Adaptive Smart Radio Environment (ASRE): New Paradigm for Wireless Communication Networks

Abstract

Recent efforts have promoted Smart Radio Environment (SRE) to enhance the reception quality in high-frequency bands via reconfigurable intelligent surfaces (RISs) for supporting hyperconnectivity of Beyond 5G/6G Era. The functionality of SRE is based on the more degrees of freedom that come from electronically controlling the environment itself rather than transmitter and receiver. Accordingly, the spectral efficiency and the sum rate throughput can be enhanced by applying customized transformations to the electromagnetic radio waves. However, futuristic applications that are envisioned to be time-critical demand much more than what SRE can handle because that too much time is spent on sensing the environment and applying customized transformations. In this work, we propose a novel concept of adaptive SRE (ASRE), which intends to start from the semantic perception of a wireless environment to explore the learning and evolutionary mechanisms of SRE through the loop of recognition, adaptation, and proaction. In combination with the techniques known in artificial intelligence (AI), such as deep reinforcement learning, knowledge graph, etc., this technology can provide enhanced ultra-reliable low-latency communication (URLLC) services. Furthermore, the wireless environment is expected to become not just adaptive, but partially intelligent by predicting and utilizing the changes in the communication environment. To corroborate the rationality and superiority of ASRE, we also present simulation results related to the typical dynamic NLOS scenario. Finally, we highlight numerous open challenges and research directions.

Keywords

• Smart radio environment (SRE)
• reconfigurable intelligent surface (RIS)
• deep reinforcement learning (DRL)
• sensing
• knowledge graph