Automatic verification of multi-agent systems security properties specified with LTL

Kholud Alghamdi, Marius Silaghi
2022 Proceedings of the ... International Florida Artificial Intelligence Research Society Conference  
We propose a way to verify security requirements of critical multi-agent system processes by using logic representationsand automatic reasoning. The typical multi-agent system considered in our work would be an election system with agentsrepresenting their users and aiming to ensure security. Relevant processes are authentication, voting, re-voting, and election verification. The security requirements commonly addressed in such a voting system are: no user can vote unless it got authenticated,
more » ... o invalid vote should be counted, no vote should be counted twice, and each valid vote should be eventually counted. We show a model of such security requirements by using system liveness properties, and exemplify their verification on a real system that we implement for this purpose.
doi:10.32473/flairs.v35i.130551 fatcat:vphuii6oxvhktkmj7nnv6yehsu