A hybrid approach to vector-based homomorphic tallying remote voting

Abstract

Vector-based homomorphic tallying remote voting schemes provide an efficient protocol for vote tallying, but they require voters to prove in zero-knowledge that the ballots they cast have been properly generated. This is usually achieved by means of the so-called zero-knowledge range proofs, which should be verified by the polling station before tallying. In this paper, we present an end-to-end verifiable hybrid proposal in which ballots are proven to be correct by making use of a zero-knowledge proof of mixing but still using a homomorphic tallying for gathering the election results. Our proposal offers all the advantages of the homomorphic tallying paradigm, while it avoids the elevated computational cost of range proofs. As a result, ballot verification performance is improved in comparison with the equivalent homomorphic systems. The proposed voting scheme is suitable for multi-candidate elections as well as for elections in which the votes have different weights.