Secure online services for private user data

点击翻译按钮获取中文摘要

When users are required to, or opt to, disclose private/sensitive data to an organization or government entity for processing, a balance must be struck between security and convenience. Online submission is often the most convenient but it also carries substantial risks. This research program proposes the development and deployment of innovative cryptographic protocols for providing secure online election systems and secure online genetic testing. Any security-critical online service must at least provide users with a process for determining they are connected to the intended party (server authentication) and then provide a confidential communication channel. These properties are intended when a user visits a website beginning with https:// (and the browser displays a lock), however, the trust assumptions and infrastructure this protocol (called SSL/TLS) relies on have been recently questioned.The research program will proceed in three thrusts: one toward improving the SSL/TLS infrastructure, one toward advancing beyond basic SSL/TLS protection for the specific application of online voting, and one that does the same for online genetic tests.When users visit a website over https://, the site provides a certificate of their identity, endorsed by a business or organization called a Certificate Authority (CA). Any CA can issue a certificate for any site. The number of CAs has proliferated, the baseline criteria for validation has declined, and high profile breeches of CAs have been publicly disclosed. This research program will design, implement, and test alternative mechanisms for certification.Online voting has been used municipally across Canada, is to be piloted federally by Elections Canada, and is being deliberated on by some provinces. Unlike online banking, where an incorrect, mistaken, or fraudulent transaction can be seen by the user, ballot secrecy mandates that individual votes cannot be displayed (which would enable vote selling). This research program will design, implement and test novel end-to-end verifiable (E2E) voting systems that provide a provably correct tally while maintaining the secrecy of each voter's ballot, even if the voter is complicit in demonstrating how they voted or uses a malware-infected personal computer to cast their ballot.The decreasing cost of whole genome sequencing has the potential to revolutionize healthcare, allowing genetic tests and personalized medicine. However, your genome is private information. Once it is disclosed or leaked, the privacy can never be recovered. This research program will design, implement and test novel cryptographic protocols for allowing these tests to be performed on an encrypted genome, while offering proof that the test was performed correctly (even if some aspects of the test remain confidential).The results of this research program should be informative and useful to government election agencies, privacy commissioners, standards institutes, internet working groups, other cryptography, security and privacy researchers, and to the elections and healthcare industry. It will further equip a set of security professionals with the skills required to work or consult in any of these areas, and to apply their knowledge and experience to novel domains.