTERMEDIATE_EXCHANGE_SUPPORTED),
[N(SUPPORTED_AUTH_METHODS),]
[N(USE_PPK_INT),]
N(USE_AUTHKEM)
Peers MUST ignore the USE_AUTHKEM notification in the message if it is not accompanied by the INTERMEDIATE_EXCHANGE_SUPPORTED notification.¶
If using KEM-based authentication is mandatory for the responder and the initiator did not propose it or failed to make a valid proposal, then the responder MUST send the NO_PROPOSAL_CHOSEN notification indicating that the proposed parameters are unacceptable. Otherwise the responder can continue establishing IKE SA using other authentication method.¶
If using the KEM-based authentication is mandatory for the initiator and it failed to negotiate it with the responder, then the initiator MUST cancel establishing IKE SA. Otherwise the initiator can continue using other authentication method.¶
After the IKE_SA_INIT exchange completes, peers calculate the session keys as defined in Section 2.14 of [RFC7296] and use these keys to protect the next exchange.¶
Once the KEM-based authentication is negotiated in IKE_SA_INIT, a series of the IKE_INTERMEDIATE exchanges takes place. These exchanges can be classified as those for the purpose of the KEM-based authentication and those for other purposes. The exchanges for the KEM base authentication MUST follow all other IKE_INTERMEDIATE exchanges, so that they take place right before the IKE_AUTH exchange.¶
First, if multiple key exchanges were negotiated, then one or more IKE_INTERMEDIATE exchanges are performed, right after the IKE_SA_INIT exchange, as defined in [RFC9370].¶
Initiator Responder
---------------------------------------------------------------------
HDR(IKE_INTERMEDIATE), SK {KEi(n)} --->
<--- HDR(IKE_INTERMEDIATE), SK {KEr(n)}
Each of these exchanges updates the session keys as defined in Section 2.2.2 of [RFC9370].¶
If the use of PPKs as defined in [RFC9867] is negotiated, then the IKE_INTERMEDIATE exchange negotiating the PPK to use is performed as defined in Section 3.1 of [RFC9867].¶
Initiator Responder
---------------------------------------------------------------------
HDR(IKE_INTERMEDIATE), SK {PPK_IDENTITY_KEY)} --->
<--- HDR(IKE_INTERMEDIATE), SK {N(PPK_IDENTITY)}
After this exchange the session keys are updated as defined in Section 3.1.1 of [RFC9867].¶
If both additional key exchange(s) and the use of PPKs as defined in [RFC9867] are negotiated, then the last additional key exchange MAY alternatively be combined with negotiating the PPK in a single IKE_INTERMEDIATE exchange as shown in Figure 3. Note that in many cases only one additional key exchange will be negotiated (as a part of PQ/T hybrid key exchange), thus in this case the last additional key exchange will also be the only one.¶
Initiator Responder
---------------------------------------------------------------------
HDR(IKE_INTERMEDIATE), SK {KEi(m),
(PPK_IDENTITY_KEY)} --->
<--- HDR(IKE_INTERMEDIATE), SK {KEr(m),
N(PPK_IDENTITY)}
If this combined IKE_INTERMEDIATE exchange takes place, then upon its completion session keys are updated twice - fist as defined in Section 2.2.2 of [RFC9370] and then, provided the peers negotiate the PPK to use, as defined in Section 3.1.1 of [RFC9867].¶
Note, that while [RFC9867] generally requires that the PPK negotiation is performed in the last IKE_INTERMEDIATE exchange right before the IKE_AUTH exchange, it also allows this requirement to be overridden by other specification. This document specifically overrides this requirement by allowing the IKE_INTERMEDIATE exchanges concerned with the KEM-based authentication to take place between the negotiation of PPK and the IKE_AUTH exchange.¶
KEM-based authentication requires two exchanges, but the first of them is usually combined with the last IKE_INTERMEDIATE exchange performed for other purposes (see Figure 2, Figure 3, and Figure 4), so that only one additional IKE_INTERMEDIATE exchange takes place in this case. If no IKE_INTERMEDIATE exchanges preceed, then two such exchanges are needed for KEM-based authentication.¶
