Non-interactive Designated Verifier Proofs and Undeniable Signatures

Non-interactive designated verifier (NIDV) proofs were first introduced by Jakobsson et al. and have widely been used as confirmation and denial proofs for undeniable signature schemes. There appears to be no formal security modelling for NIDV undeniable signatures or for NIDV proofs in general. Indeed, recent work by Wang has shown the original NIDV undeniable signature scheme of Jakobsson et al. to be flawed. We argue that NIDV proofs may have applications outside of the context of undeniable signatures and are therefore of independent interest. We therefore present two security models, one for general NIDV proof systems, and one specifically for NIDV undeniable signatures. We go on to repair the NIDV proofs of Jakobsson et al., producing secure NIDV proofs suited to combination with Chaum's original undeniable signature scheme resulting in a secure and efficient concrete NIDV undeniable signature scheme.