A Bayesian Probabilistic AOA Localization Algorithm

In the application of fifth-generation (5G) technology, the research of large-scale antenna technology and millimeter-wave beamforming technology has promoted the improvement of positioning accuracy based on the angle of arrival (AOA) method. The AOA positioning only needs to measure the incoming wave direction of the transmitter, eliminating the dependence on signal time synchronization. Due to multipath and non-line-of-sight (NLOS) propagation, the accuracy of AOA measurements is severely affected. Therefore, in most cellular systems, AOA is often not used as the main positioning parameter, but as an auxiliary positioning parameter. In this paper, we propose a positioning algorithm based on Bayesian probability. This method not only solves the multi-value problem in AOA positioning but also further eliminates the AOA deviation caused by NLOS. The method proposed in this paper can adaptively learn channel bias and modify the AOA measurement. Experiments show that our proposed method significantly improves the positioning accuracy, even if NLOS errors occur.