In the first exchange (which is combined with the preceeding IKE_INTERMEDIATE exchange, if any) the responder sends its KEM certificate to the initiator. The initiator can optionally include Certificate Request payloads in the request message to help the responder to select the right KEM certificate. The Certificate Request payload can also be used to indicate which format(s) of certificate are acceptable for the sender of this payload. This is done by setting the Cert Encoding field to appropriate value. For KEM-based authentication this field MUST be either "X.509 bundle" (new certificate encoding defined in this document, it value is <TBA5 by IANA>) or "Hash and URL of X.509 bundle" (value 13). If both encodings are acceptable, then two Certificate Request payloads MUST be included (with possibly different list of CAs) each with its own value in the Cert Encoding field. Since the Certificate Request payload is optional, its absence leaves the choice of the certificate and its encoding on the certificate sender's discretion. Note that these rules are also applicable to the Certificate Request payload sent by the responder in the IKE_SA_INIT exchange (see Section 6.1.1).¶
The initiator can also indicate which KEMs it is willing to use for authentication by including the SUPPORTED_AUTH_METHODS notification [RFC9593] in the request. In this case the notification SHOULD NOT contain announcements with the Auth Method other than the "KEM Authentication" (<TBA1> by IANA). The format of announcement is defined in Section 6.2.2.¶
The responder provides its KEM certificate in the response. Besides the KEM certificate, additional intermediate certificates and CRLs can be provided. All these certificates and CRLs are packed into the CertificateBundle ASN.1 structure defined in Section 3.6 of [RFC7296]. The KEM certificate MUST be the only KEM certificate in the bundle.¶
The certificate bundle can either be sent directly in the response or be provided by a reference using Hash and URL syntax. In both cases this content (either bundle or hash and URL) is sent in a new payload "Encrypted Certificate" (with type <TBA0 by IANA>), denoted as ECERT below. The content is encrypted using key derived from the SK_d for the next exchange (see Section 6.2.3 for details).¶
If the certificate bundle is sent inband, then the Cert Encoding field in the Encrypted Certificate payload is set to a new value "X.509 bundle", otherwise it is set to "Hash and URL of X.509 bundle". If certificate bundle is provided by reference, then a host can avoid downloading the peer's certificates if it cached them from a previous connection.¶
Exactly one Encrypted Certificate payload MUST be present in the response.¶
Since this exchange can be combined with the preceeding IKE_INTERMEDIATE exchange, we use notation "..." to indicate payloads for the purposes other than KEM-based authentication. In case of no preceeding IKE_INTERMEDIATE exchanges, no payloads for other purposes are present in this exchange.¶
Initiator Responder
---------------------------------------------------------------------
HDR(IKE_INTERMEDIATE), SK {... [CERTREQ+,]
[N(SUPPORTED_AUTH_METHODS)]} --->
<--- HDR(IKE_INTERMEDIATE), SK {... ECERT}
Whether this exchange updates the current session keys or not depends on the purpose of the preceeding exchange. In particular, in case of Figure 2, Figure 3, and Figure 4), the keys are updated as the result of those exchanges. If this exchange is not combined with other IKE_INTERMEDIATE exchanges, then it does not update the current session keys.¶
The initiator verifies the responder's certificate to make sure that it is valid, properly signed by a certificate chain leaded to a trusted CA, and that it is acceptable for responder authentication based on the initiator's local policy. The exact verification steps are out of scope of this document, but generally should follow those defined in [RFC5280].¶
Then the initiator starts the next IKE_INTERMEDIATE exchange. In the request message it sends the ciphertext CT_i, produced as a result of encapsulation of a random value SS_i using the KEM public key from the responder's certificate. CT_i is sent in a new status type notification AUTHKEM_CT (with type <TBA3> by IANA). The Protocol ID and SPI Size fields are set to 0 (meaning no SPI is present). The notification data contains the ciphertext CT_i.¶
In the same request the initiator also sends its KEM certificate. Besides the KEM certificate, additional intermediate signing certificates and CRLs can be sent. All these certificates and CRLs are packed into the CertificateBundle ASN.1 structure defined in Section 3.6 of [RFC7296]. The KEM certificate MUST be the only KEM certificate in the bundle. The bundle can either be present in the message or be provided by reference In both cases this content (either bundle or hash and URL) is sent in the "Encrypted Certificate" payload. The content is encrypted using key derived from SS_i. The details of the Certificate payload encryption are provided in Section 6.2.3.¶
