| Internet-Draft | Group OSCORE | December 2025 |
| Tiloca, et al. | Expires 26 June 2026 | [Page] |
This document defines the security protocol Group Object Security for Constrained RESTful Environments (Group OSCORE), providing end-to-end security of messages exchanged with the Constrained Application Protocol (CoAP) between members of a group, e.g., sent over IP multicast. In particular, the described protocol defines how OSCORE is used in a group communication setting to provide source authentication for CoAP group requests, sent by a client to multiple servers, and for protection of the corresponding CoAP responses. Group OSCORE also defines a pairwise mode where each member of the group can efficiently derive a symmetric pairwise key with each other member of the group for pairwise OSCORE communication. Group OSCORE can be used between endpoints communicating with CoAP or CoAP-mappable HTTP.¶
This note is to be removed before publishing as an RFC.¶
Status information for this document may be found at https://datatracker.ietf.org/doc/draft-ietf-core-oscore-groupcomm/.¶
Discussion of this document takes place on the Constrained RESTful Environments (core) Working Group mailing list (mailto:core@ietf.org), which is archived at https://mailarchive.ietf.org/arch/browse/core/. Subscribe at https://www.ietf.org/mailman/listinfo/core/.¶
Source for this draft and an issue tracker can be found at https://github.com/core-wg/oscore-groupcomm.¶
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 26 June 2026.¶
Copyright (c) 2025 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.¶
The Constrained Application Protocol (CoAP) [RFC7252] is a web transfer protocol specifically designed for constrained devices and networks [RFC7228]. Group communication for CoAP [I-D.ietf-core-groupcomm-bis] addresses use cases where deployed devices benefit from a group communication model, for example to reduce latencies, improve performance, and reduce bandwidth utilization. Use cases include lighting control, integrated building control, software and firmware updates, parameter and configuration updates, commissioning of constrained networks, and emergency multicast (see Appendix B). Group communication for CoAP [I-D.ietf-core-groupcomm-bis] mainly uses UDP/IP multicast as the underlying data transport.¶
Object Security for Constrained RESTful Environments (OSCORE) [RFC8613] describes a security protocol based on the exchange of protected CoAP messages. OSCORE builds on CBOR Object Signing and Encryption (COSE) [RFC9052][RFC9053] and provides end-to-end encryption, integrity, replay protection, and binding of response to request between a sender and a recipient, independent of the transport layer also in the presence of intermediaries. To this end, a CoAP message is protected by including its payload (if any), certain options, and header fields into a COSE object, which is conveyed within the CoAP payload and the CoAP OSCORE Option of the protected message, thereby replacing those message fields with an authenticated and encrypted object.¶
This document defines Group OSCORE, a security protocol for group communication with CoAP [I-D.ietf-core-groupcomm-bis] and for CoAP-mappable HTTP requests and responses, providing the same end-to-end security properties as OSCORE also in the case where requests have multiple recipients. In particular, the described protocol defines how OSCORE is used in a group communication setting to provide source authentication for group requests sent by a client to multiple servers, and for protection of the corresponding responses. Group OSCORE also defines a pairwise mode where each member of the group can efficiently derive a symmetric pairwise key with each other member of the group for pairwise-protected OSCORE communication. Just like OSCORE, Group OSCORE is independent of the transport layer and works wherever CoAP does.¶
As with OSCORE, it is possible to combine Group OSCORE with communication security on other layers. One example is the use of transport layer security, such as DTLS [RFC9147], between one client and one proxy, or between one proxy and one server. This prevents observers from accessing addressing information conveyed in CoAP options that would not be protected by Group OSCORE, but would be protected by DTLS. These options include Uri-Host, Uri-Port, and Proxy-Uri. Note that DTLS does not define how to secure messages sent over IP multicast and cannot be used for end-to-end protection over a proxy. Group OSCORE is also intended to work with OSCORE-capable proxies [I-D.ietf-core-oscore-capable-proxies] thereby enabling, for example, nested OSCORE operations with OSCORE-protected communication between a CoAP client and a proxy, carrying messages that are additionally protected with Group OSCORE between the CoAP client and the target CoAP servers.¶
Group OSCORE defines two modes of operation that can be used independently or together:¶
In the group mode, Group OSCORE requests and responses are digitally signed with the private key of the sender and the signature is embedded in the protected CoAP message. The group mode supports all COSE signature algorithms as well as signature verification by intermediaries. This mode is defined in Section 7.¶
In the pairwise mode, two group members exchange OSCORE requests and responses (typically) over unicast, and the messages are protected with symmetric keys not known by the other group members. These symmetric keys are derived from Diffie-Hellman shared secrets, calculated with the asymmetric keys of the sender and recipient, allowing for shorter integrity tags and therefore lower message overhead. This mode is defined in Section 8.¶
Both modes provide source authentication of CoAP messages. The application decides what mode to use, potentially on a per-message basis. Such decisions can be based, for instance, on pre-configured policies or dynamic assessing of the target recipient and/or resource, among other things. One important case is when requests are protected in group mode, and responses in pairwise mode. Since such responses convey shorter integrity tags instead of bigger, full-fledged signatures, this significantly reduces the message overhead in case of many responses to one request.¶
A special deployment of Group OSCORE consists in using the pairwise mode only. For example, consider the case of a constrained-node network [RFC7228] with a large number of CoAP endpoints and the objective to establish secure communication between any pair of endpoints with a small provisioning effort and message overhead. Since the total number of security associations that needs to be established grows with the square of the number of endpoints, it is desirable to restrict the amount of secret keying material provided to each endpoint. Moreover, a key establishment protocol would need to be executed for each security association. One solution to this issue is to deploy Group OSCORE, with the endpoints being part of a group, and to use the pairwise mode. This solution has the benefit of providing a single shared secret, while distributing only the public keys of group members or a subset of those. After that, a CoAP endpoint can locally derive the OSCORE Security Context for the other endpoint in the group, and protect CoAP communications with very low overhead [I-D.ietf-iotops-security-protocol-comparison].¶
In some circumstances, Group OSCORE messages may be transported in HTTP, e.g., when they are protected with the pairwise mode and target a single recipient, or when they are protected with the group mode and target multiple CoAP recipients through cross-protocol translators such as HTTP-to-CoAP proxies [RFC8075][I-D.ietf-core-groupcomm-proxy]. The use of Group OSCORE with HTTP is as defined for OSCORE in Section 11 of [RFC8613].¶
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.¶
Readers are expected to be familiar with the terms and concepts described in CoAP [RFC7252], including "endpoint", "client", "server", "sender", and "recipient"; group communication for CoAP [I-D.ietf-core-groupcomm-bis]; Observe [RFC7641]; Concise Binary Object Representation (CBOR) [RFC8949]; Concise Data Definition Language (CDDL) [RFC8610]; COSE [RFC9052][RFC9053] and related countersignatures [RFC9338].¶
Readers are also expected to be familiar with the terms and concepts for protection and processing of CoAP messages through OSCORE, such as "Security Context" and "Master Secret", defined in [RFC8613].¶
Terminology for constrained environments, such as "constrained device" and "constrained-node network", is defined in [RFC7228].¶
This document refers also to the following terminology.¶
Keying material: data that is necessary to establish and maintain secure communication among endpoints. This includes, for instance, keys and IVs [RFC4949].¶
Authentication credential: information associated with an entity, including that entity's public key and parameters associated with the public key. Examples of formats of authentication credentials are CBOR Web Tokens (CWTs) and CWT Claims Sets (CCSs) [RFC8392], X.509 certificates [RFC5280], and C509 certificates [I-D.ietf-cose-cbor-encoded-cert]. Further details about authentication credentials are provided in Section 2.4.¶
Group: a set of endpoints that share group keying material and security parameters (Common Context, see Section 2). That is, unless otherwise specified, the term group used in this document refers to a "security group" (see Section 2.1 of [I-D.ietf-core-groupcomm-bis]), not to be confused with "CoAP group" or "application group".¶
Group Manager: an entity responsible for a group, required neither to be an actual group member nor to take part in the group communication. The operations of the Group Manager are defined in Section 12 and its responsibilities are listed in Appendix D.¶
Silent server: a member of a group that performs only group mode processing on incoming requests and never sends responses protected with Group OSCORE. For CoAP group communications, requests are normally sent without necessarily expecting a response. A silent server may send unprotected responses, as error responses reporting a Group OSCORE error.¶
Group Identifier (Gid): identifier assigned to the group, unique within the set of groups of a given Group Manager. The Gid value changes every time the group is rekeyed (see Section 12.2).¶
Birth Gid: with respect to a group member, the Gid obtained by that group member upon (re-)joining the group.¶
Key Generation Number: an integer value identifying the current version of the keying material used in a group.¶
Source authentication: evidence that a received message in the group originated from a specific identified group member. This also provides assurance that the message was not tampered with by anyone, be it a different legitimate group member or an endpoint which is not a group member.¶
Group request: a CoAP request message sent by a client in the group to servers in that group.¶
Long exchange: an exchange of messages associated with a request that is a group request and/or an Observe request [RFC7641].¶
In either case, multiple responses can follow from the same server to the request associated with the long exchange, even if the request is not an Observe request (see Section 3.1.6 of [I-D.ietf-core-groupcomm-bis]). The client terminates a long exchange when freeing up the CoAP Token value used for the associated request, for which no further responses will be accepted afterwards.¶
This document refers to a group as a set of endpoints sharing keying material and security parameters for executing the Group OSCORE protocol, see Section 1.1. All members of a group maintain a Security Context as defined in this section.¶
How the Security Context is established by the group members is out of scope for this document, but if there is more than one Security Context applicable to a message, then the endpoints MUST be able to determine which Security Context was latest established. The management of information about the group (i.e., identifiers, OSCORE input parameters, and keying material) is described in terms of a Group Manager (see Section 12).¶
An endpoint of the group may use the group mode (see Section 7), the pairwise mode (see Section 8), or both, depending on the modes it supports and on the parameters of the Security Context. The Security Context of Group OSCORE extends the OSCORE Security Context defined in Section 3 of [RFC8613] as follows (see Figure 1).¶
One Common Context, shared by all the endpoints in the group and extended as defined below.¶
The new parameter Authentication Credential Format (see Section 2.1.5), specifying the format of authentication credentials used in the group (see Section 2.4).¶
The new parameter Group Manager Authentication Credential, specifying the authentication credential of the Group Manager responsible for the group (see Section 2.1.6).¶
For the group mode, the Common Context is extended with the following new parameters.¶
Group Encryption Algorithm, specifying the algorithm used for encrypting and decrypting messages protected in group mode (see Section 2.1.7).¶
Signature Algorithm, specifying the algorithm used for computing and verifying the countersignature of messages protected in group mode (see Section 2.1.8).¶
Signature Encryption Key, specifying the symmetric key used for deriving a keystream, which is in turn used for encrypting and decrypting the countersignature of messages protected in group mode (see Section 2.1.9).¶
For the pairwise mode, the Common Context is extended with a Pairwise Key Agreement Algorithm (see Section 2.1.10) used for the agreement on a static-static Diffie-Hellman shared secret, from which pairwise keys are derived (see Section 2.5.1).¶
The content of the Common Context is long-term, as it is meant to be stable once the Common Context is established.¶
One Sender Context, extended with the following new parameters.¶
The endpoint's own private key used to sign messages protected in group mode (see Section 7), or for deriving pairwise keys used with the pairwise mode (see Section 2.5).¶
The endpoint's own authentication credential containing its public key (see Section 2.4).¶
For the pairwise mode, the Sender Context is extended with the Pairwise Sender Keys associated with the other endpoints (see Section 2.5).¶
Except for the Sender Sequence Number defined in Section 3.1 of [RFC8613], the content of the Sender Context is long-term, as it is meant to be stable once the Sender Context is established.¶
If the endpoint is configured exclusively as a silent server (see Section 1.1), then the Sender Context is omitted.¶
One Recipient Context for each other endpoint from which messages are received. It is not necessary to maintain Recipient Contexts associated with endpoints from which messages are not (expected to be) received.¶
Each Recipient Context is extended with the authentication credential of the other endpoint, used to verify the signature of messages protected in group mode, or for deriving the pairwise keys used with the pairwise mode (see Section 2.5).¶
For the pairwise mode, each Recipient Context is extended with the Pairwise Recipient Key associated with the other endpoint (see Section 2.5).¶
Except for the Replay Window defined in Section 3.1 of [RFC8613], the content of each Recipient Context is long-term, as it is meant to be stable once the Recipient Context is established.¶
The varying part of the Group OSCORE Security Context is composed of the Sender Sequence Number in the Sender Context and the Replay Windows in the different Recipient Contexts.¶
+-------------------+-------------------------------------------------+ | Context Component | New Information Elements | +-------------------+-------------------------------------------------+ | Common Context | Authentication Credential Format | | | Group Manager Authentication Credential | | | * Group Encryption Algorithm | | | * Signature Algorithm | | | * Signature Encryption Key | | | ^ Pairwise Key Agreement Algorithm | +-------------------+-------------------------------------------------+ | Sender Context | Endpoint's own private key | | | Endpoint's own authentication credential | | | ^ Pairwise Sender Keys for the other endpoints | +-------------------+-------------------------------------------------+ | Each | Other endpoint's authentication credential | | Recipient Context | ^ Pairwise Recipient Key for the other endpoint | +-------------------+-------------------------------------------------+
The following sections specify how the Common Context is used and extended compared to [RFC8613]. The Common Context may be acquired from the Group Manager (see Section 12).¶
