Field Trial of Reconfigurable Intelligent Surface with Statistics-Based Optimization for 5G Commercial Networks

As study of reconfigurable intelligent surface (RIS) moves from theoretical analysis to practical engineering, field trials become more important to understand how RIS performs in the real world. So far comprehensive field trials have been scanty, particularly in commercial networks. In this article, we carry out a field trial of RIS-aided 5G networks where the element phases of RIS are optimized with statistics-based algorithms, instead of using explicit estimation of angle of arrival/departure or channel state information (CSI) feedback. The trial covers three scenarios: outdoor macro station to indoor parking lot, indoor coverage extension, and outdoor interference suppression. Significant performance improvements by RIS are observed in reference signal received power (RSRP), signal to interference noise ratio (SINR), and medium access control (MAC) layer throughput for both downlink and uplink. The trial results show the possibility of pre-standard deployment of RIS in 5G commercial networks and the effectiveness of statistics-based RIS phase optimization for complicated radio environments.