Adaptive LoRaWAN transmission exploiting reinforcement learning: The industrial case

Wireless technologies play a key role in the Industrial Internet of Things (IIoT) scenario, for the development of increasingly flexible and interconnected factory systems. A significant opportunity in this context is represented by the advent of Low Power Wide Area Network (LPWAN) wireless technologies, that enable a reliable, secure, and effective transmission of measurement data over long communication ranges and with very low power consumption. Nevertheless, reliability in harsh environments (as typically occurs in the industrial scenario) is a significant issue to deal with. Focusing on LoRaWAN, adaptive strategies can be profitably devised concerning the above tradeoff. To this aim, this paper proposes to exploit Reinforcement Learning (RL) techniques to design an adaptive LoRaWAN strategy for industrial applications. The RL is spreading in many fields since it allows the design of intelligent systems using a stochastic discrete-time system approach. The proposed technique has been implemented within a purposely designed simulator, allowing to draw a preliminary performance assessment in a real-world scenario. A high density of independent nodes per square km has been considered, showing a significant improvement (about 10%) of the overall reliability in terms of data extraction rate (DER) without compromising full compatibility with the standard specifications