The AEAD Algorithm (see Section 3.1 of [RFC8613]) identifies the COSE AEAD algorithm to use for encryption and decryption when messages are protected using the pairwise mode (see Section 8). This algorithm MUST provide integrity protection. If this parameter is not set, the pairwise mode is not used in the group.¶
The HKDF Algorithm (see Section 3.1 of [RFC8613]) identifies the used key derivation function, which MUST be one of the HMAC-based HKDF [RFC5869] algorithms defined for COSE (see Section 5.1 of [RFC9053]) and registered at [COSE.Algorithms].¶
The ID Context parameter (see Sections 3.1 and 3.3 of [RFC8613]) contains the Group Identifier (Gid) of the group. The choice of the Gid format is application specific. An example of specific formatting of the Gid is given in Appendix C. The application needs to specify how to handle potential collisions between Gids (see Section 14.6).¶
The Common IV parameter (see Section 3.1 of [RFC8613]) identifies the Common IV used in the group. Differently from OSCORE, the length of the Common IV is determined as follows.¶
If only one among the AEAD Algorithm and the Group Encryption Algorithm is set (see Section 2.1.1 and Section 2.1.7), the length of the Common IV is the nonce length for the set algorithm.¶
If both the AEAD Algorithm and the Group Encryption Algorithm are set, the length of the Common IV is the greatest nonce length among those of the two algorithms.¶
If the Group Encryption Algorithm is A128CTR, A192CTR, or A256CTR (see Section 4 of [RFC9459]), then the length of the nonce used by that algorithm is 12 bytes (see Section 2.1.7).¶
The new parameter Authentication Credential Format specifies the format of authentication credentials used in the group.¶
The new parameter Group Manager Authentication Credential specifies the authentication credential of the Group Manager, including the Group Manager's public key. The endpoint MUST achieve proof of possession of the corresponding private key. As an example, such proof of possession is possible to achieve during the join process provided by the realization of Group Manager specified in [I-D.ietf-ace-key-groupcomm-oscore]. Further details on the provisioning of the Group Manager's authentication credential to the group members are out of the scope of this document.¶
The new parameter Group Encryption Algorithm identifies the algorithm to use for encryption and decryption, when messages are protected in group mode (see Section 7). This algorithm MAY provide integrity protection. If it does not, integrity protection is still provided by the countersignature added to the message due to the use of the group mode. If this parameter is not set, the group mode is not used in the group.¶
In order to be eligible to use as Group Encryption Algorithm, a non-authenticated algorithm MUST ensure that the same key is not reused with the same IV or intermediate values used in the algorithm, e.g., for algorithms that increment the IV internally. If a non-authenticated algorithm does not fulfill the requirement above, that algorithm MUST NOT be used as Group Encryption Algorithm.¶
Examples of non-authenticated algorithms that can be used as Group Encryption Algorithm are A128CTR, A192CTR, and A256CTR (see Section 4 of [RFC9459]). When either of those three algorithms is used, the following applies:¶
A 12-byte nonce MUST be computed as defined in Section 3.3 of this document.¶
The Initialization Vector (IV) used in Section 4 of [RFC9459] is equivalent to the nonce above (12 bytes) concatenated with 0x00000000 (4 bytes), in this order.¶
The algorithm MUST NOT be used to encrypt a plaintext or decrypt a ciphertext whose length is larger than 64 GB (i.e., 236 bytes).¶
The non-authenticated algorithms A128CBC, A192CBC, and A256CBC (see Section 5 of [RFC9459]) MUST NOT be used as Group Encryption Algorithm.¶
Future specifications can admit alternative non-authenticated algorithms that can be used as Group Encryption Algorithm. When doing so, it MUST be defined how to securely compose the IV and possible intermediate values used in the algorithm, building on the nonce computed as defined in Section 3.3 of this document. Absent such a specification, alternative non-authenticated algorithms MUST NOT be used as Group Encryption Algorithm.¶
The new parameter Signature Algorithm identifies the digital signature algorithm used for computing and verifying the countersignature on the COSE object (see Sections 3.2 and 3.3 of [RFC9338]), when messages are protected in group mode (see Section 7). If this parameter is not set, the group mode is not used in the group.¶
The new parameter Signature Encryption Key specifies the symmetric key for deriving a keystream to encrypt/decrypt a countersignature (see Section 4.2) when a message is protected in group mode (see Section 7).¶
The Signature Encryption Key is derived as defined for Sender/Recipient Keys in Section 3.2.1 of [RFC8613], with the following differences.¶
The 'id' element of the 'info' array is the empty byte string.¶
The 'alg_aead' element of the 'info' array specifies the Group Encryption Algorithm from the Common Context (see Section 2.1.7), encoded as a CBOR integer or text string, consistently with the "Value" field in the "COSE Algorithms" Registry for this algorithm.¶
The 'type' element of the 'info' array is "SEKey". The label is an ASCII string and does not include a trailing NUL byte.¶
L and the 'L' element of the 'info' array are the size of the key for the Group Encryption Algorithm specified in the Common Context (see Section 2.1.7), in bytes. While the obtained Signature Encryption Key is never used with the Group Encryption Algorithm, its length was chosen to obtain a matching level of security.¶
The new parameter Pairwise Key Agreement Algorithm identifies the elliptic curve Diffie-Hellman algorithm used to derive a static-static Diffie-Hellman shared secret, from which pairwise keys are derived (see Section 2.5.1) to protect messages with the pairwise mode (see Section 8). If this parameter is not set, the pairwise mode is not used in the group.¶
When two endpoints compute their Diffie-Hellman shared secret, the Pairwise Key Agreement Algorithm takes as input the static-static Diffie-Hellman keys of the two endpoints. The lifetime of those keys is the same as the lifetime of the authentication credentials that the two endpoints use in the group. As detailed in Section 2.5.1, the derivation of the pairwise keys takes as input not only the Diffie-Hellman shared secret, but also group keying material from the latest established Security Context.¶
If the HKDF Algorithm specified in the Common Context is "HKDF SHA-256" (identified as "HMAC 256/256"), then the Pairwise Key Agreement Algorithm is "ECDH-SS + HKDF-256" (COSE algorithm encoding: -27).¶
If the HKDF Algorithm specified in the Common Context is "HKDF SHA-512" (identified as "HMAC 512/512"), then the Pairwise Key Agreement Algorithm is "ECDH-SS + HKDF-512" (COSE algorithm encoding: -28).¶
Note that the HKDF Algorithm in the Common Context is denoted by the corresponding COSE HMAC Algorithm. For example, the HKDF Algorithm "HKDF SHA-256" is specified as the HMAC Algorithm "HMAC 256/256".¶
More generally, if Pairwise Key Agreement Algorithm is set, it MUST identify a COSE algorithm such that: i) it performs a direct ECDH Static-Static key agreement; and ii) it indicates the use of the same HKDF Algorithm used in the group as specified in the Common Context.¶
The Sender ID SHALL be unique for each endpoint in a group with a certain triplet (Master Secret, Master Salt, Group Identifier), see Section 3.3 of [RFC8613].¶
The maximum length of a Sender ID in bytes equals L minus 6, where L is determined as follows.¶
If only one among the AEAD Algorithm and the Group Encryption Algorithm is set (see Section 2.1.1 and Section 2.1.7), then L is the nonce length for the set algorithm.¶
If both the AEAD Algorithm and the Group Encryption Algorithm are set, then L is the smallest nonce length among those of the two algorithms.¶
With the exception of the authentication credential of the sender endpoint, a receiver endpoint can derive a complete Security Context from a received Group OSCORE message and the Common Context (see Section 2.3).¶
The authentication credentials in the Recipient Contexts can be retrieved from the Group Manager (see Section 12) upon joining the group. An authentication credential can alternatively be acquired from the Group Manager at a later time, for example the first time a message is received from a particular endpoint in the group (see Section 7.2 and Section 7.4).¶
For severely constrained devices, it may be infeasible to simultaneously handle the ongoing processing of a recently received message in parallel with the retrieval of the sender endpoint's authentication credential. Such devices can be configured to drop a received message for which there is no (complete) Recipient Context, and retrieve the sender endpoint's authentication credential in order to have it available to verify subsequent messages from that endpoint.¶
An endpoint may admit a maximum number of Recipient Contexts for a same Security Context, e.g., due to memory limitations. After reaching that limit, the endpoint has to delete a current Recipient Context to install a new one (see Section 2.6.1.2). It is up to the application to define the maximum number of Recipient Contexts for a same Security Context as well as policies for deleting Recipient Contexts.¶
OSCORE defines the derivation of Sender Context and Recipient Context (specifically, of Sender/Recipient Keys) and of the Common IV, from a set of input parameters (see Section 3.2 of [RFC8613]).¶
The derivation of Sender/Recipient Keys and of the Common IV defined in OSCORE applies also to Group OSCORE, with the following modifications compared to Section 3.2.1 of [RFC8613].¶
If Group Encryption Algorithm in the Common Context is set (see Section 2.1.7), then the 'alg_aead' element of the 'info' array MUST specify Group Encryption Algorithm from the Common Context as a CBOR integer or text string, consistently with the "Value" field in the "COSE Algorithms" Registry for this algorithm.¶
If Group Encryption Algorithm in the Common Context is not set, then the 'alg_aead' element of the 'info' array MUST specify AEAD Algorithm from the Common Context (see Section 2.1.1), as per Section 5.4 of [RFC8613].¶
When deriving the Common IV, the 'L' element of the 'info' array MUST specify the length of the Common IV in bytes, which is determined as defined in Section 2.1.4.¶
The authentication credentials of the endpoints in a group MUST be encoded according to the format used in the group, as indicated by the Authentication Credential Format parameter in the Common Context (see Section 2.1.5). The authentication credential of the Group Manager SHOULD be encoded according to that same format, in order to limit the number of formats that the group members have to support and handle, unless it is infeasible or impractical for the particular realization or instance of the Group Manager to have an own authentication credential encoded in that same format.¶
The format of authentication credentials MUST provide the public key and a comprehensive set of information related to the public key algorithm, including, e.g., the used elliptic curve (when applicable). If Group Encryption Algorithm in the Common Context is not set (see Section 2.1.7), then the public key algorithm is the Pairwise Key Agreement Algorithm used in the group (see Section 2.1.10), else the Signature Algorithm used in the group (see Section 2.1.8).¶
Examples of formats of authentication credentials are CBOR Web Tokens (CWTs) and CWT Claims Sets (CCSs) [RFC8392], X.509 certificates [RFC5280], and C509 certificates [I-D.ietf-cose-cbor-encoded-cert].¶
If the authentication credentials are X.509 certificates or C509 certificates, the public key algorithm is identified by the "algorithm" field of the "SubjectPublicKeyInfo" structure, and by the "subjectPublicKeyAlgorithm" element, respectively.¶
If authentication credentials are CBOR Web Tokens (CWTs) or CWT Claims Sets (CCSs), then a COSE Key structure and its "kty" and "crv" parameters identify the types of pertinent public key algorithms. For example: the pair ("crv" = X25519, "kty" = OKP) indicates that the public key is meant to be used with X25519 ECDH key agreement; the pair ("crv" = Ed25519, "kty" = OKP) indicates that the public key is meant to be used with the signature algorithm EdDSA; the pair ("crv" = P-256, "kty" = EC2) indicates that the public key is meant to be used with the signature algorithm ECDSA and/or with P-256 ECDH key agreement.¶
Authentication credentials are used to derive pairwise keys (see Section 2.5.1) and are included in the external additional authenticated data when processing messages (see Section 3.4). In both these cases, an endpoint in a group MUST treat authentication credentials as opaque data, i.e., by considering the same binary representation made available to other endpoints in the group, possibly through a designated trusted source (e.g., the Group Manager).¶
For example, an X.509 certificate is provided as its direct binary serialization. If C509 certificates or CWTs are used as authentication credentials, each is provided as the binary serialization of a (possibly tagged) CBOR array. If CCSs are used as authentication credentials, each is provided as the binary serialization of a (possibly tagged) CBOR map.¶
If authentication credentials are CWTs or CCSs, then the untagged CWT or CCS associated with an entity is stored in the Security Context and used as authentication credential for that entity.¶
If authentication credentials are X.509 / C509 certificates, CWTs, or CCSs and the authentication credential associated with an entity is provided within a chain or a bag, then only the end-entity certificate or end-entity untagged CWT / CCS is stored in the Security Context and used as authentication credential for that entity.¶
Storing whole authentication credentials rather than only a subset of those may result in a non-negligible storage overhead. On the other hand, it also ensures that authentication credentials are correctly used in a simple, flexible and non-error-prone way, also taking into account future credential formats as entirely new or extending existing ones. In particular, it is ensured that:¶
When used to derive pairwise keys and when included in the external additional authenticated data, authentication credentials can also specify possible metadata and parameters related to the included public key. Besides the public key algorithm, these comprise other relevant pieces of information such as key usage, expiration time, issuer, and subject.¶
All endpoints using another endpoint's authentication credential use exactly the same binary serialization, as obtained and distributed by the credential provider (e.g., the Group Manager), and as originally crafted by the credential issuer. In turn, this does not require to define and maintain canonical subsets of authentication credentials and their corresponding encoding, and spares endpoints from error-prone re-encoding operations.¶
Depending on the particular deployment and the intended group size, limiting the storage overhead of endpoints in a group can be an incentive for system/network administrators to prefer using a compact format of authentication credentials in the first place.¶
In certain Elliptic Curve Cryptographic schemes, it is possible to use public/private key pairs with both signature operations (ECDSA or EdDSA) and key agreement operations (ECDH). This section specifies the derivation of "pairwise keys" for use in the pairwise mode defined in Section 8.¶
Group OSCORE keys used for both signature operations and key agreement operations MUST be used only for purposes related to Group OSCORE. These include the processing of messages with Group OSCORE, as well as performing proof of possession of private keys, e.g., upon joining a group through the Group Manager (see Section 12).¶
Using the Group OSCORE Security Context (see Section 2), a group member can derive AEAD keys, to protect point-to-point communication between itself and each other endpoint X in the group by means of the AEAD Algorithm from the Common Context (see Section 2.1.1).¶
Analogous to the construction used by OSCORE in Section 3.2.1 of [RFC8613], the key derivation of these so-called pairwise keys relies on an HKDF algorithm and is as defined below:¶
Pairwise Sender Key = HKDF(Sender Key, IKM-Sender, info, L) Pairwise Recipient Key = HKDF(Recipient Key, IKM-Recipient, info, L) with IKM-Sender = Sender Auth Cred | Recipient Auth Cred | Shared Secret IKM-Recipient = Recipient Auth Cred | Sender Auth Cred | Shared Secret¶
where:¶
The Pairwise Sender Key is the AEAD key for processing outgoing messages addressed to endpoint X.¶
