Anomalous and Specular Reflections of Reconfigurable Intelligent Surfaces: Configuration Strategies and System Performance

Reconfigurable intelligent surfaces (RISs) enable wireless communication in harsh propagation conditions. In fact, when impinging on an RIS, wireless signals can be anomalously reflected, i.e., not obeying to the Snell’s law, to reach intended destinations at arbitrary locations in space and time. However, current reconfiguration strategies do not consider the presence of classical specular reflections (SRs), which exist due to hardware realizations/impairments and/or the fact that RISs are placed over larger non-reconfigurable surfaces in practical scenarios (e.g., walls). This phenomenon can harm wireless systems performance due to the energy spilled over non-intended directions, thus increasing interference and other metrics if not taken into account when configuring RISs. In this paper, we focus on the SRs generating from a surface behind the RIS, and we propose an SR-aware end-to-end system model and RIS reconfiguration that take into account this undesired effect. Numerical results show the effect of SRs on the achievable rate regions of an interference channel under various RIS configurations, with a considerable difference between SR-aware and SR-agnostic strategies.