Hybrid RIS‐Assisted UAV‐IoT Communications With Wireless Power Transfer Employing Reinforcement Learning

ABSTRACT

In recent years, Internet of Things (IoT) technology has been widely used; however, many devices have limited energy. Therefore, it is essential to power IoT devices while maintaining seamless connectivity. Against this background, in this paper, we propose a simultaneous wireless information and power transmission (SWIPT) scheme based on a hybrid active–passive reconfigurable intelligent surface (Hybrid‐RIS)–assisted UAV communication system. Hybrid‐RIS not only intelligently reflects signals but also amplifies them to improve performance. To improve the learning efficiency of the proposed system, this paper proposes to use reinforcement learning algorithms to optimize UAV power, UAV position, energy reception time of IoT devices, phase of RIS, and power allocation of active RIS. We also improve the Q‐learning algorithm to classify the action space and select the action type before selecting the action. This improvement allows complex environments that initially relied on neural networks to converge. Ultimately, the research results also proved that hybrid RIS has a higher throughput than traditional passive RIS. These results provide a reliable reference for applying Hybrid‐RIS technology in complex urban environments.