The Pairwise Recipient Key is the AEAD key for processing incoming messages from endpoint X.¶
HKDF is the OSCORE HKDF algorithm [RFC8613] from the Common Context.¶
The Sender Key from the Sender Context is used as salt in the HKDF, when deriving the Pairwise Sender Key.¶
The Recipient Key from the Recipient Context associated with endpoint X is used as salt in the HKDF, when deriving the Pairwise Recipient Key.¶
Sender Auth Cred is the endpoint's own authentication credential from the Sender Context.¶
Recipient Auth Cred is the endpoint X's authentication credential from the Recipient Context associated with the endpoint X.¶
The Shared Secret is computed as a cofactor Diffie-Hellman shared secret, see Section 5.7.1.2 of [NIST-800-56A], using the Pairwise Key Agreement Algorithm. The endpoint uses its private key from the Sender Context and the other endpoint's public key included in Recipient Auth Cred. Note the requirement of validation of public keys in Section 14.16.¶
In case the other endpoint's public key has COSE Key Type "EC2" [RFC9053] (e.g., for the curves P-256, P-384, and P-521), then the public key is used as is. In case the other endpoint's public key has COSE Key Type "OKP" [RFC9053], the procedure is described in Section 5 of [RFC7748]. In particular, if the public key is for X25519 or X448, it is used as is. Otherwise, if the public key is for the curve Ed25519 or Ed448, it is first mapped to Montgomery coordinates (see Section 2.5.2).¶
IKM-Sender is the Input Keying Material (IKM) used in the HKDF for the derivation of the Pairwise Sender Key. IKM-Sender is the byte string concatenation of Sender Auth Cred, Recipient Auth Cred, and the Shared Secret. The authentication credentials Sender Auth Cred and Recipient Auth Cred are binary encoded as defined in Section 2.4.¶
IKM-Recipient is the Input Keying Material (IKM) used in the HKDF for the derivation of the Pairwise Recipient Key. IKM-Recipient is the byte string concatenation of Recipient Auth Cred, Sender Auth Cred, and the Shared Secret. The authentication credentials Recipient Auth Cred and Sender Auth Cred are binary encoded as defined in Section 2.4.¶
info and L are as defined in Section 3.2.1 of [RFC8613]. That is:¶
The 'alg_aead' element of the 'info' array takes the value of AEAD Algorithm from the Common Context (see Section 2.1.1).¶
L and the 'L' element of the 'info' array are the size of the key for the AEAD Algorithm from the Common Context (see Section 2.1.1), in bytes.¶
If EdDSA asymmetric keys are used, the Edward coordinates are mapped to Montgomery coordinates using the maps defined in Sections 4.1 and 4.2 of [RFC7748], before using the X25519 or X448 function defined in Section 5 of [RFC7748]. For further details, see Section 2.5.2. ECC asymmetric keys in Montgomery or Weierstrass form are used directly in the key agreement algorithm, without coordinate mapping.¶
As long as any two group members preserve the same asymmetric keys, their Diffie-Hellman shared secret does not change across updates of the group keying material. The lifetime of those keys is the same as the lifetime of the authentication credentials Sender Auth Cred and Recipient Auth Cred.¶
After establishing a partially or completely new Security Context (see Section 2.6 and Section 12.2), the old pairwise keys MUST be deleted. Since new Sender/Recipient Keys are derived from the new group keying material (see Section 2.2), every group member MUST use the new Sender/Recipient Keys when deriving new pairwise keys.¶
The y-coordinate of the other endpoint's Ed25519 public key is decoded as specified in Section 5.1.3 of [RFC8032]. The Curve25519 u-coordinate is recovered as u = (1 + y) / (1 - y) (mod p) following the map in Section 4.1 of [RFC7748]. Note that the mapping is not defined for y = 1, and that y = -1 maps to u = 0 which corresponds to the neutral group element and thus will result in a degenerate shared secret. Therefore, implementations MUST abort if the y-coordinate of the other endpoint's Ed25519 public key is 1 or -1 (mod p).¶
The private signing key byte strings (i.e., the lower 32 bytes used for generating the public key, see Step 1 of Section 5.1.5 of [RFC8032]) are decoded the same way for signing in Ed25519 and scalar multiplication in X25519. Hence, in order to compute the shared secret, the endpoint applies the X25519 function to the Ed25519 private signing key byte string and the encoded u-coordinate byte string as specified in Section 5 of [RFC7748].¶
The y-coordinate of the other endpoint's Ed448 public key is decoded as specified in Section 5.2.3. of [RFC8032]. The Curve448 u-coordinate is recovered as u = y^2 * (d * y^2 - 1) / (y^2 - 1) (mod p) following the map from "edwards448" in Section 4.2 of [RFC7748], and also using the relation x^2 = (y^2 - 1)/(d * y^2 - 1) from the curve equation. Note that the mapping is not defined for y = 1 or -1. Therefore, implementations MUST abort if the y-coordinate of the peer endpoint's Ed448 public key is 1 or -1 (mod p).¶
The private signing key byte strings (i.e., the lower 57 bytes used for generating the public key, see Step 1 of Section 5.2.5 of [RFC8032]) are decoded the same way for signing in Ed448 and scalar multiplication in X448. Hence, in order to compute the shared secret, the endpoint applies the X448 function to the Ed448 private signing key byte string and the encoded u-coordinate byte string as specified in Section 5 of [RFC7748].¶
When using any of its Pairwise Sender Keys, a sender endpoint including the 'Partial IV' parameter in the protected message MUST use the current fresh value of the Sender Sequence Number from its Sender Context (see Section 2.2). That is, the same Sender Sequence Number space is used for all outgoing messages protected with Group OSCORE, thus limiting both storage and complexity.¶
When combining communications with the group mode and the pairwise mode, this may result in the Partial IV values moving forward more often than when using OSCORE [RFC8613]. This can happen when a client engages in frequent or long sequences of one-to-one exchanges with servers in the group, by sending requests over unicast. In turn, this contributes to a sooner exhaustion of the Sender Sequence Number space of the client, which would then require to take actions for deriving a new Sender Context before resuming communications in the group (see Section 2.6.2).¶
If the pairwise mode is supported, the Security Context additionally includes the Pairwise Key Agreement Algorithm and the pairwise keys, as described at the beginning of Section 2.¶
The pairwise keys as well as the shared secrets used in their derivation (see Section 2.5.1) may be stored in memory or recomputed every time they are needed. The shared secret changes only when a public/private key pair used for its derivation changes, which results in the pairwise keys also changing. Additionally, the pairwise keys change if the Sender ID changes or if a new Security Context is established for the group (see Section 2.6.3). In order to optimize protocol performance, an endpoint may store the derived pairwise keys for easy retrieval.¶
In the pairwise mode, the Sender Context includes the Pairwise Sender Keys to use with the other endpoints (see Figure 1). In order to identify the right key to use, the Pairwise Sender Key for endpoint X may be associated with the Recipient ID of endpoint X, as defined in the Recipient Context (i.e., the Sender ID from the point of view of endpoint X). In this way, the Recipient ID can be used to lookup for the right Pairwise Sender Key. This association may be implemented in different ways, e.g., by storing the pair (Recipient ID, Pairwise Sender Key) or linking a Pairwise Sender Key to a Recipient Context.¶
It is RECOMMENDED that the long-term part of the Security Context is stored in non-volatile memory, or that it can otherwise be reliably accessed throughout the operation of the group, e.g., after device reboots. However, also data in the long-term part of the Security Context may need to be updated, for example due to scheduled key renewal, new or re-joining members in the group, or the fact that the endpoint changes Sender ID (see Section 2.6.3).¶
The data in the varying part of the Security Context are updated by the endpoint when executing the security protocol, but may be lost (see Section 2.6.1) or become outdated by exhaustion of Sender Sequence Numbers (see Section 2.6.2).¶
An endpoint may lose the varying part of its Security Context due to accidental events, e.g., if a reboot occurred in an unprepared way (see Section 2.6.1.1) or due to a deliberately deleted Recipient Context (see Section 2.6.1.2).¶
If it is not feasible or practically possible to store and maintain up-to-date the varying part in non-volatile memory (e.g., due to limited number of write operations), the endpoint MUST be able to detect a loss of the varying part of the Security Context, to prevent the re-use of a nonce with the same key and to handle incoming replayed messages.¶
In case a loss of the Sender Context and/or of the Recipient Contexts is detected (e.g., if a reboot occurred in an unprepared way), the endpoint MUST NOT protect further messages using this Security Context, to avoid reusing a nonce with the same key.¶
Before resuming its operations in the group, the endpoint MUST retrieve new Security Context parameters from the Group Manager (see Section 2.6.3) and use them to derive a new Sender Context and Recipient Contexts (see Section 2.2). Since the new Sender Context includes newly derived encryption keys, an endpoint will not reuse the same pair (key, nonce), even when it is a server using the Partial IV of (old re-injected) requests to build the nonce for protecting the responses.¶
From then on, the endpoint MUST use the latest installed Sender Context to protect outgoing messages. Newly derived Recipient Contexts will have a Replay Window which is initialized as valid.¶
If an endpoint is not configured as a silent server and is not able to establish an updated Sender Context, e.g., because of lack of connectivity with the Group Manager, then the endpoint MUST NOT protect further messages using the current Security Context and MUST NOT accept incoming messages from other group members, as it is currently unable to detect possible replays.¶
If an endpoint is configured as a silent server and is not able to establish an updated Security Context, e.g., because of lack of connectivity with the Group Manager, then the endpoint MUST NOT accept incoming messages from other group members, as it is currently unable to detect possible replays.¶
The Security Context may contain a large and variable number of Recipient Contexts. While Group OSCORE in itself does not establish a maximum number of Recipient Contexts, there are circumstances by which implementations might choose to discard Recipient Contexts or have to do so in accordance with enforced application policies. Such circumstances include the need to reclaim memory or other resources on the node hosting the endpoint, for example because the predefined maximum number of Recipient Contexts has been reached in the Security Context (see Section 2.2). Implementations can provide means for the application to gain knowledge about the deliberate deletion of Recipient Contexts, e.g., through notifications sent to the application and/or logs made available to the application.¶
When a Recipient Context is deleted, this not only results in losing information about previously received messages from the corresponding other endpoint. It also results in the inability to be aware of the Security Contexts from which information has been lost.¶
Therefore, if the Recipient Context is derived again from the same Security Context, there is a risk that a replayed message is not detected. If any Recipient Context associated with any peer has ever been deleted from the current Security Context, then the Replay Window of any new Recipient Context in this Security Context MUST be initialized as invalid. An exception applies when the deleted Recipient Context was created upon receiving a message and that message was not verified successfully (see Section 7.2, Section 7.4, Section 8.4, and Section 8.6). Messages associated with a Recipient Context that has an invalid Replay Window MUST NOT be delivered to the application.¶
If the endpoint receives a message to be processed with any such new Recipient Context whose Replay Window is invalid, then the endpoint MUST take one of the following courses of action.¶
The endpoint discards the message.¶
The endpoint follows the procedure based on the CoAP Echo Option [RFC9175] and specified in Section 9, in order to establish a valid Replay Window. In particular, the endpoint MUST use its Partial IV when generating the nonce and MUST include the Partial IV in the response message conveying the Echo Option. If the endpoint supports the CoAP Echo Option, then it is RECOMMENDED to take this course of action.¶
The endpoint retrieves or waits for new Security Context parameters from the Group Manager and derives new Sender and Recipient Contexts, as defined in Section 2.6.1.1. In this case the Replay Windows of all Recipient Contexts become valid if they are not already. In particular, any invalid Replay Window is re-initialized as valid and with 0 as its current lower limit.¶
Since an endpoint increments its Sender Sequence Number for each new outgoing message including a Partial IV, an endpoint can eventually exhaust the Sender Sequence Number space.¶
Implementations MUST be able to detect an exhaustion of Sender Sequence Number space, after the endpoint has consumed the largest usable value. This may be influenced by additional limitations besides the mere 40-bit size limit of the Partial IV.¶
Upon exhausting the Sender Sequence Number space, the endpoint MUST NOT use this Security Context to protect further messages including a Partial IV.¶
When approaching the exhaustion of the Sender Sequence Number space, the endpoint SHOULD inform the Group Manager, retrieve new Security Context parameters from the Group Manager (see Section 2.6.3), and use them to derive a new Sender Context (see Section 2.2). It is RECOMMENDED that the endpoint takes this course of action with some margin, i.e., well before exhausting the Sender Sequence Number space, in order to avoid a period of inability to protect messages including a Partial IV.¶
From then on, the endpoint MUST use its latest installed Sender Context to protect outgoing messages.¶
The Group Manager can assist an endpoint with an incomplete Sender Context to retrieve missing data of the Security Context and thereby become fully operational in the group again. The two main options for the Group Manager are: i) assignment of a new Sender ID to the endpoint (see Section 2.6.3.1); and ii) establishment of a new Security Context for the group (see Section 2.6.3.2). The update of the Replay Window in each of the Recipient Contexts is discussed in Section 2.6.1.¶
As group membership changes, or as group members get new Sender IDs (see Section 2.6.3.1), so do the relevant Recipient IDs that the other endpoints need to keep track of. As a consequence, group members may end up retaining stale Recipient Contexts that are no longer useful to verify incoming secure messages.¶
The Recipient ID ('kid') SHOULD NOT be considered as a persistent and reliable identifier of a group member. Such an indication can be achieved only by using that member's public key, when verifying countersignatures of received messages (in group mode), or when verifying messages integrity-protected with pairwise keying material derived from authentication credentials and associated asymmetric keys (in pairwise mode).¶
Furthermore, applications MAY define policies to: i) delete (long-)unused Recipient Contexts and reduce the impact on storage space; as well as ii) check with the Group Manager that an authentication credential with the public key included therein is currently the one associated with a 'kid' value, after a number of consecutive failed verifications.¶
The Group Manager may assign a new Sender ID to an endpoint, while leaving the Gid, Master Secret, and Master Salt unchanged in the group. In this case, the Group Manager assigns a Sender ID that has not been used in the group since the latest time when the current Gid value was assigned to the group (see Section 12.2).¶
Having retrieved the new Sender ID, and potentially other missing data for the long-term part of the Security Context, the endpoint can derive a new Sender Context (see Section 2.2). When doing so, the endpoint resets the Sender Sequence Number in its Sender Context to 0, and derives a new Sender Key. This is in turn used to possibly derive new Pairwise Sender Keys.¶
