Towards Practical and Secure Coercion-Resistant Electronic Elections

摘要

Coercion-resistance is the most effective property to fight coercive attacks in Internet elections. This notion was introduced by Juels, Catalano, and Jakobsson (JCJ) at WPES 2005 together with a voting protocol that satisfies such a stringent security requirement. Unfortunately, their scheme has a quadratic complexity (the overhead for tallying authorities is quadratic in the number of votes) and would therefore not be suitable for large scale elections. Based on the work of JCJ, Schweisgut proposed a more efficient scheme. In this paper, we first show that Schweisgut's scheme is insecure. In particular, we describe an attack that allows a coercer to check whether a voter followed or not his instructions. We then present a new coercion-resistant election scheme with a linear complexity that overcomes the drawbacks of these previous proposals. Our solution relies on special anonymous credentials and is proven secure, in the random oracle model, under the q-Strong Diffie-Hellman and Strong Decisional Diffie-Hellman Inversion assumptions.