Statistical Characterization of Path-Loss for 900 MHz LoRa Using a Bike-Share IoT Network

The Internet of Things (IoT) has evolved over the past decade and is now deployed in a myriad of applications such as environmental variable sensing, agriculture, healthcare, smart homes, and smart cities, to name just a few. In this paper, we leverage a bike-share IoT network to measure RF path-loss over a wide area within a city. Compact sensors are attached to the bikes, providing real-time measurements of several variables, including received RF power. Our methodology can be replicated to measure air quality indicators, traffic, or any other variables that can be sensed in compact modules mounted on bikes. Using this system, we perform a statistical characterization of path-loss for LoRa at a frequency of 905.3 MHz. The measurement campaign took place in Pasto, Colombia, a mountainous city in the Andes mountains with medium-size buildings and narrow streets. We found that the path-loss exponent in this environment is 2.4, with a standard deviation of 7.9 dB. Cell radius for LoRa gateways under this propagation conditions is 410 m for 5% outage and 729 m for 10% outage. This path-loss characterization can be used for initial design and for optimization of LoRa IoT networks.