From then on, the endpoint MUST use its latest installed Sender Context to protect outgoing messages.¶
The assignment of a new Sender ID may be the result of different processes. The endpoint may request a new Sender ID, e.g., because of the impending exhaustion of the Sender Sequence Number space (see Section 2.6.2). An endpoint may request to re-join the group, e.g., because of losing the varying part of its Security Context (see Section 2.6.1), and is provided with a new Sender ID together with the latest data for the long-term part of the Security Context.¶
For the other group members, the Recipient Context corresponding to the old Sender ID becomes stale (see Section 12.2).¶
The Group Manager may establish a new Security Context for the group (see Section 12.2). The Group Manager does not necessarily establish a new Security Context for the group if one member has an outdated Security Context (see Section 2.6.3.1), unless that was already planned or required for other reasons.¶
All the group members need to acquire new Security Context parameters from the Group Manager. Once having acquired new Security Context parameters, each group member performs the following actions.¶
From then on, it MUST NOT use the current Security Context to start processing new messages for the considered group.¶
It completes any ongoing message processing for the considered group.¶
It derives and installs a new Security Context. In particular:¶
It re-derives the keying material stored in its Sender Context and Recipient Contexts (see Section 2.2). The Master Salt used for the re-derivations is the updated Master Salt parameter if provided by the Group Manager, or the empty byte string otherwise.¶
It resets its Sender Sequence Number in its Sender Context to 0.¶
It re-initializes the Replay Window of each Recipient Context as valid and with 0 as its current lower limit.¶
For each long exchange where it is a client and that it wants to keep active, it sets the Response Number of each associated server as not initialized (see Section 5.1).¶
From then on, it can resume processing new messages for the considered group. In particular:¶
It MUST use its latest installed Sender Context to protect outgoing messages.¶
It SHOULD use only its latest installed Recipient Contexts to process incoming messages, unless application policies admit to temporarily retain and use the old, recent, Security Context (see Section 14.5.1).¶
The distribution of a new Gid and Master Secret may result in temporarily misaligned Security Contexts among group members. In particular, this may result in a group member not being able to process messages received right after a new Gid and Master Secret have been distributed. A discussion on practical consequences and possible ways to address them, as well as on how to handle the old Security Context, is provided in Section 14.5.¶
Building on Section 5 of [RFC8613], this section defines how to use COSE [RFC9052] to wrap and protect data in the original message. Like OSCORE, Group OSCORE uses the untagged COSE_Encrypt0 structure with an Authenticated Encryption with Associated Data (AEAD) algorithm. Unless otherwise specified, the following modifications to what is defined for OSCORE apply for both the group mode and the pairwise mode of Group OSCORE.¶
When protecting a message in group mode, the 'unprotected' field MUST additionally include the following parameter:¶
Countersignature0 version 2: its value is set to the countersignature of the COSE object.¶
The countersignature is computed by the sender as described in Sections 3.2 and 3.3 of [RFC9338], by using its private key and according to the Signature Algorithm in the Security Context.¶
In particular, the Countersign_structure contains the context text string "CounterSignature0", the external_aad as defined in Section 3.4 of this document, and the ciphertext of the COSE object as payload.¶
The literature commonly refers to a countersignature as a signature computed by an entity A over a document already protected by a different entity B.¶
However, the COSE_Countersignature0 structure belongs to the set of abbreviated countersignatures defined in Sections 3.2 and 3.3 of [RFC9338], which were designed primarily to deal with the problem of encrypted group messaging, but where it is required to know who originated the message.¶
Since the parameters for computing or verifying the abbreviated countersignature generated by A are provided by the same context used to describe the security processing performed by B and to be countersigned, these structures are applicable also when the two entities A and B are actually the same one, like the sender of a Group OSCORE message protected in group mode.¶
The value of the 'kid' parameter in the 'unprotected' field of response messages MUST be set to the Sender ID of the endpoint transmitting the message, if the request was protected in group mode. That is, unlike in [RFC8613], the 'kid' parameter is always present in responses to a request that was protected in group mode.¶
The value of the 'kid context' parameter in the 'unprotected' field of request messages MUST be set to the ID Context, i.e., the Group Identifier value (Gid) of the group. That is, unlike in [RFC8613], the 'kid context' parameter is always present in requests.¶
The nonce is constructed like in OSCORE, with the difference that Step 4 in Section 5.2 of [RFC8613] is replaced with:¶
and then XOR with X bytes from the Common IV's start, where X is the length in bytes of the nonce.¶
For example, if X = 7 and the Common IV is 0x00112233445566778899aabbcc (13 bytes), then the bytes to XOR are 0x00112233445566 (7 bytes).¶
The constructed nonce is used both by the AEAD Algorithm (see Section 2.1.1) and by the Group Encryption Algorithm (see Section 2.1.7), independent of whether they are AEAD or plain encryption algorithms. Algorithms that do not use a nonce are not supported, as per Section 2.1.7.¶
The external_aad of the Additional Authenticated Data (AAD) is different compared to OSCORE [RFC8613] and is defined in this section.¶
The same external_aad structure is used in group mode and pairwise mode for encryption/decryption (see Section 5.3 of [RFC9052]), as well as in group mode for computing and verifying the countersignature (see Sections 3.2 and 3.3 of [RFC9338]).¶
In particular, the external_aad includes also the Signature Algorithm, the Group Encryption Algorithm, the Pairwise Key Agreement Algorithm, the value of the 'kid context' in the COSE object of the request, the OSCORE Option of the protected message, the sender's authentication credential, and the Group Manager's authentication credential.¶
The external_aad SHALL be a CBOR array wrapped in a bstr object as defined below, following the notation of [RFC8610]:¶
external_aad = bstr .cbor aad_array
aad_array = [
oscore_version : uint,
algorithms : [alg_aead : int / tstr / null,
alg_group_enc : int / tstr / null,
alg_signature : int / tstr / null,
alg_pairwise_key_agreement : int / tstr / null],
request_kid : bstr,
request_piv : bstr,
options : bstr,
request_kid_context : bstr,
OSCORE_option : bstr,
sender_cred : bstr,
gm_cred : bstr
]
Compared with Section 5.4 of [RFC8613], the aad_array has the following differences.¶
The 'algorithms' array is extended as follows.¶
The parameter 'alg_aead' MUST be set to the CBOR simple value null (0xf6) if the parameter AEAD Algorithm is not set in the Common Context of the Security Context used (see Section 2.1.1). Otherwise, regardless of whether the endpoint supports the pairwise mode or not, this parameter MUST specify AEAD Algorithm from the Common Context (see Section 2.1.1) as per Section 5.4 of [RFC8613].¶
Furthermore, the 'algorithms' array additionally includes:¶
'alg_group_enc', which specifies Group Encryption Algorithm from the Common Context of the Security Context used (see Section 2.1.7). This parameter MUST be set to the CBOR simple value null (0xf6) if the parameter Group Encryption Algorithm in the Common Context is not set. Otherwise, regardless of whether the endpoint supports the group mode or not, this parameter MUST specify Group Encryption Algorithm as a CBOR integer or text string, consistently with the "Value" field in the "COSE Algorithms" Registry for this algorithm.¶
'alg_signature', which specifies Signature Algorithm from the Common Context of the Security Context used (see Section 2.1.8). This parameter MUST be set to the CBOR simple value null (0xf6) if the parameter Signature Algorithm in the Common Context is not set. Otherwise, regardless of whether the endpoint supports the group mode or not, this parameter MUST specify Signature Algorithm as a CBOR integer or text string, consistently with the "Value" field in the "COSE Algorithms" Registry for this algorithm.¶
'alg_pairwise_key_agreement', which specifies Pairwise Key Agreement Algorithm from the Common Context of the Security Context used (see Section 2.1.10). This parameter MUST be set to the CBOR simple value null (0xf6) if Pairwise Key Agreement Algorithm in the Common Context is not set. Otherwise, regardless of whether the endpoint supports the pairwise mode or not, this parameter MUST specify Pairwise Key Agreement Algorithm as a CBOR integer or text string, consistently with the "Value" field in the "COSE Algorithms" Registry for this algorithm.¶
The new element 'request_kid_context' contains the value of the 'kid context' in the COSE object of the request (see Section 3.2).¶
This enables endpoints to safely keep a long exchange active beyond a possible change of Gid (i.e., of ID Context), following a group rekeying (see Section 12.2). In fact, it ensures that every response within a long exchange cryptographically matches with only one request (i.e., the request associated with that long exchange), rather than with multiple requests that were protected with different keying material but share the same 'request_kid' and 'request_piv' values.¶
The new element 'OSCORE_option', containing the value of the OSCORE Option present in the protected message, encoded as a binary string. This prevents the attack described in Section 14.7 when using the group mode, as further explained in Section 14.7.2.¶
Note for implementation: this construction requires the OSCORE Option of the message to be generated and finalized before computing the ciphertext of the COSE_Encrypt0 object (when using the group mode or the pairwise mode) and before calculating the countersignature (when using the group mode). Also, the aad_array needs to be large enough to contain the largest possible OSCORE Option.¶
The new element 'sender_cred', containing the sender's authentication credential. This parameter MUST be set to a CBOR byte string, which encodes the sender's authentication credential in its original binary representation made available to other endpoints in the group (see Section 2.4).¶
The new element 'gm_cred', containing the Group Manager's authentication credential. This parameter MUST be set to a CBOR byte string, which encodes the Group Manager's authentication credential in its original binary representation made available to other endpoints in the group (see Section 2.4). This prevents the attack described in Section 14.8.¶
Group OSCORE relies on a header compression mechanism similar to the one used by OSCORE and specified in Section 4.1. Examples are provided in Section 4.3.¶
The OSCORE header compression defined in Section 6 of [RFC8613] is used for compactly encoding the COSE_Encrypt0 object specified in Section 3 of this document, with the following differences.¶
When the Group OSCORE message is protected in group mode, the message payload SHALL encode the ciphertext of the COSE object, concatenated with the encrypted countersignature of the COSE object. That is:¶
The plain, original countersignature of the COSE object, namely SIGNATURE, is specified in the "Countersignature0 version 2" parameter within the 'unprotected' field of the COSE object (see Section 3.1).¶
The encrypted countersignature, namely ENC_SIGNATURE, is computed as¶
ENC_SIGNATURE = SIGNATURE XOR KEYSTREAM¶
where KEYSTREAM is derived as per Section 4.2.¶
When the Group OSCORE message is protected in pairwise mode, the message payload SHALL encode the ciphertext of the COSE object.¶
This document defines the usage of the sixth least significant bit, called "Group Flag", in the first byte of the OSCORE Option containing the OSCORE flag bits. This flag bit is specified in Section 15.1.¶
The Group Flag MUST be set to 1 if the Group OSCORE message is protected using the group mode (see Section 7).¶
The Group Flag MUST be set to 0 if the Group OSCORE message is protected using the pairwise mode (see Section 8). The Group Flag MUST also be set to 0 for ordinary OSCORE messages processed according to [RFC8613].¶
The following defines how an endpoint derives the keystream KEYSTREAM, used to encrypt/decrypt the countersignature of an outgoing/incoming message M protected in group mode.¶
The keystream SHALL be derived as follows, by using the HKDF Algorithm from the Common Context (see Section 3.2 of [RFC8613]), which consists of composing the HKDF-Extract and HKDF-Expand steps [RFC5869].¶
KEYSTREAM = HKDF(salt, IKM, info, L)¶
The input parameters of HKDF are as follows.¶
salt takes as value the Partial IV (PIV) used to protect M. Note that, if M is a response, salt takes as value either: i) the fresh Partial IV generated by the server and included in the response; or ii) the same Partial IV of the request generated by the client and not included in the response.¶
IKM is the Signature Encryption Key from the Common Context (see Section 2.1.9).¶
info is the serialization of a CBOR array with the structure defined below, following the notation of [RFC8610]:¶
info = [
id : bstr,
id_context : bstr,
type : bool,
L : uint
]
¶
where:¶
id is the Sender ID of the endpoint that generated PIV.¶
id_context is the ID Context (Gid) used when protecting M.¶
Note that, in the case of group rekeying, a server might use a different Gid when protecting a response, compared to the Gid that it used to verify (that the client used to protect) the request, see Section 7.3.¶
type is the CBOR simple value true (0xf5) if M is a request, or the CBOR simple value false (0xf4) otherwise.¶
L is the size of the countersignature, as per Signature Algorithm from the Common Context (see Section 2.1.8), in bytes.¶
This section covers a list of OSCORE Header Compression examples of Group OSCORE used in group mode (see Section 4.3.1) or in pairwise mode (see Section 4.3.2).¶
The examples assume that the COSE_Encrypt0 object is set (which means the CoAP message and cryptographic material is known). Note that the examples do not include the full CoAP unprotected message or the full Security Context, but only the input necessary to the compression mechanism, i.e., the COSE_Encrypt0 object. The output is the compressed COSE object as defined in Section 4.1 and divided into two parts, since the object is transported in two CoAP fields: OSCORE Option and payload.¶
The examples assume that the plaintext (see Section 5.3 of [RFC8613]) is 6 bytes long, and that the AEAD tag is 8 bytes long, hence resulting in a ciphertext which is 14 bytes long. When using the group mode, the COSE_Countersignature0 byte string as described in Section 3 is assumed to be 64 bytes long.¶
Request with ciphertext = 0xaea0155667924dff8a24e4cb35b9, kid = 0x25, Partial IV = 5 and kid context = 0x44616c.¶
Before compression (96 bytes):¶
[
/ protected / h'',
/ unprotected / {
/ kid / 4 : h'25',
/ Partial IV / 6 : h'05',
/ kid context / 10 : h'44616c',
/ Countersignature0 version 2 / 12 : h'66e6d9b0
db009f3e105a673f8855611726caed57f530f8cae9d0b168
513ab949fedc3e80a96ebe94ba08d3f8d3bf83487458e2ab
4c2f936ff78b50e33c885e35'
},
/ ciphertext / h'aea0155667924dff8a24e4cb35b9'
]
¶
After compression (85 bytes):¶
Flag byte: 0b00111001 = 0x39 (1 byte) Option Value: 0x39 05 03 44 61 6c 25 (7 bytes) Payload: 0xaea0155667924dff8a24e4cb35b9 de9e ... f1 (14 bytes + size of the encrypted countersignature)¶
Response with ciphertext = 0x60b035059d9ef5667c5a0710823b, kid = 0x52 and no Partial IV.¶
Before compression (88 bytes):¶
[
/ protected / h'',
/ unprotected / {
/ kid / 4 : h'52',
/ Countersignature0 version 2 / 12 : h'f5b659b8
24487eb349c5c5c8a3fe401784cade2892725438e8be0fab
daa2867ee6d29f68edb0818e50ebf98c28b923d0205f5162
e73662e27c1a3ec562a49b80'
},
/ ciphertext / h'60b035059d9ef5667c5a0710823b'
]
¶
After compression (80 bytes):¶
Flag byte: 0b00101000 = 0x28 (1 byte) Option Value: 0x28 52 (2 bytes) Payload: 0x60b035059d9ef5667c5a0710823b ca1e ... b3 (14 bytes + size of the encrypted countersignature)¶
Request with ciphertext = 0xaea0155667924dff8a24e4cb35b9, kid = 0x25, Partial IV = 5 and kid context = 0x44616c.¶
Before compression (29 bytes):¶
[
/ protected / h'',
