Fault Detection and Exclusion for INS/GPS/5G Tightly-Coupled Navigation

A solution separation-based fault detection and exclusion (FDE) framework is developed for GPS and 5G signal of opportunity (SOP) aided inertial navigation system (INS). The proposed framework fuses an inertial measurement unit (IMU) with GPS and 5G pseudorange measurements in a tightly-coupled fashion via an extended Kalman filter to estimate the ground vehicles' attitude, position, velocity, and clock errors. Solution separation tests are exploited to detect and exclude faults from GPS and 5G signals due to transmitter failures and local threats in urban environments (e.g., multipath). Experimental results are presented to evaluate the efficacy of the proposed framework under different sensor fusion scenarios. It is shown that fusing 5G signals enhances the FDE performance of the multi-sensor system in a suburban scenario: while INS/GPS fails to detect faulty GPS measurements, the INS/GPS/SOP is able to detect the fault. Moreover, over a trajectory of 1.91 km traversed in 200 s, using signals from two 5G gNBs, the INS/GPS/5G system achieved a position root-mean squared error (RMSE) of 0.81 m and maximum position error of 2.17 m. The undetected GPS fault in the INS/GPS system increased the RMSE and maximum position error to 1.83 m and 4.25 m, respectively.