Performance of TVWS-based LoRa Transmissions using Multi-Armed Bandit

Abstract

Low Power Wide Area Networks (LPWANs) support long-range communication that suits them for varied IoT applications such as asset tracking, environmental monitoring, agricultural management, work site monitoring, fleet management, and smart cities. Even with the large number of connected devices, LPWANs are more power efficient than traditional IoT networks. Long Range (LoRa) is a noteworthy LPWAN technology that employs the ISM band, which experiences congestion due to the growing IoT and smart network applications. LoRa networks can utilize the available TV white spaces to overcome the congestion problem. Performance enhancement of the LoRa network in terms of energy efficiency is a significant concern. This paper proposes, for the first time, Multi-Armed Bandit (MAB) to improve the performance of the LoRa network using TVWS. We have developed a novel algorithm, LoRaT-MAB, for TVWS-based LoRa, and results show that the mean rewards increase by about 12.5% over conventional LoRa and the energy consumption for TVWS-based LoRa using LoRaT-MAB decreases by about 11.7% over conventional LoRa. The findings are encouraging and provide a basis for further research on the TVWS-based LoRa and other IoT applications.