/ unprotected / {
/ kid / 4 : h'25',
/ Partial IV / 6 : h'05',
/ kid context / 10 : h'44616c'
},
/ ciphertext / h'aea0155667924dff8a24e4cb35b9'
]
¶
After compression (21 bytes):¶
Flag byte: 0b00011001 = 0x19 (1 byte) Option Value: 0x19 05 03 44 61 6c 25 (7 bytes) Payload: 0xaea0155667924dff8a24e4cb35b9 (14 bytes)¶
Response with ciphertext = 0x60b035059d9ef5667c5a0710823b and no Partial IV.¶
Before compression (18 bytes):¶
[
/ protected / h'',
/ unprotected / {},
/ ciphertext / h'60b035059d9ef5667c5a0710823b'
]
¶
After compression (14 bytes):¶
Flag byte: 0b00000000 = 0x00 (1 byte) Option Value: 0x (0 bytes) Payload: 0x60b035059d9ef5667c5a0710823b (14 bytes)¶
Like OSCORE, Group OSCORE provides message binding of responses to requests, as well as uniqueness of (key, nonce) pairs (see Sections 7.1 and 7.2 of [RFC8613], respectively).¶
For each of its ongoing long exchanges, a client maintains one Response Number for each different server. Then, separately for each server, the client uses the associated Response Number to perform ordering and replay protection of responses received from that server within that long exchange (see Section 5.3.1).¶
That is, the Response Number has the same purpose that the Notification Number has in OSCORE (see Section 4.1.3.5.2 of [RFC8613]), but a client uses it for handling any response from the associated server within a long exchange.¶
Group OSCORE allows a long exchange to remain active, even if the group is rekeyed (thus changing the ID Context) or the client obtains a new Sender ID.¶
As defined in Section 7, this is achieved by the client and server(s) storing the 'kid' and 'kid context' used in the original request, throughout the whole duration of the long exchange.¶
Upon leaving the group or before re-joining the group, a group member MUST terminate all the ongoing long exchanges that it has started in the group as a client. This frees up the CoAP Token associated with the corresponding request.¶
If the application requires freshness, e.g., according to time- or event-based policies (see Section 2.5.1 of [RFC9175]), a server can use the approach in Section 9 as a variant of the Challenge-Response procedure based on the Echo Option [RFC9175] before delivering request messages from a client to the application.¶
Like in OSCORE [RFC8613], assuming an honest server, the message binding guarantees that a response is not older than the request it replies to. Therefore, building on Section 7.3 of [RFC8613], the following properties hold for Group OSCORE.¶
The freshness of a response can be assessed if it is received soon after the request.¶
For responses within a long exchange, this assessment gets weaker with time. If such responses are Observe notifications [RFC7641], it is RECOMMENDED that the client regularly re-register the observation.¶
If the request was neither a group request nor an Observe request, there is at most a single valid response and only from one, individually targeted server in the group. Thus, freshness can be assessed depending on when the request was sent.¶
It is not guaranteed that a misbehaving server did not create the response before receiving the request, i.e., Group OSCORE does not verify server aliveness.¶
For requests and responses, the received Partial IV allows a recipient to determine the relative order of requests or responses.¶
Like in OSCORE [RFC8613], the replay protection relies on the Partial IV of incoming messages. A server updates the Replay Window of its Recipient Contexts based on the Partial IV values in received request messages, which correspond to the Sender Sequence Numbers of the clients. Note that there can be large jumps in these Sender Sequence Number values, for example when a client exchanges unicast messages with other servers. The operation of validating the Partial IV and performing replay protection MUST be atomic. Section 2.6.1 and Section 2.6.3.2 describe the update of Replay Windows after the loss of data from the Security Context and the retrieving of new Security Context parameters.¶
The protection from replay of requests is performed as per Section 7.4 of [RFC8613], separately for each client and by leveraging the Replay Window in the corresponding Recipient Context. The protection from replay of responses in a long exchange is performed as defined in Section 5.3.1 of this document.¶
A client uses the method defined in this section in order to check whether a received response is a replay.¶
This especially applies to responses received within a long exchange, during which multiple such responses can be received from the same server to the corresponding request. These include Observe notifications [RFC7641]; and non-notification responses as a reply to a group request, which the client can receive until the CoAP Token value associated with the group request is freed up (see Section 3.1.6 of [I-D.ietf-core-groupcomm-bis]).¶
When sending a response (both successful and error), a server MUST include its Sender Sequence Number as Partial IV in the response, except when sending the first response to the corresponding request, in which case the Partial IV in the response MAY be omitted.¶
In order to protect against replay, the client SHALL maintain for each ongoing long exchange one Response Number for each different server. The Response Number is a non-negative integer containing the largest Partial IV of the responses received from that server during the long exchange, while using the same Security Context.¶
Then, separately for each server, the client uses the associated Response Number to perform ordering and replay protection of the responses from that server during the long exchange, by comparing their Partial IVs with one another and against the Response Number.¶
For every long exchange, the Response Number associated with a server is initialized to the Partial IV of the response from that server such that, within the long exchange, it is the first response from that server to include a Partial IV and to be successfully verified with the used Security Context. Note that, when a new Security Context is established in the group, the client sets the Response Number of each associated server as not initialized (see Section 2.6.3.2), hence later responses within the same long exchange and protected with the new Security Context will result in a new initialization of Response Numbers. Furthermore, for every long exchange, a client MUST only accept at most one response without Partial IV from each server, and treat it as the oldest response from that server within that long exchange.¶
During a long exchange, a client receiving a response containing a Partial IV SHALL compare the Partial IV with the Response Number associated with the replying server within that long exchange. The client MUST stop processing a response from a server, if that response has a Partial IV that has been previously received from that server during that long exchange, while using the same Security Context.¶
Applications MAY decide that a client only processes responses within a long exchange if those have a greater Partial IV than the Response Number associated with the replying server within that long exchange. This limits the storage overhead on the client to maintaining one Response Number per replying server within the long exchange. Conversely, more permissive applications can allow clients to also process responses that have a smaller Partial IV than the Response Number associated with the replying server. For a client, the ability to detect previously received Partial IVs while admitting the processing of such responses comes at the cost of additional storage overhead, for which a reasonable bound has to be defined by the application. A possible way to achieve that relies on using a sliding Replay Window uniquely paired with the replying server within the long exchange, similarly to that used by a server for detecting replayed requests.¶
If the verification of the response succeeds, and the received Partial IV (when included) was greater than the Response Number associated with the replying server, then the client SHALL overwrite that Response Number with the received Partial IV.¶
As long as a server uses the same Security Context to protect its responses to the same request, the client MUST consider the response with the highest Partial IV as the freshest response from that server among those protected with that Security Context, regardless of the order of arrival. Within a long exchange, implementations need to make sure that a response without Partial IV is considered the oldest response from the replying server within that long exchange.¶
The method defined in this section is not relevant for responses to requests that are neither group requests nor Observe requests. In fact, for each of such requests, there is at most one response and only from one individually targeted server in the group.¶
Upon receiving a protected message, a recipient endpoint retrieves a Security Context as in [RFC8613]. An endpoint MUST be able to distinguish between a Security Context to process OSCORE messages as in [RFC8613] and a Group OSCORE Security Context to process Group OSCORE messages as defined in this document.¶
The way to accomplish this distinction is implementation specific. For example, an endpoint can take into account the different structure of the Security Context defined in Section 2, e.g., based on the presence of Signature Algorithm and Pairwise Key Agreement Algorithm in the Common Context. Alternatively, at the cost of increasing storage, implementations can use an additional parameter in the Security Context, to explicitly mark that it is intended for processing Group OSCORE messages.¶
If either of the following conditions holds, a recipient endpoint MUST discard the incoming protected message:¶
The Group Flag is set to 0 and the retrieved Security Context is associated with an OSCORE group, but the endpoint does not support the pairwise mode or any of the following parameters is not set in the Security Context: the AEAD Algorithm and the Pairwise Key Agreement Algorithm.¶
The Group Flag is set to 1 and the retrieved Security Context is associated with an OSCORE group, but the endpoint does not support the group mode or any of the following parameters is not set in the Security Context: the Group Encryption Algorithm and the Signature Algorithm.¶
The Group Flag is set to 1 but there is no Security Context associated with an OSCORE group.¶
Future specifications may define how to process incoming messages protected with Security Contexts as in [RFC8613], when the Group Flag bit is set to 1.¶
Otherwise, if a Security Context associated with an OSCORE group is retrieved, the recipient endpoint processes the message with Group OSCORE, using the group mode (see Section 7) if the Group Flag is set to 1, or the pairwise mode (see Section 8) if the Group Flag is set to 0.¶
Note that if the Group Flag is set to 0 and the recipient endpoint retrieves a Security Context which is valid to process the message but is not associated with an OSCORE group, then the message is processed according to [RFC8613].¶
When using the group mode, messages are protected and processed as specified in [RFC8613] with the modifications described in this section. The security objectives of the group mode are discussed in Appendix A.2.¶
The possible use of the group mode is indicated by the Group Manager as part of the group data provided to new group members when joining the group, according to which the parameters Group Encryption Algorithm and Signature Algorithm in the Security Context are set (see Section 2).¶
During all the steps of the message processing, an endpoint MUST use the same Security Context for the considered group. That is, an endpoint MUST NOT install a new Security Context for that group (see Section 2.6.3.2) until the message processing is completed.¶
The group mode SHOULD be used to protect group requests intended for multiple recipients or for the whole group. This applies to both requests directly addressed to multiple recipients, e.g., sent by the client over multicast, as well as requests sent by the client over unicast to a proxy that forwards them to the intended recipients over multicast [I-D.ietf-core-groupcomm-bis]. Exceptions where the requirement above is not fulfilled and the pairwise mode is used to protect group requests include: the efficient discovery of a server's address in the group (see Section 8.1); or the enabling of simple constructions where a variation of the pairwise mode protects requests possibly intended to multiple servers, in such a way that the corresponding responses are effectively cacheable by intermediaries (e.g., see [I-D.ietf-core-cacheable-oscore]).¶
As per [RFC7252][I-D.ietf-core-groupcomm-bis], group requests sent over multicast are always Non-confirmable, and thus are not retransmitted by the CoAP messaging layer. Instead, applications should store such outgoing messages for a predefined, sufficient amount of time, in order to correctly perform potential retransmissions at the application layer. If performed, these retransmissions are repetitions of previous protected messages, which the sender endpoint does not protect again with Group OSCORE.¶
According to Section 5.2.3 of [RFC7252], "[i]f the request message is Non-confirmable, then the response SHOULD be returned in a Non-confirmable message as well. However, an endpoint MUST be prepared to receive (...) a Confirmable response in reply to a Non-confirmable request." Confirmable group requests are acknowledged when sent over non-multicast transports, as specified in [RFC7252].¶
Furthermore, endpoints in the group locally perform error handling and processing of invalid messages according to the same principles adopted in [RFC8613]. In addition, a recipient MUST stop processing and reject any message that is malformed and that does not follow the format specified in Section 3 of this document, or that is not cryptographically validated in a successful way as per the processing defined in Section 7.2 and Section 7.4 of this document.¶
In either case, it is RECOMMENDED that a server does not send back any error message in reply to a received request if either of the following conditions holds:¶
The server is not able to identify whether the received request is a group request, i.e., as sent to all servers in the group.¶
The server identifies the received request as a group request.¶
This prevents servers from replying with multiple error messages to a client sending a group request, so avoiding the risk of flooding and possibly congesting the network.¶
When using the group mode to protect a request, a client proceeds as described in Section 8.1 of [RFC8613] with the following modifications.¶
In Step 2, the Additional Authenticated Data is modified as described in Section 3 of this document.¶
In Step 4, the encryption of the COSE object is modified as described in Section 3 of this document. The encoding of the compressed COSE object is modified as described in Section 4 of this document. In particular, the Group Flag MUST be set to 1. The Group Encryption Algorithm from the Common Context MUST be used.¶
In Step 5, the countersignature is computed and the format of the OSCORE message is modified as described in Section 3 and Section 4 of this document. In particular, the payload of the Group OSCORE message includes also the encrypted countersignature.¶
In addition, the following applies when sending a request that establishes a long exchange.¶
If the client intends to keep the long exchange active beyond a possible change of Sender ID, the client MUST store the value of the 'kid' parameter from the request, and retain it until the long exchange is terminated. Even in case the client is individually rekeyed and receives a new Sender ID from the Group Manager (see Section 2.6.3.1), the client MUST NOT update the stored 'kid' parameter value associated with the long exchange and the corresponding request.¶
If the client intends to keep the long exchange active beyond a possible change of ID Context following a group rekeying (see Section 12.2), then the following applies.¶
The client MUST store the value of the 'kid context' parameter from the request, and retain it until the long exchange is terminated. Upon establishing a new Security Context with a new Gid as ID Context (see Section 2.6.3.2), the client MUST NOT update the stored 'kid context' parameter value associated with the long exchange and the corresponding request.¶
The client MUST store an invariant identifier of the group, which is immutable even if the Security Context of the group is re-established. For example, this invariant identifier can be the "group name" in [I-D.ietf-ace-key-groupcomm-oscore], where it is used for joining the group and retrieving the current group keying material from the Group Manager.¶
