Multi-Armed Bandits for Resource Allocation in UAV-Assisted LoRa Networks

This work taps into the power of Reinforcement Learning (RL) and drone technology to improve the performance of Long Range Wide Area Network (LoRaWAN) networks with potential use in different domains including the Internet of Things and military applications. An Unmanned Aerial Vehicle (UAV)-assisted LoRa network is investigated, where the UAV hovers above a considered area to work as a LoRa Gateway (GW) for collecting information from ground LoRa end nodes. Three efficient Multi-Armed Bandits (MAB)-based resource allocation (RA) algorithms are exploited to grant end nodes (ENs) the autonomy to fine-tune their transmission parameters independently. The simulation results show improvement compared with traditional fixed GW deployment scenarios in terms of average Packet Delivery Ratio (PDR).