Research and Analysis of Indoor Positioning Accuracy Based on Machine Learning and Particle Filtering Algorithm

The positioning problem in the indoor environment has always been an unresolved area, and the severe attenuation of the signal and the multipath effect has led to its low positioning accuracy. In this paper, the location fingerprint method is used for indoor positioning. Firstly, the ray tracing technology is used to obtain the indoor positioning experimental data, and then the multiple classification regression algorithms in machine learning are used to carry out experiments respectively. Finally, the Particle Filter (PF) algorithm optimizes the above algorithm. By analyzing the experimental results, it is shown that machine learning combined with the particle filter algorithm can greatly improve the accuracy of indoor positioning.