After a group rekeying, the client might have missed both the rekeying messages and the servers' first responses that are protected with the new Security Context and include the new ID Context (Gid) in the 'kid context' parameter (see Section 7.3). In such a case, while still not knowing the new ID Context (Gid) used in the group, the client is able to retrieve the current group keying material from the Group Manager, using the invariant identifier to unambiguously refer to the group.¶
Upon receiving a protected request with the Group Flag set to 1, following the procedure in Section 6, a server proceeds as described in Section 8.2 of [RFC8613] with the following modifications.¶
In Step 2, the decoding of the compressed COSE object follows Section 4 of this document. In particular:¶
If the server discards the request due to not retrieving a Security Context associated with the OSCORE group, the server MAY respond with a 4.01 (Unauthorized) error message. When doing so, the server MAY set an Outer Max-Age Option with value zero, and MAY include a descriptive string as diagnostic payload.¶
If the received 'kid context' matches an existing ID Context (Gid) but the received 'kid' does not match any Recipient ID in this Security Context, then the server MAY create a new Recipient Context for this Recipient ID and initialize it according to Section 3 of [RFC8613], and also retrieve the authentication credential associated with the Recipient ID to be stored in the new Recipient Context. Such a configuration is application specific. If the application does not specify dynamic derivation of new Recipient Contexts, then the server SHALL stop processing the request.¶
In Step 4, the Additional Authenticated Data is modified as described in Section 3 of this document.¶
In Step 6, the server also verifies the countersignature by using the public key from the client's authentication credential stored in the associated Recipient Context. In particular:¶
If the server does not have the public key of the client yet, the server MUST stop processing the request and MAY respond with a 5.03 (Service Unavailable) response. The response MAY include a Max-Age Option, indicating to the client the number of seconds after which to retry. If the Max-Age Option is not present, Section 5.10.5 of [RFC7252] specifies a default retry time of 60 seconds.¶
The signature verification as defined below SHOULD be performed before decrypting the COSE object. An exception applies to implementations that cannot perform the two steps in this order. Those implementations MUST ensure that no access to the plaintext is possible before a successful signature verification and MUST prevent any possible leak of time-related information that can yield side-channel attacks.¶
The server retrieves the encrypted countersignature ENC_SIGNATURE from the message payload, and computes the original countersignature SIGNATURE as¶
SIGNATURE = ENC_SIGNATURE XOR KEYSTREAM¶
where KEYSTREAM is derived as per Section 4.2.¶
The server verifies the original countersignature SIGNATURE as described in Sections 3.2 and 3.3 of [RFC9338] by using the client's public key and according to the Signature Algorithm in the Security Context.¶
In particular, the Countersign_structure contains the context text string "CounterSignature0", the external_aad as defined in Section 3.4 of this document, and the ciphertext of the COSE object as payload.¶
If the signature verification fails, the server SHALL stop processing the request, SHALL NOT update the Replay Window, and MAY respond with a 4.00 (Bad Request) response. Such a response MAY include an Outer Max-Age Option with value zero, and its diagnostic payload MAY contain a string, which, if present, MUST be "Decryption failed" as if the decryption of the COSE object had failed.¶
When decrypting the COSE object using the Recipient Key, the Group Encryption Algorithm from the Common Context MUST be used.¶
Additionally, if the used Recipient Context was created upon receiving this request and the message is not verified successfully, the server MAY delete that Recipient Context. When this behavior is specified by the application, it mitigates attacks that aim at overloading the server's storage.¶
If the server deletes the used Recipient Context in this particular circumstance, then this deletion does not require the server to initialize as invalid the Replay Window of any new Recipient Context created later within the Security Context (see Section 2.6.1.2).¶
A server SHOULD NOT process a request if the received Recipient ID ('kid') is equal to its own Sender ID in its own Sender Context. However, in some applications the server can prepare a request to be sent to itself (e.g., see [I-D.ietf-core-observe-multicast-notifications]), in which case such requests would be expected.¶
In addition, the following applies if the request establishes a long exchange and the server intends to reply with multiple responses.¶
The server MUST store the value of the 'kid' parameter from the request, and retain it until the last response has been sent. The server MUST NOT update the stored value of the 'kid' parameter associated with the request, even if the client is individually rekeyed and starts using a new Sender ID received from the Group Manager (see Section 2.6.3.1).¶
The server MUST store the value of the 'kid context' parameter from the request, and retain it until the last response has been sent, i.e., beyond a possible change of ID Context following a group rekeying (see Section 12.2). That is, upon establishing a new Security Context with a new Gid as ID Context (see Section 2.6.3.2), the server MUST NOT update the stored value of a 'kid context' parameter associated with the request.¶
When using the group mode to protect a response, a server proceeds as described in Section 8.3 of [RFC8613] with the following modifications.¶
Note that the server always protects a response with the Sender Context from its latest Security Context, and that establishing a new Security Context resets the Sender Sequence Number to 0 (see Section 2.6.3.1 and Section 2.6.3.2).¶
In Step 2, the Additional Authenticated Data is modified as described in Section 3 of this document.¶
In addition, the following applies if the server intends to reply with multiple responses within the long exchange established by the corresponding request.¶
The server MUST use the stored value of the 'kid' parameter from the request (see Section 7.2), as value for the 'request_kid' parameter in the external_aad (see Section 3.4).¶
The server MUST use the stored value of the 'kid context' parameter from the request (see Section 7.2), as value for the 'request_kid_context' parameter in the external_aad (see Section 3.4).¶
In Step 3, if either of the following conditions holds, the server MUST include its Sender Sequence Number as Partial IV in the response and use it to build the nonce to protect the response. This prevents the server from reusing the nonce from the request together with the same encryption key.¶
The response is not the first response that the server sends to the request.¶
The server is using a different Security Context for the response than the one that was used to verify the request (see Section 12.2).¶
In Step 4, the encryption of the COSE object is modified as described in Section 3 of this document. The encoding of the compressed COSE object is modified as described in Section 4 of this document. In particular, the Group Flag MUST be set to 1. The Group Encryption Algorithm from the Common Context MUST be used.¶
In addition, the following applies.¶
If the server is using a different ID Context (Gid) for the response than the one that was used to verify the request (see Section 12.2) and this is the first response from the server to that request, then the new ID Context MUST be included in the 'kid context' parameter of the response.¶
The server may be replying to a request that was protected with an old Security Context. After completing the establishment of a new Security Context, the server MUST protect all the responses to that request with the Sender Context of the new Security Context.¶
For each ongoing long exchange, the server can help the client to synchronize, by including also the 'kid context' parameter in responses following a group rekeying, with value set to the ID Context (Gid) of the new Security Context.¶
If there is a known upper limit to the duration of a group rekeying, the server SHOULD include the 'kid context' parameter during that time. Otherwise, the server SHOULD include it until the Max-Age has expired for the last response sent before the installation of the new Security Context.¶
The server can obtain a new Sender ID from the Group Manager when individually rekeyed (see Section 2.6.3.1) or when re-joining the group. In such a case, the server can help the client to synchronize by including the 'kid' parameter in a response protected in group mode even when the request was protected in pairwise mode (see Section 8.3).¶
That is, when responding to a request protected in pairwise mode, the server SHOULD include the 'kid' parameter in a response protected in group mode, if it is replying to that client for the first time since the assignment of its new Sender ID.¶
In Step 5, the countersignature is computed and the format of the OSCORE message is modified as described in Section 3 and Section 4 of this document. In particular the payload of the Group OSCORE message includes also the encrypted countersignature (see Section 4.1).¶
Upon receiving a protected response with the Group Flag set to 1, following the procedure in Section 6, a client proceeds as described in Section 8.4 of [RFC8613] with the modifications described in this section.¶
Note that a client may receive a response protected with a Security Context different from the one used to protect the corresponding request, and that, upon the establishment of a new Security Context, the client re-initializes its Replay Windows in its Recipient Contexts (see Section 12.2).¶
In Step 2, the decoding of the compressed COSE object is modified as described in Section 4 of this document. In particular, a 'kid' may not be present if the response is a reply to a request protected in pairwise mode. In such a case, the client assumes the response 'kid' to be the Recipient ID for the server for which the request protected in pairwise mode was intended.¶
If the response 'kid context' matches an existing ID Context (Gid) but the received/assumed 'kid' does not match any Recipient ID in this Security Context, then the client MAY create a new Recipient Context for this Recipient ID and initialize it according to Section 3 of [RFC8613], and also retrieve the authentication credential associated with the Recipient ID to be stored in the new Recipient Context. If the application does not specify dynamic derivation of new Recipient Contexts, then the client SHALL stop processing the response.¶
In Step 3, the Additional Authenticated Data is modified as described in Section 3 of this document.¶
In addition, the following applies if the client processes a response to a request within a long exchange.¶
The client MUST use the stored value of the 'kid' parameter from the request (see Section 7.1), as value for the 'request_kid' parameter in the external_aad (see Section 3.4).¶
The client MUST use the stored value of the 'kid context' parameter from the request (see Section 7.1), as value for the 'request_kid_context' parameter in the external_aad (see Section 3.4).¶
This ensures that, throughout a long exchange, the client can correctly verify the received responses, even if the client is individually rekeyed and starts using a new Sender ID received from the Group Manager (see Section 2.6.3.1), as well as when it installs a new Security Context with a new ID Context (Gid) following a group rekeying (see Section 12.2).¶
In Step 5, the client also verifies the countersignature by using the public key from the server's authentication credential stored in the associated Recipient Context. In particular:¶
The signature verification as defined below SHOULD be performed before decrypting the COSE object. An exception applies to implementations that cannot perform the two steps in this order. Those implementations MUST ensure that no access to the plaintext is possible before a successful signature verification and MUST prevent any possible leak of time-related information that can yield side-channel attacks.¶
The client retrieves the encrypted countersignature ENC_SIGNATURE from the message payload, and computes the original countersignature SIGNATURE as¶
SIGNATURE = ENC_SIGNATURE XOR KEYSTREAM¶
where KEYSTREAM is derived as per Section 4.2.¶
The client verifies the original countersignature SIGNATURE.¶
If the verification of the countersignature fails, the client: i) SHALL stop processing the response; and ii) SHALL NOT update the Response Number associated with the server.¶
After a successful verification of the countersignature, the client performs also the following actions in case the request was protected in pairwise mode (see Section 8.3).¶
If the 'kid' parameter is present in the response, the client checks whether this received 'kid' is equal to the expected 'kid', i.e., the known Recipient ID for the server for which the request was intended.¶
If the 'kid' parameter is not present in the response, the client checks whether the server that has sent the response is the same one for which the request was intended. This can be done by checking that the public key used to verify the countersignature of the response is equal to the public key included in the authentication credential Recipient Auth Cred, which was taken as input to derive the Pairwise Sender Key used for protecting the request (see Section 2.5.1).¶
In either case, if the client determines that the response has come from a different server than the expected one, then the client: i) SHALL discard the response and SHALL NOT deliver it to the application; ii) SHALL NOT update the Response Number associated with the server.¶
Otherwise, the client hereafter considers the received 'kid' as the current Recipient ID for the server.¶
In Step 5, when decrypting the COSE object using the Recipient Key, the Group Encryption Algorithm from the Common Context MUST be used.¶
In addition, the client performs the following actions if the response is received within a long exchange.¶
The ordering and the replay protection of responses received from the server during the long exchange are performed as per Section 5.3.1 of this document, by using the Response Number associated with that server within that long exchange. In case of unsuccessful decryption and verification of a response, the client SHALL NOT update the Response Number associated with the server.¶
When receiving the first valid response from the server within the long exchange, the client MUST store the kid "kid1" of that server for that long exchange. If the 'kid' field is included in the OSCORE Option of the response, its value specifies "kid1". If the request was protected in pairwise mode (see Section 8.3), the 'kid' field may not be present in the OSCORE Option of the response (see Section 3.2). In this case, the client assumes "kid1" to be the Recipient ID for the server for which the request was intended.¶
When receiving another valid response to the same request from the same server - which can be identified and recognized through the same public key used to verify the countersignature and included in the server's authentication credential - the client determines the kid "kid2" of the server as above for "kid1", and MUST check whether "kid2" is equal to the stored "kid1".¶
If "kid1" and "kid2" are different, the client SHOULD NOT accept the response as valid to be delivered to the application, and SHOULD NOT update the Response Number associated with the server. Exceptions can apply as the client can retain the information required to order the responses, or if the client application does not require response ordering altogether. Servers MUST NOT rely on clients tolerating this, unless it was explicitly agreed on (e.g., as part of the group's setup).¶
Note that, if "kid2" is different from "kid1" and the 'kid' field is omitted from the response - which is possible if the request was protected in pairwise mode - then the client will compute a wrong keystream to decrypt the countersignature (i.e., by using "kid1" rather than "kid2" in the 'id' field of the 'info' array in Section 4.2), thus subsequently failing to verify the countersignature and discarding the response.¶
