Empirical Performance Analysis and Channel Modeling of UAV-Assisted LoRa Networks

Abstract

Long range (LoRa) technology has been widely proposed to empower cost-efficient and scalable communication backbone for the battery-constrained internet of thing (IoT) device. However, the LoRa-based IoT network suffers from a limitation in coverage caused by high signal attenuation due to the terrestrial blockages especially in the urban and suburban environments. To overcome this issue, the use of unmanned aerial vehicles (UAVs) as a flying LoRa gateway to transfer messages from ground LoRa nodes to the cellular base station is proposed. Thus, it is of great interest to accurately model the propagation channel and performance analysis of UAV-Assisted LoRa network based on real measurements. In this paper, a UAV-Assisted LoRa Network is empirically implemented and analyzed. First, LoRa node and gateway are designed and practically implemented using LoRa Ra-02 module, Arduino Uno board, and SIM800l GSM module. Then, the LoRa gateway is mounted over DJI Phantom 4 Pro drone to have a flying LoRa gateway. In this UAV-assisted LoRa Network, the LoRa node sends data to the LoRa gateway over LoRa links, and then LoRa gateway sends data over cellular link to the server. The UAV-Assisted LoRa Network is set up over the campus of Shahid Chamran University of Ahvaz. For this network, different experiment scenarios including line-of-sight (LoS) or none-LoS (NLoS) links and ground or flying LoRa gateway are considered. Following the real-world measurement in these scenarios, the channel model parameters of LoRa links between flying or ground gateway and node are empirically estimated. Our results show that the path loss exponent of LoS LoRa link is 2.57 for ground LoRa gateway, while the LoS path loss exponent of LoRa link decreases to 1.65 for 80m-height flying LoRa gateway. Moreover, our practical analysis shows that the useful coverage area of network rapidly increases when the flying LoRa gateway is used instead of ground LoRa one.