Integrating Direction-of-Arrival Estimation Function With Reconfigurable Intelligent Surfaces for Self-Adaptive Wireless Networks

This paper presents a self-adaptive reconfigurable intelligent surface (RIS) control system that operates online without human intervention, aimed at addressing future communication challenges in dynamic environments. The system incorporates a cost-effective processor with a direction of arrival (DoA) estimation module to realize an end-to-end autonomous control loop, in which received power measurements are processed on-device to infer DoA and to update the RIS bias configuration for beamforming at 29.5 GHz. By employing a preconfigured bias mapping dataset, the RIS adjusts reflection angles in increments of 10°, ranging from 0° to 60°, thereby manipulating the electromagnetic (EM) environment. Measurements indicate that the proposed system achieves an improvement of up to 15 dB in received power when utilizing the RIS, even under non-ideal conditions involving variations in distance and angle. Furthermore, the system demonstrates its capability to enhance received power through atypical reflection, underscoring the potential of RIS for energy-efficient wireless communication. This study confirms the feasibility of self-adaptive RIS control using low-complexity embedded hardware, showcasing its practical applicability in dynamic EM environments.