Beamforming Design in Reconfigurable Intelligent Surface-Assisted IoT Networks Based on Discrete Phase Shifters and Imperfect CSI.

In this paper, we study reconfigurable intelligent surface (RIS)-assisted networks to support Internet of Things (IoT) devices. We propose RIS beamforming strategies to maximize sum-rate and fairness and analyze the RIS location. We derive a theoretical lower-bound of the minimum number of RIS elements needed to guarantee specific network performance metrics and validate our results via simulations.We present two RIS scenarios in this study, both with the same total number of RIS elements: (i) centralized RIS, where a single RIS assists the network, and (ii) distributed RIS, where each transmitter has its own dedicated RIS. We study addressing two practical challenges related to RIS elements and channel state information (CSI) assumptions. First, we consider hardware limitations by assuming that each RIS element is equipped with a discrete phase shifter (PS). Second, we investigate the impact of CSI perfectness and availability in the network; therefore, we evaluate the performance of the RISassisted network under two scenarios: (a) centralized RIS with imperfect global CSI and (b) distributed RIS, where imperfect local CSI is available at each transmitter.