Implementation of Encrypted Control with Possessing State Matrices with Integer Elements in Linear Discrete-time Controllers

The study examines a method to enhance the performance of the control system without the need to reset controller states to zero in encrypted control. Possessing state matrices with integer elements in linear discrete-time controllers lets these controllers work for infinite time in partially encryption. However, computation costs will increase but performance is enhanced. The computational overhead and trade-offs with performance when applying various theories that have been extended to encrypted controllers are discussed. In cryptographic algorithms, computation volume often increases due to data encryption. Therefore, these controllers must be designed to consider resource constraints for computations, such as hardware limitations and required response time. These methodologies pave the way for future advancements, enabling the integration of increasingly sophisticated controllers. Examples include machine learning-driven systems such as reinforcement learning or other AI-based controllers. This innovation opens up opportunities for implementing more intricate control mechanisms across various domains, heralding a new era of control system design and application.

Keywords:

Cloud computing , cyber security , encrypted control