A Self-Organizing Map Neural Network Based RAIM Method for Standalone BDS Receivers

BeiDou Navigation Satellite System (BDS) has completed its global constellation deployment and has commenced providing navigation, positioning, and timing services to users worldwide. Considering the autonomy and security, standalone BDS receivers are adopted by receiving BDS signals completely. To improve the sensitivity of the Receiver Autonomous Integrity Monitoring (RAIM) to detect fault deviations and address the limitations of conventional threshold-based methods, this paper proposes a Self-Organizing Map (SOM) neural network based RAIM, named SOM-RAIM. The core innovation lies in redefining fault identification as an clustering problem, where multidimensional least-squares residuals (LSR) of satellites serve as features. Experimental results demonstrate that the proposed SOM-RAIM achieves a reduction in minimum detectable bias compare to threshold-based methods, while improving fault identification probability for small magnitude faults. These advancements enable earlier fault alarm and enhance robustness for BDS integrity assurance in dynamic environments.