The encryption provides protection of initiator's identity against active attackers, since only a responder with the private key corresponding to the responder's certificate is able to decrypt the initiator's KEM certificate.¶
If the certificate bundle is sent inband, then the Cert Encoding field in the Encrypted Certificate payload is set to a "X.509 bundle", otherwise it is set to "Hash and URL of X.509 bundle". If certificate bundle is provided by reference, then a host can avoid downloading the peer's certificates if it cached them from a previous connection.¶
Exactly one Encrypted Certificate payload MUST be present in the message.¶
The responder decrypts the initiator's certificate, verifies it to make sure that it is valid, properly signed by a certificate chain leaded to a trusted CA, and that it is acceptable for initiator authentication based on the responder's local policy. The exact verification steps are out of scope of this document, but generally should follow those defined in [RFC5280].¶
The initiator's certificate and the responder's one, sent in the previous exchange, MAY contain public keys for different KEMs, provided that peers support these KEMs, - there is no requirement that the KEMs must be the same.¶
The responder then sends a response message containing the ciphertext CT_r, produced as a result of encapsulation of a random value SS_r using the KEM public keys from the initiator's certificate. CT_r is sent in the AUTHKEM_CT notification in the same manner as SS_i.¶
The responder also includes a new status type notification AUTHKEM_SSCONF (with type <TBA3> by IANA). The Protocol ID and SPI Size fields are set to 0 (meaning no SPI is present). The notification data contains an SSCONF value computed as:¶
SSCONF = prf(SK_pr, SS_r)¶
where prf is the negotiated PRF for this SA and SK_pr is from the current session keys (see Section 3.3.1 of [RFC9242]).¶
Initiator Responder
---------------------------------------------------------------------
HDR(IKE_INTERMEDIATE), SK {N(AUTHKEM_CT), ECERT} --->
<--- HDR(IKE_INTERMEDIATE), SK {N(AUTHKEM_CT),
N(AUTHKEM_SSCONF)}
Upon receiving the response message, the initiator decapsulates the received ciphertext CT_r and obtains the SS_r key. Then the initiator computes re-computes the SSCONF value using the obtained SS_r and verifies that the value matches the content of the AUTHKEM_SSCONF notification data. This ensures that there are no mis-configurations and the received SS_r value is indeed the same that was encapsulated.¶
Upon completion of this exchange the session keys are updated. A new value of SKEYSEED is computed as:¶
SKEYSEED' = prf(SK_d, SS_i | SS_r)¶
Then all the session keys (SK_d, SK_ai, SK_ar, SK_ei, SK_er, SK_pi, and SK_pr) are computed as defined in Section 2.14 of [RFC7296]. These session keys are then used for protecting all subsequent exchanges starting from the IKE_AUTH exchange.¶
The IKE_AUTH exchange immediately follows the IKE_INTERMEDIATE exchange where peers sent each other encrypted SS values. The IKE_AUTH exchange is basically unchanged from that defined in [RFC7296]. The Auth Method field in the AUTH payloads MUST be "KEM Authentication". The format of the AUTH payload is computed as defined in Section 6.2.1.¶
Initiator Responder
---------------------------------------------------------------------
HDR(IKE_AUTH), SK{IDi, [IDr,]
AUTH, SAi2, TSi, TSr} --->
<--- HDR(IKE_AUTH), SK{IDr,
AUTH, SAr2, TSi, TSr}
The AUTH payloads are composed as defined in Section 6.2.1.¶
With KEM based authentication peers have to exchange their KEM certificates before actual authentication of each other. Certificates usually contain some identity information, that the CA asserts the included public key is bound to. This makes peers aware of each other's identity before the IKE_AUTH exchange starts.¶
However, IKEv2 doesn't rely on the identities from certificates when performing authentication. Instead, Identification payloads (IDi and IDr) are mandatory in the IKE_AUTH exchange and exactly they are included into the AUTH payload calculation. The content of the IDi and IDr payload may differ from the identification information in the corresponding certificate, as stated in Section 3.5 of [RFC7296].¶
The format of the Authentication payload is defined in Section 3.8 of [RFC7296]. In case of KEM-based authentication the Auth Method is set to "KEM Authentication" (<TBA1> by IANA). The Authentication Data is computed as:¶
For the initiator:
AUTH = prf( prf(SK_pi, "Key Pad for IKEv2"),
<InitiatorSignedOctets>)
For the responder:
AUTH = prf( prf(SK_pr, "Key Pad for IKEv2"),
