Explicit Model Predictive Control with Quantum Encryption

A team of researchers including Yingjie Mi, Zihao Ren, and Daniel E. Quevedo has introduced a novel security framework for industrial control systems in a new arXiv paper. The work, 'Explicit Model Predictive Control with Quantum Encryption,' merges two advanced technologies: Explicit Model Predictive Control (MPC)—a method for controlling constrained systems by solving optimization problems offline—and quantum cryptography. The core innovation is a cloud-based architecture where the heavy computation of MPC is performed ahead of time, resulting in a lightweight, piecewise-affine controller. This controller's critical online evaluation between sensors and actuators is then protected using encryption keys derived from quantum mechanics.

The security relies on Quantum Key Distribution (QKD), which uses entangled Bell pairs to generate shared secret keys that are theoretically immune to computational attacks. The authors develop a specific protocol that uses these quantum keys to encrypt the control signals, proving that the original control action can be exactly recovered by the authorized actuator. Their numerical analysis shows this quantum-encrypted approach has lower online computational overhead than traditional classically-encrypted MPC, making it more efficient for real-time operation. The paper frames its security benefits in terms of ensuring the confidentiality of both the private dynamic data of the industrial plant (the 'state') and the control commands themselves, which could be targets for sabotage or espionage.