This ensures that the client remains able to correctly perform the ordering and replay protection of responses within a long exchange, even if the server legitimately starts using a new Sender ID, as received from the Group Manager when individually rekeyed (see Section 2.6.3.1) or when re-joining the group.¶
In Step 8, if the used Recipient Context was created upon receiving this response and the message is not verified successfully, the client MAY delete that Recipient Context. When this behavior is specified by the application, it mitigates attacks that aim at overloading the client's storage.¶
If the client deletes the used Recipient Context in this particular circumstance, then this deletion does not require the client to initialize as invalid the Replay Window of any new Recipient Context created later within the Security Context (see Section 2.6.1.2).¶
When a message is protected in group mode, it is possible for designated external signature checkers to verify the countersignature of the message. For example, an external signature checker can be an intermediary gateway that intercepts messages protected in group mode and ensures that they reach the intended recipients only if it successfully verifies their countersignatures.¶
Since they do not join a group as members, external signature checkers need to retrieve from the Group Manager the authentication credentials of group members and other selected group data, such as the current Signature Encryption Key (see Section 2.1.9). Section 12.3 describes how the Group Manager supports signature checkers.¶
When receiving a message protected in group mode, a signature checker proceeds as follows.¶
The signature checker retrieves the encrypted countersignature ENC_SIGNATURE from the message payload, and computes the original countersignature SIGNATURE as¶
SIGNATURE = ENC_SIGNATURE XOR KEYSTREAM¶
where KEYSTREAM is derived as per Section 4.2.¶
The signature checker verifies the original countersignature SIGNATURE, by using the public key of the sender endpoint as included in that endpoint's authentication credential. The signature checker determines the right authentication credential based on the ID Context (Gid) and the Sender ID of the sender endpoint.¶
Note that the following applies when attempting to verify the countersignature of a response message.¶
The response may not include a Partial IV and/or an ID Context. In such a case, the signature checker considers the same values from the corresponding request, i.e., the request matching with the response by CoAP Token value.¶
The response may not include a Sender ID. This can happen when the response protected in group mode matches a request protected in pairwise mode (see Section 8.1), with a case in point provided by [I-D.ietf-core-cacheable-oscore]. In such a case, the signature checker needs to use other means (e.g., source addressing information of the server endpoint) to identify the correct authentication credential including the public key to use for verifying the countersignature of the response.¶
The particular actions following a successful or unsuccessful verification of the countersignature are application specific and out of the scope of this document.¶
When using the pairwise mode of Group OSCORE, messages are protected and processed as in [RFC8613] with the modifications described in this section. The security objectives of the pairwise mode are discussed in Appendix A.2.¶
The possible use of the pairwise mode is indicated by the Group Manager as part of the group data provided to new group members when joining the group, according to which the parameters AEAD Algorithm and Pairwise Key Agreement Algorithm in the Security Context are set (see Section 2).¶
The pairwise mode takes advantage of an existing Security Context to establish keying material shared exclusively with each other member. For encryption and decryption operations in pairwise mode, the AEAD Algorithm from the Common Context is used (see Section 2.1.1).¶
In order to use the pairwise mode in a group where the group mode is also used (i.e., Group Encryption Algorithm and Signature Algorithm in the Security Context are set), the public/private key pairs used for signature operations of the group mode MUST be possible to also use for key agreement operations. For example, this can rely on signing operations using ECDSA, and encryption operations using AES-CCM with keying material derived through ECDH.¶
The pairwise mode does not support external verifiers of source authentication and message integrity like the group mode does, e.g., for external signature checkers (see Section 7.5).¶
An endpoint implementing only a silent server does not support the pairwise mode.¶
Endpoints using the CoAP Echo Option [RFC9175] in a group where the AEAD Algorithm and Pairwise Key Agreement Algorithm are set MUST support the pairwise mode. When using the challenge-response method defined in Section 9, this prevents the attack described in Section 14.9, which leverages requests sent over unicast to a single group member and protected in group mode.¶
The pairwise mode cannot be used to protect messages intended for multiple recipients, as the keying material used for the pairwise mode is shared only between two endpoints.¶
However, a sender can use the pairwise mode to protect a message sent to (but not intended for) multiple recipients, if interested in a response from only one of them. For instance, this is useful to support the address discovery service defined in Section 8.1, when a single 'kid' value is indicated in the payload of a request sent to multiple recipients, e.g., over multicast.¶
In order to protect an outgoing message in pairwise mode, the sender needs to know the authentication credential and the Recipient ID for the recipient endpoint, as stored in the Recipient Context associated with that endpoint (see Section 2.5.4).¶
Typically, the sender endpoint sends the message protected in pairwise mode over unicast, so that the message is delivered only to the intended recipient endpoint for which it is protected. This requires the sender to know the individual address of that recipient endpoint, which the sender may not know at any given point in time. For instance, right after having joined the group, a client may know the authentication credential and Recipient ID for a given server, but not the addressing information required to reach it with an individual, one-to-one request.¶
In order to make addressing information of individual endpoints available, servers in the group MAY expose a resource to which a client can send a request targeting a set of servers, identified by their 'kid' values specified in the request payload, or implicitly if the request is sent in pairwise mode. Further details of such an interface are out of scope for this document.¶
The pairwise mode protects messages between two members of a group, essentially following [RFC8613] but with the following notable differences.¶
The 'kid' and 'kid context' parameters of the COSE object are used as defined in Section 3.2 of this document.¶
The external_aad defined in Section 3.4 of this document is used for the encryption and decryption process.¶
The Pairwise Sender/Recipient Keys used as Sender/Recipient keys are derived as defined in Section 2.5 of this document.¶
When using the pairwise mode to protect a request, a client SHALL proceed as described in Section 8.1 of [RFC8613] with the differences summarized in Section 8.2 of this document.¶
Furthermore, when sending a request that establishes a long exchange, what is specified in Section 7.1 of this document holds, with respect to storing the value of the 'kid' and 'kid context' parameters, and to storing an invariant identifier of the group.¶
Upon receiving a protected request with the Group Flag set to 0, following the procedure in Section 6, a server SHALL proceed as described in Section 8.2 of [RFC8613] with the differences summarized in Section 8.2 of this document. The following differences also apply.¶
If the server discards the request due to not retrieving a Security Context associated with the OSCORE group or to not supporting the pairwise mode, the server MAY respond with a 4.01 (Unauthorized) error message or a 4.02 (Bad Option) error message, respectively. When doing so, the server MAY set an Outer Max-Age Option with value zero, and MAY include a descriptive string as diagnostic payload.¶
If a new Recipient Context is created for this Recipient ID, new Pairwise Sender/Recipient Keys are also derived (see Section 2.5.1). The new Pairwise Sender/Recipient Keys are deleted if the Recipient Context is deleted as a result of the message not being successfully verified.¶
What is specified in Section 7.2 of this document holds with respect to the following points.¶
The possible, dynamic creation and configuration of a Recipient Context upon receiving the request.¶
The possible deletion of a Recipient Context created upon receiving the request, in case the request is not verified successfully.¶
The rule about processing the request where the received Recipient ID ('kid') is equal to the server's Sender ID.¶
The storing of the value of the 'kid' and 'kid context' parameters from the request, if the server intends to reply with multiple responses within the long exchange established by the request.¶
When using the pairwise mode to protect a response, a server SHALL proceed as described in Section 8.3 of [RFC8613] with the differences summarized in Section 8.2 of this document. The following differences also apply.¶
What is specified in Section 7.3 of this document holds with respect to the following points.¶
The protection of a response when using a different Security Context than the one used to verify the corresponding request (see Section 12.2). That is, the server always protects a response with the Sender Context from its latest Security Context, and establishing a new Security Context resets the Sender Sequence Number to 0 (see Section 2.6.3.1 and Section 2.6.3.2).¶
The use of the stored value of the 'kid' and 'kid context' parameters, if the server intends to reply with multiple responses within the long exchange established by the request.¶
The rules for the inclusion of the server's Sender Sequence Number as Partial IV in a response, as used to build the nonce to protect the response.¶
The rules for the inclusion of the ID Context (Gid) in the 'kid context' parameter of a response, if the ID Context used for the response differs from the one used to verify the request (see Section 12.2), also for helping the client to synchronize.¶
The rules for the inclusion of the Sender ID in the 'kid' parameter of a response to a request that was protected in pairwise mode, if the server has obtained a new Sender ID from the Group Manager when individually rekeyed (see Section 2.6.3.1), thus helping the client to synchronize.¶
Upon receiving a protected response with the Group Flag set to 0, following the procedure in Section 6, a client SHALL proceed as described in Section 8.4 of [RFC8613] with the differences summarized in Section 8.2 of this document. The following differences also apply.¶
The client may receive a response protected with a Security Context different from the one used to protect the corresponding request. Also, upon the establishment of a new Security Context, the client re-initializes its Replay Windows in its Recipient Contexts (see Section 2.2).¶
The same as described in Section 7.4 holds with respect to handling the 'kid' parameter of the response, when received as a reply to a request protected in pairwise mode. The client can also in this case check whether the replying server is the expected one, by relying on the server's public key. However, since the response is protected in pairwise mode, the public key is not used for verifying a countersignature as in Section 7.4. Instead, the expected server's authentication credential - namely Recipient Auth Cred and including the server's public key - was taken as input to derive the Pairwise Recipient Key used to decrypt and verify the response (see Section 2.5.1).¶
If a new Recipient Context is created for this Recipient ID, new Pairwise Sender/Recipient Keys are also derived (see Section 2.5.1). The new Pairwise Sender/Recipient Keys are deleted if the Recipient Context is deleted as a result of the message not being successfully verified.¶
What is specified in Section 7.4 of this document holds with respect to the following points.¶
The possible, dynamic creation and configuration of a Recipient Context upon receiving the response.¶
The use of the stored value of the 'kid' and 'kid context' parameters, when processing a response received within a long exchange.¶
The performing of ordering and replay protection for responses received within a long exchange.¶
The possible deletion of a Recipient Context created upon receiving the response, in case the response is not verified successfully.¶
This section describes how a server endpoint can verify freshness of a request by means of a challenge-response exchange with a client using the Echo Option for CoAP specified in Section 2 of [RFC9175]. The same mechanism, with small alterations, is also used by the server when first processing a request using a Recipient Context whose Replay Window was initialized as invalid.¶
If the application requires freshness, e.g., according to time- or event-based policies (see Section 2.5.1 of [RFC9175]), a server proceeds as described below upon receiving a request from a particular client for the first time.¶
The server processes the message as described in this document, but, even if valid, does not deliver it to the application. Instead, the server replies to the client with a Group OSCORE protected 4.01 (Unauthorized) response message, including only the Echo Option and no diagnostic payload. The server MUST use its Partial IV when generating the nonce for protecting the response conveying the Echo Option, and MUST include the Partial IV in the response.¶
The Echo Option value SHOULD NOT be reused; if it is reused, it MUST be highly unlikely to have been recently used with this client. Since this response is protected with the Security Context used in the group, the client will consider the response valid upon successfully decrypting and verifying it.¶
The server stores the Echo Option value included in the response together with the pair (gid,kid), where 'gid' is the Group Identifier of the OSCORE group and 'kid' is the Sender ID of the client in the group. These are specified in the 'kid context' and 'kid' fields of the OSCORE Option of the request, respectively. After a group rekeying has been completed and a new Security Context has been established in the group, which results also in a new Group Identifier (see Section 12.2), the server MUST delete all the stored Echo values associated with members of the group.¶
After receiving a 4.01 (Unauthorized) response that includes an Echo Option and originates from a verified group member, a subsequent client request sent to that server and echoing the Echo Option value MUST be a message sent unicast to that server.¶
If in the group the AEAD Algorithm and Pairwise Key Agreement Algorithm are set in the Security Context, the client MUST use the pairwise mode to protect the request, as per Section 8.3. Note that, as defined in Section 8, endpoints that are members of such a group and that use the Echo Option support the pairwise mode. In a group where the AEAD Algorithm and Pairwise Key Agreement Algorithm are not set, only the group mode can be used. Hence, requests including the Echo Option can be protected only with the Group Mode, with the caveat due to the risk for those requests to be redirected to a different server than the intended one, as discussed in Section 14.9.¶
The client does not necessarily resend the same request, but can instead send a more recent one if the application permits it. This allows the client to not retain previously sent requests for full retransmission, unless the application explicitly requires otherwise. In either case, the client uses a fresh Sender Sequence Number value from its own Sender Context. If the client stores requests for possible retransmission with the Echo Option, it should not store a given request for longer than a preconfigured time interval. Note that the unicast request echoing the Echo Option is correctly treated and processed, since the 'kid context' field including the Group Identifier of the OSCORE group is still present in the OSCORE Option as part of the COSE object (see Section 3).¶
Upon receiving the unicast request including the Echo Option, the server performs the following verifications.¶