<ResponderSignedOctets>)
¶
The keys SK_pi and SK_pr used for these calculations are from the latest generation of the SK* keys that is computed from SKEYSEED' (see Section 6.1.2.2).¶
The content of the InitiatorSignedOctets and the ResponderSignedOctets is defined in Section 3.3.2 of [RFC9242]. If full transcript authentication is employed as specified in [I-D.ietf-ipsecme-ikev2-downgrade-prevention], then the InitiatorSignedOctets and the ResponderSignedOctets are additionally modified as defined in [I-D.ietf-ipsecme-ikev2-downgrade-prevention].¶
The format of the announcement for the "KEM Authentication" is the same as for "Digital Signature" as shown in Section 3.2.3 of [RFC9593]. For convenience the format is also shown below.¶
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Length (>3) | Auth Method | Cert Link | |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +
| |
~ AlgorithmIdentifier ASN.1 object ~
| |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Length - Length of the announcement in octets, must be greater than 3.¶
Auth Method - Announced authentication method, in this case it is "KEM Authentication" <TBA1 by IANA>.¶
Cert Link - Links this announcement with a particular CA; see Section 3.2.2 of [RFC9593] for details.¶
AlgorithmIdentifier ASN.1 object - the AlgorithmIdentifier of PKIX (see Section 4.1.1.2 of [RFC5280]), encoded using distinguished encoding rules (DER) [X.690].¶
For "KEM Authentication" announcement the AlgorithmIdentifier MUST contain identifier of a KEM algorithm (and not, for example, of a signature algorithm). Identifiers for KEM algorithms are specified in the corresponding documents for these algorithms (e.g., for ML-KEM see [RFC9935]).¶
The Encrypted Certificate payload, denoted ECERT in this document, provides a means to transport certificates or other authentication-related information in encrypted form. The payload type for the Encrypted Certificate payload is <TBA0 by IANA>. This payload is semantically equivalent to the Certificate payload (defined in Section 3.6 of [RFC7296], but it provides a means to encrypt the certificate information. For this purpose the format of the Encrypted Certificate payload mirrors the format of the Encrypted payload (defined in Section 3.14 of [RFC7296]) with the difference, that it contains not the encrypted IKE payloads, but the encrypted Cert Encoding and Certificate Data as defined in Section 3.6 of [RFC7296].¶
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Next Payload |C| RESERVED | Payload Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
~ Initialization Vector ~
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
~ Encrypted Cert Encoding and Certificate Data ~
| |
| +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| | Padding (0-255 octets) |
+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+
| | Pad Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
~ Integrity Checksum Data ~
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Initialization Vector, Padding, Pad Length and Integrity Checksum Data are defined in Section 3.14 of [RFC7296].¶
This document restricts the possible values of Cert Encoding in the ECERT payload to either "X.509 bundle" (<TBA5 by IANA>) or "Hash and URL of X.509 bundle" (13). Future specifications may define additional encodings allowed for use in this payload.¶
The cryptographic processing of the Encrypted Certificate payload is identical to that described in Section 3.14 of [RFC7296] (as well as any document updating such processing for particular algorithms or modes, such as [RFC5282]).¶
The payload is encrypted and authenticated using the same encryption and integrity algorithms as were negotiated for IKE SA. The encryption key EC_e and the authentication key EC_a for protecting the Encrypted Certificate payload play the same role as SK_ei/SK_er and SK_ai/SK_ar keys for the Encrypted payload. They are calculated as follows:¶
EC_a | EC_e = prf+ ( K, Ni | Nr )¶
where Ni and Nr are the nonces from the IKE_SA_INIT exchange, stripped of any headers, and K is a key used to protect the certificate data in a particular step of the protocol. Depending on the place where the Encrypted Certificate is sent in the protocol, the value of K will be different. In particular:¶
For ECERT sent by the responder (as per Figure 5) the SK_d value for the next IKE_INTERMEDIATE is used as the K, which resulted to the following calculation:¶
EC_a | EC_e = prf+ ( SK_d(next), Ni | Nr )
¶
