A Novel Path Planning and Node Selection Method Using Reinforcement Learning in NTN IoT Networks

With the rapid deployment of 5G networks in recent years, the characteristics of high bandwidth, low latency, and low energy consumption of 5G networks have enabled the rapid development of IoT (Internet of things) technology. However, 5G networks cannot provide high-quality wireless coverage for many IoT devices in border areas and hotspots with a high signal density that lack fixed infrastructure. Therefore, this paper uses the UAV (unmanned aerial vehicle) to carry the communication platform to build the NTN (nonterrestrial network) to provide wireless coverage for terrestrial fixed and mobile IoT devices. Meanwhile, since the NTN needs to provide wireless coverage for many IoT devices, we use deep reinforcement learning to provide path planning for the UAV communication platform to improve the efficiency of wireless coverage. We build a simulation environment to evaluate the performance of the NTN network for wireless coverage of IoT devices in urban hotspot areas. Experimental results show that the method proposed in this paper can provide higher downlink rates for more IoT devices than NB-IoT (narrowband Internet of things).