International Association
for Cryptologic Research

Key Blinding Signature Schemes allow to derive so-called blinded keys from public keys, which can be used to verify signatures created with the secret key. At the same time, neither the blinded keys nor their signatures disclose from which public key they were derived, effectively implementing pseudonyms for one’s identity.

In search of conservative schemes, we deviate from the homomorphism- based re-randomization approach in favor of a novel proof of knowledge- based approach. To authenticate a message, a signer proves that they know an original keypair and a valid way to commit to the corresponding verification key to derive a given blinded key. We provide a framework for such constructions and indicate how MPC-friendly block ciphers and one-way functions may be used for efficient instantiations. While the general framework’s security arguments are stated in the random oracle model, we show a natural instantiation approach whose security can be based on collision-resistance and pseudorandomness instead. The result is the first standard model construction of key blinding.

Using our framework, we identify a shortcoming in the usual definition of unlinkability for key blinding signature schemes, which we rectify by considering an additional notion called targeted unlinkability.