For ECERT sent by the initiator (as per Figure 6) the SS_i value is used as the K, which resulted to the following calculation:¶
EC_a | EC_e = prf+ ( SS_i, Ni | Nr )
¶
Note: Initial thoughts, need expanding.¶
KEM can be replaced as a composite KEM, as defined in Lamps WG. The composition job is done at crypto-engineering level, so using a composite KEM for KEM based authentication does not require to change any thing at the protocol layer of IKEv2.¶
Extension is straightforward using reliable transport.¶
The security considerations of IKEv2 and the extensions used with KEM based mechanism continue to apply. In particular, encryption of an exchange before completion of authentication does not, by itself, establish the identity of the communicating peer. Implementations MUST distinguish protection under provisional session keys from successful peer authentication. See [RFC7296] and [RFC9242].¶
The intended security properties are mutual authentication, confidentiality of established session keys, binding of authentication to the current exchange, and protection against replay and modification. Identity protection and forward secrecy require additional conditions described below.¶
These properties depend on secure KEMs, a cryptographically strong pseudorandom function, secure encryption and integrity algorithms, adequate randomness, correct credential validation, and correct implementation of the complete protocol state machine. Selection of a post-quantum KEM alone does not establish post-quantum security for the complete protocol.¶
KEM-based authentication establishes possession of a private decapsulation key through confirmation of secrets encapsulated to the corresponding public key. A valid certificate establishes a binding between a public key and an identity; receipt of that certificate does not establish that the current peer possesses its private key.¶
An implementation MUST NOT declare authentication successful solely because a certificate validates, a ciphertext decapsulates, or an encrypted payload passes its integrity check. Authentication requires successful verification of the peer's final AUTH payload using the keys and authenticated octets prescribed for the current exchange.¶
Certificate validation and local authorization policy MUST establish that the credential is acceptable for the claimed IKE identity and requested access. An implementation MUST NOT infer authorization merely from membership in a trusted certification hierarchy. In particular, differences between a certificate's identity information and an Identification payload require an explicit policy decision.¶
A group preshared key establishes, at most, membership in the group of entities possessing that key. It MUST NOT replace individual peer authentication where individual accountability or authorization is required.¶
Authentication MUST bind the peer's identity and role to the current protocol instance. Implementations MUST preserve the prescribed ordering of exchanges, key updates, and transcript processing. A message accepted in one role, session, or key generation MUST NOT be accepted as evidence of authentication in another.¶
The intermediate-exchange authentication rules are security-critical: they connect exchanges protected before authentication to the eventual AUTH verification. Implementations MUST apply those rules consistently, including when an exchange combines several extensions. See [RFC9242], Section 3.3.2.¶
Deployments requiring KEM authentication, post-quantum key establishment, or preshared-key protection MUST enforce those requirements through local policy. Failure to negotiate a required mechanism MUST terminate establishment rather than trigger an automatic retry with weaker requirements.¶
Full-transcript authentication [I-D.ietf-ipsecme-ikev2-downgrade-prevention] [RFC9370] SHOULD be used. Implementations MUST NOT treat an unauthenticated capability announcement, error, or timeout as authorization to weaken security policy.¶
Long-term authentication keys MUST be separated from ephemeral key-establishment keys. A long-term authentication private key MUST NOT also be exposed through a service that returns raw decapsulation results or otherwise enables an attacker to obtain the secret needed to authenticate in this protocol.¶
Certificates and key-management policy SHOULD restrict authentication KEM keys to their intended use. Where an applicable certificate profile provides usage restrictions, implementations MUST enforce them.¶