If the server does not store an Echo Option value for the pair (gid,kid), it considers: i) the time t1 when it has established the Security Context used to protect the received request; and ii) the time t2 when the request has been received. Since a valid request cannot be older than the Security Context used to protect it, the server verifies that (t2 - t1) is less than the largest amount of time acceptable to consider the request fresh.¶
If the server stores an Echo Option value for the pair (gid,kid) associated with that same client in the same group, the server verifies that the option value equals that same stored value previously sent to that client.¶
If the verifications above fail, the server MUST NOT process the request further and MAY send a 4.01 (Unauthorized) response including an Echo Option, hence performing a new challenge-response exchange.¶
If the verifications above are successful, the server considers the Recipient Context associated with the sender client and proceeds as follows.¶
If the Replay Window is invalid, the steps below occur.¶
The server updates the Replay Window by marking as received the Sender Sequence Number from the latest received request. This becomes the lower limit of the Replay Window, while all the greater Sender Sequence Number values within the Replay Window are marked as not received.¶
The server makes the Replay Window valid, and accepts the request as fresh.¶
If the Replay Window is already valid, the server discards the verification result and accepts the request as fresh or treats it as a replay, according to the existing Replay Window.¶
A server should not deliver requests from a given client to the application until one valid request from that same client has been verified as fresh via an echoed Echo Option included therein. A server may perform the challenge-response described above at any time, e.g., after a device reboot occurred in an unprepared way. A client has to be ready to perform the challenge-response based on the Echo Option if a server starts it.¶
Message freshness is further discussed in Section 14.14.¶
Like in [RFC8613], HKDF SHA-256 is the mandatory-to-implement HKDF.¶
An endpoint may support only the group mode, or only the pairwise mode, or both.¶
For endpoints that support the group mode, the following applies.¶
For endpoints that use authenticated encryption, the AEAD algorithm AES-CCM-16-64-128 defined in Section 4.2 of [RFC9053] is mandatory to implement as Group Encryption Algorithm (see Section 2.1.7).¶
For endpoints that use non-authenticated encryption, the algorithm A128CTR defined in Section 4 of [RFC9459] is mandatory to implement as Group Encryption Algorithm (see Section 2.1.7).¶
Section 6 of [RFC9459] mandates that COSE libraries supporting the AES-CTR algorithm and accepting Additional Authenticated Data (AAD) as input must return an error if AAD is provided when such a non-AEAD content encryption algorithm is selected.¶
In case the used Group Encryption Algorithm (see Section 2.1.7) does not provide integrity protection, the following applies.¶
When invoking the execution of the Group Encryption Algorithm, the Group OSCORE implementation MUST NOT provide any AAD to the COSE library, unless AAD is always expected as input. In the latter case, the AAD will not be protected by the Group Encryption Algorithm, which is unable to do so.¶
If the used COSE library adheres to the mandate in Section 6 of [RFC9459], then a Group OSCORE implementation requires that the COSE library supports using the Group Encryption Algorithm without taking AAD as input.¶
For many constrained IoT devices, it is problematic to support more than one signature algorithm. The following applies with respect to the Signature Algorithm (see Section 2.1.8).¶
Less constrained endpoints MUST implement at least one of the following and SHOULD implement both: the EdDSA signature algorithm together with the elliptic curve Ed25519 [RFC8032]; the ECDSA signature algorithm together with the elliptic curve P-256.¶
Constrained endpoints MUST implement at least one of the following and, if affordable, SHOULD implement both: the EdDSA signature algorithm together with the elliptic curve Ed25519 [RFC8032]; the ECDSA signature algorithm together with the elliptic curve P-256.¶
Endpoints that implement the ECDSA signature algorithm MAY use "deterministic ECDSA" as specified in [RFC6979]. Pure deterministic elliptic-curve signature algorithms such as deterministic ECDSA and EdDSA have the advantage of not requiring access to a source of high-quality randomness. However, these signature algorithms have been shown vulnerable to some side-channel and fault injection attacks due to their determinism, which can result in extracting a device's private key. As suggested in Section 2.1.1 of [RFC9053], this can be addressed by combining both randomness and determinism [I-D.irtf-cfrg-det-sigs-with-noise].¶
For endpoints that support the pairwise mode, the following applies.¶
The AEAD algorithm AES-CCM-16-64-128 defined in Section 4.2 of [RFC9053] is mandatory to implement as AEAD Algorithm (see Section 2.1.1).¶
The ECDH-SS + HKDF-256 algorithm specified in Section 6.3.1 of [RFC9053] is mandatory to implement as Pairwise Key Agreement Algorithm (see Section 2.1.10).¶
The following applies with respect to ECDH curves.¶
Less constrained endpoints MUST implement at least one of the following ECDH curves and SHOULD implement both: the X25519 curve [RFC7748]; the P-256 curve.¶
Constrained endpoints MUST implement at least one of the following ECDH curves and, if affordable, SHOULD implement both: the X25519 curve [RFC7748]; the P-256 curve.¶
Constrained IoT devices may alternatively represent Montgomery curves and (twisted) Edwards curves [RFC7748] in the short-Weierstrass form Wei25519, with which the algorithms ECDSA25519 and ECDH25519 can be used for signature operations and Diffie-Hellman secret calculation, respectively [I-D.ietf-lwig-curve-representations].¶
The use of Group OSCORE or OSCORE [RFC8613] MAY be indicated by a target "gosc" attribute in a web link [RFC8288] to a resource, e.g., using a link-format document [RFC6690] if the resource is accessible over CoAP.¶
The "gosc" attribute is a hint indicating that the destination of that link is only accessible using Group OSCORE or OSCORE, and unprotected access to it is not supported. Note that this is simply a hint, it does not include any security context material or any other information required to run Group OSCORE or OSCORE.¶
A value MUST NOT be given for the "gosc" attribute; any present value MUST be ignored by parsers. The "gosc" attribute MUST NOT appear more than once in a given link-value; occurrences after the first MUST be ignored by parsers.¶
When a link-value includes the "gosc" attribute, the link-value MUST also include the "osc" attribute defined in Section 9 of [RFC8613]. If the endpoint parsing the link-value supports Group OSCORE and understands the "gosc" attribute, then the parser MUST ignore the "osc" attribute, which is overshadowed by the "gosc" attribute.¶
The example in Figure 3 shows a use of the "gosc" attribute: the client does resource discovery on a server and gets back a list of resources, one of which includes the "gosc" attribute indicating that the resource is protected with Group OSCORE or OSCORE. The link-format notation (see Section 5 of [RFC6690]) is used.¶
REQ: GET /.well-known/core
RES: 2.05 Content
</sensors/temp>;gosc;osc,
</sensors/light>;if="sensor"
As with OSCORE, endpoints communicating with Group OSCORE need to establish the relevant Security Context. Group OSCORE endpoints need to acquire OSCORE input parameters, information about the group(s) and about other endpoints in the group(s).¶
Every group is associated with a Group Manager that is responsible for distributing security parameters and keying material within the group, among other tasks. The details of how the Group Manager interacts with (candidate) group members or with external entities like signature checkers, as well as the protocols used for those interactions, are out of scope.¶
The Group Manager assigns unique Group Identifiers (Gids) to the groups under its control. Within each of such groups, the Group Manager assigns unique Sender IDs (and thus Recipient IDs) to the respective group members. The maximum length of Sender IDs depends on the length of the nonce for the algorithms used in the group (see Section 2.2).¶
The Gid value assigned to a group is associated with a dedicated space for the values of Sender ID and Recipient ID of the members of that group. When an endpoint (re-)joins a group, it is provided with the current Gid to use in the group. The Group Manager also assigns an integer Key Generation Number counter to each of its groups, identifying the current version of the keying material used in that group. Further details about identifiers and keys are provided in Section 12.2.¶
The Group Manager maintains records of the authentication credentials of endpoints in a group, and provides information about the group and its members to other group members (see Section 12.1). Optionally, the Group Manager provides information about the group and its members to external entities with a specific role, such as signature checkers (see Section 12.3).¶
The list of responsibilities of the Group Manager is compiled in Appendix D.¶
One realization of a Group Manager is specified in [I-D.ietf-ace-key-groupcomm-oscore], where the process by which an endpoint (re-)joins a group is based on the ACE framework for authentication and authorization in constrained environments [RFC9200].¶
From the Group Manager, an endpoint acquires group data such as the Gid and OSCORE input parameters including its own Sender ID, with which it can derive the Sender Context.¶
When joining the group or later on as a group member, an endpoint can also retrieve from the Group Manager the authentication credential of the Group Manager as well as the authentication credential and other information associated with other members of the group, with which it can derive the corresponding Recipient Context. An application can configure a group member to asynchronously retrieve information about Recipient Contexts, e.g., by Observing [RFC7641] a resource at the Group Manager to get updates on the group membership.¶
Upon endpoints' joining, the Group Manager collects their authentication credentials and MUST verify proof of possession of the respective private key. As an example, such proof of possession is possible to achieve during the join process provided by the realization of Group Manager specified in [I-D.ietf-ace-key-groupcomm-oscore]. Together with the requested authentication credentials of other group members, the Group Manager MUST provide the joining endpoints with the Sender ID of the associated group members and the current Key Generation Number in the group (see Section 12.2).¶
An endpoint may join a group, for example, by explicitly interacting with the responsible Group Manager, or by being configured with some tool performing the tasks of the Group Manager. When becoming members of a group, endpoints are not required to know how many and what endpoints are in the same group.¶
Communications that the Group Manager has with joining endpoints and group members MUST be secured. Specific details on how to secure such communications are out of the scope of this document.¶
The Group Manager MUST verify that the joining endpoint is authorized to join the group. To this end, the Group Manager can directly authorize the joining endpoint, or expect it to provide authorization evidence previously obtained from a trusted entity. Further details about the authorization of joining endpoints are out of the scope of this document.¶
In case of successful authorization check, the Group Manager provides the joining endpoint with the keying material to initialize the Security Context. The actual provisioning of keying material and parameters to the joining endpoint is out of the scope of this document.¶
In order to establish a new Security Context for a group, the Group Manager MUST generate and assign to the group a new Group Identifier (Gid) and a new value for the Master Secret parameter. When doing so, a new value for the Master Salt parameter MAY also be generated and assigned to the group. When establishing the new Security Context, the Group Manager SHOULD preserve the current value of the Sender ID of each group member in order to ensure an efficient key rollover. Exceptions can apply if there are compelling reasons for making available again some of the Sender ID values currently used.¶
The specific group key management scheme used to distribute new keying material is out of the scope of this document. A simple group key management scheme is defined in [I-D.ietf-ace-key-groupcomm-oscore]. Different group key management schemes rely on different approaches to compose and deliver rekeying messages, i.e., individually targeting single recipients, or targeting multiple recipients at once (e.g., over UDP/IP multicast), or a combination of the two approaches. As long as it is viable for the specific rekeying message to be delivered and it is supported by the intended message recipient(s), using a reliable transport to deliver a rekeying message should be preferred, as it reduces chances of group members missing a rekeying instance.¶
Irrespective of the transport used being reliable or unreliable, appropriate congestion control MUST be enforced. If the key distribution traffic uses CoAP over UDP or over other unreliable transports, mechanisms for enforcing congestion control are specified in Section 4.7 of [RFC7252] and in Section 3.6 of [I-D.ietf-core-groupcomm-bis] for the case of group communication (e.g., over UDP/IP multicast). If, irrespective of using CoAP, the key distribution traffic relies on alternative setups with unreliable transports, one can rely on general congestion-control mechanisms such as DCCP [RFC4340], or on dedicated congestion control mechanisms for the transport specifically used (e.g., those defined in [RFC9002] for QUIC [RFC9000]).¶
The set of group members should not be assumed as fixed, i.e., the group membership is subject to changes, possibly on a frequent basis.¶
The Group Manager MUST rekey the group without undue delay when one or more endpoints leave the group. An endpoint may leave the group at own initiative, or may be evicted from the group by the Group Manager, e.g., in case the endpoint is compromised, or is suspected to be compromised (as determined by the Group Manager through its own means or based on information that it obtains from a trusted source such as an Intrusion Detection System or an issuer of authentication credentials). In either case, rekeying the group excludes such endpoints from future communications in the group, and thus preserves forward security. If a network node is compromised or suspected to be compromised, the Group Manager MUST evict from the group all the endpoints hosted by that node that are members of the group and rekey the group accordingly.¶
If required by the application, the Group Manager MUST also rekey the group when one or more new joining endpoints are added to the group, thus preserving backward security.¶
The Group Manager MAY also rekey the group for other reasons, e.g., according to an application-specific rekeying period or scheduling.¶
Separately for each group, the value of the Key Generation Number increases by one each time the Group Manager distributes new keying material to that group (see below).¶
The establishment of the new Security Context for the group takes the following steps.¶
The Group Manager MUST increment the Key Generation Number for the group by 1. It is up to the Group Manager what actions to take when a wrap-around of the Key Generation Number is detected.¶
The Group Manager MUST build a set of stale Sender IDs including:¶
The Sender IDs that, during the current Gid, were both assigned to an endpoint and subsequently relinquished (see Section 2.6.3.1).¶
The current Sender IDs of the group members that the upcoming group rekeying aims to exclude from future group communications, if any.¶
The Group Manager rekeys the group, by distributing:¶
The new keying material, i.e., the new Master Secret, the new Gid and (optionally) the new Master Salt.¶
The new Key Generation Number from Step 1.¶
The set of stale Sender IDs from Step 2.¶
Further information may be distributed, depending on the specific group key management scheme used in the group.¶
When receiving the new group keying material, a group member considers the received stale Sender IDs and performs the following actions.¶
The group member MUST remove every authentication credential associated with a stale Sender ID from its list of group members' authentication credentials used in the group.¶
The group member MUST delete each of its Recipient Contexts used in the group whose corresponding Recipient ID is a stale Sen