Algorithm Diversity for Resilent Systems
Scott Stoller, Annie Liu
In cyberspace, as in many other domains, diversity provides resilience and is a robust defense against attacks. Many ways of varying computer programs have been proposed to produce diversity from a given initial program. However, these techniques do not vary the core or essence of a program—the algorithms it embodies— and therefore cannot achieve full diversity. Achieving essential diversity requires an algorithm design method that is both powerful and systematic: powerful, so that it is able to generate fundamentally different new algorithms, and systematic, so that it is able to best explore the large design space to ensure the desired resilience through diversity, while also ensuring algorithm correctness and efficiency. This project aims to develop such a method.
Prof Stoller and Liu: “This is a uniquely exciting research endeavor for the Stony Brook team, because it is an entirely new dimension for systematic algorithm design, expanding our two decades of research on systematic program design and optimization.”