Revisiting Reconfigurable Intelligent Surface for Improved Coverage in Nuclear Plants

Advance reactor's operation and monitoring can be significantly improved using wireless technologies. These wireless solutions can also support plant modernization while enhancing the operational efficiency of existing nuclear fleet, and other mission-critical energy sectors. However, the adaptation of these wireless technologies can be challenging due to the heavy metallic environment within the nuclear plant. The unique environment of nuclear power plants (NPPs) with metallic structures often blocks signal propagation resulting in reduced coverage within the facility. This paper proposes to develop a Reconfigurable Intelligent Surface (RIS)-assisted architecture for nuclear plants to improve wireless signal coverage. The aim is for RIS to intelligently reflect the signal toward the intended direction in the event of any blockade. In addition, we applied a reinforcement learning solution, named multi-armed bandit, that picks the best RIS location for a dynamic network scenario. Then, we design the RIS to optimally direct the reflected signal to maximize the signal strength at the receiver. The simulations represent a practical nuclear environment with RIS-assisted architecture and compare that with a no RIS scenario as the benchmark. Proving RIS-assisted architecture with enhanced coverage for wireless data transmission could serve as a blueprint advancing concept for operations of future NPPs.