CCF Profile for COSE Receipts

1.1. Requirements Notation

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.¶

2.1. Merkle Tree Shape

A CCF ledger is a binary Merkle Tree constructed from a hash function H, which is defined from the log type. For instance, the hash function for CCF_LEDGER_SHA256 is SHA256, whose HASH_SIZE is 32 bytes.¶

The Merkle tree encodes an ordered list of n transactions T_n = {T[0], T[1], ..., T[n-1]}. We define the Merkle Tree Hash (MTH) function, which takes as input a list of serialized transactions (as byte strings), and outputs a single HASH_SIZE byte string called the Merkle root hash, by induction on the list.¶

This function is defined as follows:¶

The hash of an empty list is the hash of an empty string:¶

MTH({}) = HASH().

The hash of a list with one entry (also known as a leaf hash) is:¶

MTH({d[0]}) = HASH(d[0]).

For n > 1, let k be the largest power of two smaller than n (i.e., k < n <= 2k). The Merkle Tree Hash of an n-element list D_n is then defined recursively as:¶

MTH(D_n) = HASH(MTH(D[0:k]) || MTH(D[k:n])),

where:¶

• || denotes concatenation¶

• : denotes concatenation of lists¶

• D[k1:k2] = D'_(k2-k1) denotes the list {d'[0] = d[k1], d'[1] = d[k1+1], ..., d'[k2-k1-1] = d[k2-1]} of length (k2 - k1).¶

2.2. Transaction Components

Each leaf in a CCF ledger carries the following components:¶

ccf-leaf = [
  ; Byte string of size HASH_SIZE(32)
  internal-transaction-hash: bstr .size 32

  ; Text string of at most 1024 bytes
  internal-evidence: tstr .size (1..1024)

  ; Byte string of size HASH_SIZE(32)
  data-hash: bstr .size 32
]

The internal-transaction-hash and internal-evidence byte strings are internal to the CCF implementation. They can be safely ignored by receipt Verifiers, but they commit the transparency service (TS) to the whole tree contents and may be used for additional, CCF-specific auditing.¶

internal-transaction-hash is a hash over the complete entry in the [CCF-Ledger-Format], and internal-evidence is a revealable [CCF-Commit-Evidence] value that allows early persistence of ledger entries before distributed consensus can be established. This mechanism is useful to implement high-throughput transparency applications in Trusted Execution Environments (TEEs) that only provide a limited amount of memory, while maintaining high availability afforded by distributed consensus.¶

data-hash summarizes the application data included in the ledger at this transaction, which is a Signed Statement as defined by [I-D.ietf-wg-scitt-architecture].¶

3.1. CCF Inclusion Proof Signature

The proof signature for a CCF inclusion proof is a COSE signature (encoded with the COSE_Sign1 CBOR type) which includes the following additional requirements for protected and unprotected headers. Please note that there may be additional header parameters defined by the application.¶

The protected header parameters for the CCF inclusion proof signature MUST include the following:¶

• verifiable-data-structure: int/tstr. This header MUST be set to the verifiable data structure algorithm identifier for ccf-ledger (TBD_1).¶

• label: int. This header MUST be set to the value of the inclusion proof type in the IANA registry of Verifiable Data Structure Proof Type (-1).¶

The unprotected header for a CCF inclusion proof signature MUST include the following:¶

• inclusion-proof: bstr .cbor ccf-inclusion-proof. This contains the serialized CCF inclusion proof, as defined above.¶

The payload of the signature is the CCF ledger Merkle root digest, and MUST be detached in order to force verifiers to recompute the root from the inclusion proof in the unprotected header. This provides a safeguard against implementation errors that use the payload of the signature but do not recompute the root from the inclusion proof.¶

3.2. Inclusion Proof Verification Algorithm

CCF uses the following algorithm to verify an inclusion receipt:¶

compute_root(proof):
  h := proof.leaf.internal-transaction-hash
       || HASH(proof.leaf.internal-evidence)
       || proof.leaf.data-hash

  for [left, hash] in proof:
      h := HASH(hash + h) if left
           HASH(h + hash) else
  return h

verify_inclusion_receipt(inclusion_receipt):
  let label = INCLUSION_PROOF_LABEL
  assert(label in inclusion_receipt.unprotected_header)
  let proof = inclusion_receipt.unprotected_header[label]
  assert(inclusion_receipt.payload == nil)
  let payload = compute_root(proof)

  # Use the Merkle Root as the detached payload
  return verify_cose(inclusion_receipt, payload)

A description can also be found at [CCF-Receipt-Verification].¶

6.1. Trusted Execution Environments

CCF networks of nodes rely on executing in TEEs to secure their function, in particular:¶

1. The evaluation of registration policies¶

2. The creation and usage of receipt signing keys¶

A compromise in the TEE platform used to execute the network may allow an attacker to produce invalid and divergent ledger branches. Clients can mitigate this risk in two ways: by regularly auditing the consistency of the CCF ledger; and by regularly fetching attestation information about the TEE instances, available in the ledger and from the network itself, and confirming that the nodes composing the network are running up-to-date, trusted platform components.¶

6.2. Operators

An operator has the ability to start successor networks with a distinct identity. The operator of a CCF network can recover the service by starting a successor network, for example a new CCF network with its own service identity, that endorses the ledger state of the previous instance. This provides service continuity after a catastrophic failure of a majority of the nodes. However, a malicious operator could exploit this mechanism and truncate the ledger’s history by initializing the successor network from an earlier ledger prefix, thereby omitting some later entries. Clients can mitigate this risk by auditing the successor ledger and verifying that their latest known receipts from the prior service are included in the successor’s ledger.¶

7.1. Additions to Existing Registries