CAREER: Understanding and Ensuring Secure-by-design Microarchitecture in Modern Era of Computing

Microarchitectural attacks (i.e., side and covert channels) have opened a new chapter in computer system security. These attacks manifest by exploiting timing observations from microarchitectural events, leading to either illicit communications among isolated entities or exfiltration of private data. Due to the rapid integration of new microarchitecture innovations by both industry and academia, it is widely believed that the microarchitectural attack surface will continuously expand. Therefore, a comprehensive and scalable evaluation of microarchitecture security beyond manual inspection is imperative. This project investigates systematic and automated leakage reasoning techniques on given microarchitecture designs to identify new vulnerable microarchitectural states and discover novel exploitation mechanisms. The project then explores principled mitigation and prevention techniques to thwart microarchitectural information leakage. This project includes the following complementary research efforts: (1) Building models using microarchitecture-level abstraction and constructing model checking-based frameworks that reason timing explicitly for microarchitectural leakage analysis; (2) Developing ways to instantiate microarchitectural vulnerable abstract patterns to ISA-level executions, validate and quantify new leakage attack vectors in both commercial-off-the-shelf hardware and novel microarchitectural mechanisms for future systems; (3) Exploring leakage-free and composable secure microarchitecture schemes with low overhead by employing mitigation techniques according to performance characteristics of the underlying hardware. The proof-of-concept attack/defense artifacts and analysis tools developed from this project could benefit computer architects and system designers in assessing and limiting microarchitectural information leakage. This project also offers extensive opportunities for undergraduate students to engage in the field of computer architecture and security through a synergy of education and research training efforts. Successful outcomes of this project can provide holistic understandings of microarchitecture security and contribute to the development of secure-by-design architectures in modern and future computing. 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.