On remote and voter-verifiable voting

Voting is an important tool for modern societies. It is fundamental for the democracy. This mechanism must be fair and accurate. However, threats intrinsic in the voting process may compromise these requisites. In conventional voting, dishonest talliers may corrupt the ballot box, for instance. Remote voting has more threats due to the uncontrolled environment. In this work, we propose secure protocols for polling station and remote voting with attractive properties. Voter-verifiability is a novel security feature provided by recent voting systems. It allows voters to confirm that their votes are accurately counted in the tally. These systems provide better security guarantees than the conventional ones as voters are able to detect corruptions of votes and counting errors. They usually rely on cryptography to implement this feature. We introduce a novel concept of ballot box based on the Farnel approach. It provides voters, when they cast their votes, with one or more random selected receipts. This allows a different notion of voter-verifiability in which the voter may verify one or more random votes. The idea has a number of attractive features: ballot secrecy is achieved up front and does not have to be provided by anonymising mixes during tabulation. Also, it mitigates randomization attacks that are inherent in some voter-verifiable schemes. Based on the new ballot box, we introduce three new voter-verifiable paperbased schemes. One of these schemes does not employ cryptography and thus can be more easily understood by the average voter. The others require just a single cryptographic primitive and achieve better security guarantees. In addition to these schemes, we propose improvements to the Threeballot and RandellRyan voting protocols. These schemes are voter-verifiable and do not employ cryptography. However, they ensure less guarantees than cryptographic based systems. The voter-verifiable schemes given consider a controlled environment as conventional voting. Remote Internet voting, however, has many benefits over the conventional ones and is challenging. Although the existing technology render possible such voting, it has a number of associated threats as coercion and vote selling. We also propose a new scheme for remote Internet voting. The proposal mitigates coercive attacks and is more efficient for large scale voting than the previous solutions.