In view of the various faults that may occur while tracking the satellite navigation signal, the residual observations of the Kalman filter are used for filtering, and the faults can usually be quickly detected and isolated without the use of redundant observations. This paper establishes a software receiver platform, uses Kalman filter for vector tracking of navigation signals, simulates the occurrence of faults in the tracking loop and relies on residual sequence filtering to detect and eliminate erroneous observations. Experiments show that under the short-term and large-scale interference simulated by the algorithm, the integrity method can quickly detect and eliminate the influence of erroneous observations on loop tracking without redundant observations.