Raw KEM shared secrets MUST NOT be transmitted as authentication evidence or recorded in ordinary diagnostic logs. Key confirmation MUST use the protocol-defined cryptographic computations and session context.¶
Keys for certificate protection, final authentication, IKE protection, and application traffic require separation by purpose and, where applicable, direction. A protocol revision introducing additional derivations MUST specify unambiguous inputs and domain separation; implementations MUST NOT invent incompatible derivation rules locally.¶
Forward secrecy concerns compromise of long-term credentials after completion of an otherwise uncompromised session. It does not protect a session whose ephemeral secrets or traffic keys were exposed during execution.¶
Post-quantum forward secrecy requires a suitable ephemeral post-quantum contribution, or a suitable hybrid construction containing such a contribution, that remains secure against the adversary under consideration. Long-term KEM authentication secrets alone do not provide this property: an adversary retaining ciphertexts AUTHKEM_CTs may recover their shared secrets after obtaining the corresponding long-term KEM private keys.¶
The selected key schedule must preserve the protection provided by the ephemeral contribution. Implementations MUST erase obsolete ephemeral private keys, encapsulation randomness, intermediate shared secrets, and superseded key material when they are no longer required.¶
When multiple key exchanges are used, their negotiation, combination, and key updates MUST follow the applicable specification. Security of the resulting composition must not be inferred simply from the presence of several algorithms (See [RFC9370]).¶
A claim of post-quantum authentication also depends on the credential-validation path. A quantum-vulnerable certification signature can undermine authentication against an attacker capable of forging that signature, even when the certified public key belongs to a post-quantum KEM.¶
Identity protection must be evaluated separately for passive observers, active attackers, authorized peers, and holders of any group preshared key.¶
For the initiator, as shown in Figure 6, confidentiality of credentials encrypted using a secret encapsulated to the responder depends on validating the responder's credential and checking that it represents an acceptable intended peer before releasing the initiator's credentials. Accepting an arbitrary trusted credential can disclose identity information to an unintended but credentialed responder.¶
An unauthenticated active initiator can establish its own ephemeral exchange with a responder. Consequently, as shown in Figure 5, ephemeral encryption alone does not prevent that initiator from obtaining responder identity information. A group preshared key can restrict this exposure to parties possessing that key, provided its contribution protects the credentials before they are disclosed. The placement of preshared-key mixing therefore matters; RFC 9867 provides protection earlier in IKE establishment than the mechanism it extends.¶
Neither credential encryption nor a group preshared key hides identities from an authorized peer or a compromised group member. Group-key compromise expands the population able to probe protected identities and requires an appropriate revocation and replacement policy.¶
Certificate retrieval, revocation checking, message sizes, timing, addresses, and stable identifiers can reveal information independently of credential encryption. Privacy-sensitive deployments SHOULD prefer in-band credential delivery and SHOULD assess the observable behavior of external retrieval and validation services. Session-key forward secrecy does not automatically imply forward secrecy of identity information. Recorded encrypted credentials may become readable after later compromise of the secrets used specifically to protect them.¶
Implementations MUST validate payload lengths, algorithm identifiers, encodings, certificate structures, and algorithm-specific public-key and ciphertext requirements before using them. Validation MUST include bounds on allocation, certificate-chain processing, and nested structures.¶
For KEMs providing implicit rejection, implementations MUST preserve that behavior. They MUST NOT expose whether a correctly sized ciphertext produced an ordinary shared secret or an implicit-rejection secret through explicit notifications, logging interfaces accessible to the attacker, or avoidable timing differences.¶
Cryptographic comparisons MUST use appropriate constant-time operations. Private-key operations and subsequent failure processing SHOULD avoid secret-dependent observable behavior.¶
