Analysis and design of robust reconfigurable intelligent surfaces using a statistical approach

Reconfigurable Intelligent Surfaces (RIS) and Smart Skins are well-established metasurface-based technologies for the wireless electromagnetic environments of the future. Despite the accurate modeling and fabrication techniques of such devices, their performances may be affected by unpredictable mutual coupling, fabrication defects, mechanical and thermal/mechanical stress that change the local response of the metasurface, realizing an electromagnetic roughness of the surface properties. In this contribution, we statistically model such electromagnetic roughness to predict the modifications to the local surface impedance. We derive a profile transformation function for the surface impedance that allows making the RIS more resilient to such unpredictable defects. A proper set of numerical simulations demonstrates the robustness of such design considering an anomalous reflective metasurface operating at 10 GHz.