Cryptographic Computation for Control and Optimization

My research focus is on establishing the security and privacy foundations for decision-making and control in cyber-physical systems within adversarial environments. In these systems, information from physical space is aggregated in cyber space, where intelligent decision-making and real-time control are executed. A fundamental challenge in this paradigm is the inherent trade-off between the intensive data utilization required for such computations and the rigorous guarantees of security and privacy.

To bridge this gap, I am developing a theoretical framework for encrypted control by integrating control theory with cryptography. This framework leverages cryptographic computation techniques, such as homomorphic encryption and multi-party computation, to perform control and optimization algorithms directly on encrypted data, including feedback gains, sensor measurements, and control actions. By employing encrypted control, we enable the secure outsourcing of control computations to untrusted cloud servers and facilitate privacy-preserving distributed control among multiple agents. Furthermore, I am interested in the mathematical formalization of security and privacy within the context of control systems.