Lifetime Improvement in Rechargeable Mobile IoT Networks Using Deep Reinforcement Learning

The rapid advancement of Internet of Things (IoT) technology has revolutionized industries and daily life through enhanced connectivity and automation. Moreover, the development of mobile IoT devices (IoTD) has extended the capabilities of these networks beyond fixed cyber-physical infrastructures, resulting in the Internet of Mobile Things (IoMT). Leveraging the IoMT applications' productivity demands judicious usage of the limited battery of the IoTD. In this regard, Mobile Energy Transmitters (MET) aided energy harvesting can improve the operational lifetime of the IoMT networks. However, IoTD mobility and non-uniform energy utilization make MET scheduling challenging in IoMT networks. Moreover, they also result in dynamic energy holes in the network. In this regard, we propose a novel approach to mitigate the emergence of energy holes by employing a deep reinforcement learning (DRL) framework for MET scheduling in IoMT networks. The proposed algorithm designs a suitable sequence of charging locations for MET visits. The simulation results indicate the superiority of the proposed algorithm over other MET-scheduling algorithms. Furthermore, the proposed DRL algorithm significantly enhances the operational lifetime of IoMT networks, thereby increasing network stability and continuous functionality. The results motivate using the proposed DRL algorithm in self-sustained IoMT networks.