International Association for Cryptologic Research

International Association
for Cryptologic Research


Aldar C-F. Chan


Conditionally Verifiable Signatures
Aldar C-F. Chan Ian F. Blake
We introduce a new digital signature model, called conditionally verifiable signature (CVS), which allows a signer to specify and convince a recipient under what conditions his signature would become valid and verifiable; the resulting signature is not publicly verifiable immediately but can be converted back into an ordinary one (verifiable by anyone) after the recipient has obtained proofs, in the form of signatures/endorsements from a number of third party witnesses, that all the specified conditions have been fulfilled. A fairly wide set of conditions could be specified in CVS. The only job of the witnesses is to certify the fulfillment of a condition and none of them need to be actively involved in the actual signature conversion, thus protecting user privacy. It is guaranteed that the recipient cannot cheat as long as at least one of the specified witnesses does not collude. We formalize the concept of CVS and give a generic CVS construction based on any CPA-secure identity based encryption (IBE) scheme. Theoretically, we show that the existence of IBE with indistinguishability under a chosen plaintext attack (a weaker notion than the standard one) is necessary and sufficient for the construction of a secure CVS.\footnote{Due to page limit, some proofs are omitted here but could be found in the full version \cite{CB05ibecvs}.}
Scalable, Server-Passive, User-Anonymous Timed Release Public Key Encryption from Bilinear Pairing
Ian F. Blake Aldar C-F. Chan
We consider the problem of sending messages into the future, commonly known as timed release cryptography. Existing schemes for this task either solve the relative time problem with uncontrollable, coarse-grained release time (time-lock puzzle approach) or do not provide anonymity to sender and/or receiver and are not scalable (server-based approach). Using a bilinear paring on any Gap Diffie-Hellman group, we solve this problem by giving a scalable, server-passive and user-anonymous timed release public-key encryption scheme which allows precise absolute release time specifications. Unlike the existing server-based schemes, the trusted time server in our scheme is completely passive --- no interaction between it and the sender or receiver is needed; it is even not aware of the existence of a user, thus assuring the anonymity of both the sender and receiver of a message and the privacy of the message. Besides, our scheme also has a number of desirable properties including self-authenticated time-bound key updates, a single form of update for all receivers, simple public-key renewal and key insulation, making it a scalable and appealing solution.


Ian F. Blake (2)