Malformed framing may require immediate rejection. Failures involving credential identity, authorization, or decapsulation require additional care because distinguishable responses can disclose identity or private-key information. Implementations SHOULD minimize such distinctions without reporting successful authentication or allowing unauthorized access.¶
Failure handling MUST release provisional state and prevent application use of keys from an unsuccessful exchange.¶
Certificate processing, KEM operations, external retrieval, and fragment reassembly create opportunities for denial of service before peer authentication completes.¶
Implementations SHOULD apply bounded concurrency, timeouts, rate limits, and limits on message size, certificate-chain depth, retrieval activity, and reassembly storage. Appropriate IKEv2 cookie mechanisms SHOULD be used where applicable.¶
Retransmission MUST preserve the original exchange semantics. In particular, an implementation MUST NOT accidentally repeat a key update, change an encapsulated secret within the same retransmitted exchange, or reuse an encryption nonce with different plaintext under the same key.¶
URLs received from an unauthenticated peer MUST NOT permit unrestricted access to local files, internal services, or unbounded remote resources.¶
IANA is requested to make the following changes in the IKEv2 registries "Internet Key Exchange Version 2 (IKEv2) Parameters" [IANA-IKEv2]:¶
Define a new payload type in the "IKEv2 Payload Types" registry:¶
| Value | Next Payload Type | Notation |
|---|---|---|
| TBA0 | Encrypted Certificate | ECERT |
Define a new authentication method in the "IKEv2 Authentication Method" registry:¶
| Value | IKEv2 Authentication Method |
|---|---|
| TBA1 | KEM Authentication |
Define three new status type notifications in the "IKEv2 Notify Message Status Types" registry:¶
| Value | Notify Message Status Type |
|---|---|
| TBA2 | USE_AUTHKEM |
| TBA3 | AUTHKEM_CT |
| TBA4 | AUTHKEM_SSCONF |
Define a new certificate encoding in the "IKEv2 Certificate Encodings" registry:¶
| Value | Certificate Encoding |
|---|---|
| TBA5 | X.509 bundle |
PQuAKE achieves authenticated key establishment with four conceptual stages:¶
This construction avoids requiring each endpoint to generate an online digital signature over the handshake. Certificate signatures may still be required to validate the long-term public keys.¶
The mechanism in this document adopts the same general approach within IKEv2. Verion-04 of this draft explicitly identifies alignment with PQuAKE [I-D.uri-cfrg-pquake] as a design objective. Its IKE-specific processing remains authoritative for interoperability (see Section 6 of the draft).¶
| PQuAKE [I-D.uri-cfrg-pquake] Component | Corresponding IKEv2 Mechanism Specified in This Draft |
|---|---|
Ephemeral shared secret
|
Initial and any additional key exchanges |
Optional identity-hiding preshared secret
|
PPK contribution before credential disclosure |
Protected certificate exchange
|
ECERT payloads |
Encapsulation to long-term peer keys
|
AUTHKEM_CT notifications |
Combination of authentication secrets
|
Post-encapsulation IKE key update |
Directional key confirmation
|
Final IKE_AUTH AUTH payloads |
Binding of exchanged messages
|
IKEv2 authenticated octets and intermediate-exchange transcript processing |
This correspondence identifies functional roles. It does not necessarily imply identical encodings, derivation functions, or security proofs.¶
PQuAKE defines a protocol-specific transcript and derives separate confirmation and session keys. An IKEv2 integration must instead be evaluated using the actual IKEv2 key hierarchy, authenticated octets, extension interactions, and acceptance conditions.¶
Revision -04 also introduces asymmetric credential protection: the initiator's ECERT protection depends on a secret encapsulated to the responder. Its AUTHKEM_SSCONF check precedes final authentication and MUST NOT be treated as a replacement for AUTH verification. See Section 6.1.2.2 for detail.¶
These differences affect what each participant knows at each stage. In particular, analysis must distinguish knowledge of a provisional encryption key, possession of a certified private key, agreement on an identity, and successful authorization.¶
Thom Wiggers shared an idea to provide identity protection of one of the peers against active attackers.¶
Uri Blumenthal and Brandon Luo contributed a lot to the design of this protocol as authors of PQuAKE protocol.¶