Design and implementation of LoRa geolocation algorithm in NS3 network simulator

In recent years, TDOA-based LoRa geolocation technology is favored by IOT researchers. However, the current researches on this technology are mostly physical experiments, which are not scalable enough. In this paper, the highly open source NS3 network simulator is used to design and implement the LoRa geolocation process based on TDOA. On the basis of the existing LoRaWAN module, the gateway receiving time extraction, clock noise simulation and TDOA-based coordinate solving algorithm are added to simulate the performance of LoRa geolocation in the noise environment. Simulation results show that the improved LoRaWAN module can support the realization of LoRa geolocation well, and the simulation results conform to the TDOA positioning accuracy range in the LoRaWAN geolocation whitepaper published by LoRa Alliance. Specifically, when the standard deviation of gateway clock noise is 100ns and the gateways adopt the strategy of uneven distribution, the mean error of LoRa geolocation using three gateways is 130.36m, and the mean error of LoRa geolocation using five gateways is 33.00m. This simulation provides a basic platform for LoRa geolocation research and can be used for reference by subsequent researchers.