Voting is the most fundamental act of participation in our democracy, yet voting technology has suffered diverse failures and crises. Voting is a remarkably hard, interdisciplinary problem, with seemingly conflicting goals of verifying ballots while maintaining secret ballots to safeguard against bribery and coercion. This project is producing innovative and fundamental computer security technologies supporting remote and kiosk voting scenarios, i.e. high assurance applications on un-trusted hardware/software. This research is designing and developing a secure execution partition (SEP) prototype based on an un-trusted hardware and software base. The research leverages the trusted components, e.g. Trusted Platform Module (TPM) and vPro chipsets, currently available on production motherboards, and programming language constructs producing a prototype that enables easy formal verification by third parties. The execution of high assurance software on un-trusted components has been a research problem for decades. While specifically addressing the remote voting problem in this effort, results will also significantly impact health care privacy, remote financial transactions, and remote authentication. The findings and experience from this research will be incorporated into UMD?s system's programming/architecture, operating systems, and information security courses at the graduate and under graduate level. Exceptional undergraduate students will be active solicited for participation in the research. Finally, the results of this research will impact society through improvements in future operating systems.
Secure Execution Partition (SEP) for Remote Voting