Controlled Quantum Semantic Communication for Industrial CPS Networks

Computing-intensive semantic communication emphasizes context, enabling the extraction of task-specific semantics from the source data and the reconstruction of the intended meaning at the destination. In industrial cyber-physical systems (CPSs), this approach can optimize automation processes while minimizing communication overhead with efficient bandwidth use in environments where machines, sensors, and controllers must communicate frequently. By integrating quantum communication with computing-empowered semantic methods, we can achieve unprecedented efficiency and security in task-oriented data transmission, effectively safeguarding against eavesdropping and other attacks. This paper presents a controlled quantum semantic communication (QSC) framework that leverages semantic extraction for anomaly detection in industrial CPS networks and employs controlled quantum communication to send the data securely with high semantic fidelity. A machine learning model extracts semantic information from images as the hull point data representing defective regions as pixel points. This data is then transmitted with high fidelity using quantum communication with controlled quantum state preparation. We use discrete- and continuous-variable states to simulate quantum binary phaseshift keying (BPSK) and $M$-ary pulse position modulation ($M$-PPM), respectively. At the receiver, these quantum states are measured using optimal quantum decision-making and converted back into the hull point data, thereby generating the anomaly map. This map is overlaid on a template image to highlight defect positions, which can be used for industrial quality control. Furthermore, we simulate the controlled QSC framework (BPSK and $M$-PPM) across a diverse set of anomaly detection examples and evaluate the QSC performance in industrial CPS networks.