SaTC: CORE: Small: A Robust Framework with Rigorous Semantics and Security Guarantees for Election-Day Voter Check-in

The broad concerns surrounding the integrity and security of electronic election systems used in the 2016 and 2020 Presidential Elections underscored the need for a rigorous scientific approach to designing and implementing such systems. While substantial research has been dedicated to electronic processing of ballots, vote aggregation, and audits, the very first step that enables voters to cast their votes on Election Day and that ensures the 'one voter, one vote' imperative has not been the target of sufficient research. Moreover, in 2018 the Committee on the Future of Voting of the National Academies of Science, Engineering, and Medicine expressed serious concerns and risks associated with the use of electronic pollbooks. This project focuses on the problem of voter check-in in elections, an area that received surprisingly little attention from the community of researchers working on electronic election systems. The project's novelties are the development of a rigorous scientific foundation for electronic check-in systems and the construction on that basis of a human-centric reference implementation that reflects the legal requirements for such systems. The project's broader significance and importance are potentially substantial savings to the taxpayer by developing and making available the technology ready to be used in constructing secure and trustworthy electronic poll book systems.

This project proceeds along three dimensions: (a) socio-political, (b) software foundations, and (c) system implementation and evaluation. The project advances the state of the art in addressing the electronic check-in challenges, and the project includes the development of a reference implementation, serving as a proof of concept and used for evaluation purposes. The resulting system will be a secure and trustworthy electronic pollbook solution whose main purpose is to ensure 'one voter, one vote' on Election Day. An electronic pollbook system is an inherently a dynamic distributed system where multiple check-in devices and/or servers must operate in concert in providing 'one voter, one vote' guarantee, with security, integrity and auditability, and despite possible failures and the resulting need to dynamically reconfigure this distributed system on the fly. Such a system represents a convergence of several difficult problems in distributed computing as this includes: the need to reach consensus, the need to replicate for availability and fault-tolerance, the challenge of guaranteeing data consistency and longevity, the requirement for secure operation, the need to guarantee correctness in all executions, the ability to dynamically evolve the systems either because of failure or for reasons of performance. Equally importantly, electronic pollbooks is not a theoretical endeavor, but an important component of our democratic process in the digital age. This project develops and will make available the technology for constructing electronic poll book systems that are human-centric and that are based on rigorous research. The project involves graduate students, who are mentored by the investigators, and who are involved in research on the critically important problems of security and integrity in electronic election systems.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.