Receiver Beacon Transmission Interval Design Using Q-Learning focused on Packet Delivery Rate for Multi-Stage Wireless Sensor Networks

U-Bus Air is a protocol that can transmit packets by multi-hop communication, and is adopted in high-rise buildings and so on. Since U-Bus Air is time asynchronous, the communication quality deteriorates due to packet collision as the number of nodes increases. In previous research, centralized control that collects topology information in a Fusion Center (FC), calculates the communication interval, and then performs feedback. However, since information collection and feedback packet communication are required, communication overhead occurs, leading to degradation of communication quality. In addition, the authors have proposed a machine learning method that focuses on packet holding time, but local solutions have not been considered. In this paper, we propose a method for controlling the autonomously distributed receiver beacon transmission interval considering local solution by using machine learning. In this protocol, each node calculates its own PDR and transmits it to surrounding nodes with link table exchange. After a certain period of time has passed, the communication interval is determined using machine learning based on the packet delivery rate (PDR) of the surrounding nodes on the link table. Using computer simulation, we show improvements in PDR, throughput and power consumption compared to the previous research.