Multi-scene LoRa positioning algorithm based on Kalman filter and its implementation on NS3.

In recent years, LoRa positioning technology has been favored by researchers of the Internet of Things because of its wide coverage, low power consumption, low cost and applicability to indoor and outdoor positioning. However, the positioning process is susceptible to interference, fading, clock accuracy and other factors. The fine-grained positioning accuracy is generally only in the range of 20 similar to 200 m, which is still far from the actual application requirements. At present, there are many physical experiments in the research of LoRa positioning, whose scalability is insufficient. The simulation research on LoRa positioning uses more commercial software, and there are few simulation experiments based on simulation platforms with high open source and good portability. Therefore, this paper proposes a multi-scene LoRa positioning algorithm based on Kalman filter. The positioning process is implemented in the open-source NS-3 network simulator, and Kalman filter is used to improve LoRa positioning accuracy. The simulation results show that the improved LoRaWAN module can well support the implementation of LoRa positioning in various scenarios, and Kalman filter can significantly improve the positioning accuracy. In addition, the fluctuation of positioning error after filtering is greatly reduced. By improving the LoRaWAN module, it not only realizes the positioning simulation of fixed node scenario, moving node scenario and different number of gateways scenario, but also provides a platform for the follow-up research of LoRa positioning.