Moving Average Filters for Faster GPS Receiver Autonomous Integrity Monitoring

Yi-Hsueh Tsai, Fan-Ren Chang, and Wen-Chieh Yang

Abstract:	In this paper, the Moving Average filter (MA-filter) is proposed for Receiver Autonomous Integrity Monitoring (RAIM) to guarantee the safety of satellite navigation. The conventional algorithms for RAIM use the least-squares residue method, in which only current information is considered. However, the historical data may be helpful in satellite failure detection. Taking advantage of the dynamic characteristic, we use the MA-filter in RAIM. The weighting of each data inside the moving window is equal. The output data of the MA-filter is compared with the threshold associated with a given fixed false alarm rate (FAR) to guarantee normal navigation. To obtain the threshold, the MA-filter is approximated as a “discrete finite-state Markov chain” with a stationary transition probability matrix. Although the calculation of threshold for MA-filter is complex, it is independent of satellite geometry. Therefore, it can be computed off-line and tabulated in advance. Based on a given fixed FAR, the threshold value is decreased in MA compared with those of the conventional algorithms, because the variance of the output of MA-filter is reduced. Thus, the detection speed is higher for the MA-filter. The change of the number of satellites in view is solved via the Probability Integral Transformation (PIT) method. A fast PIT algorithm is also introduced to significantly reduce the computation time. In the simulation, a 24-satellite constellation with perfectly circular orbits is assumed. A total of 1152 space-time sample points are produced according to the RTCA DO-208. Two types of pseudorange error are used to simulate the satellite malfunction: the ramp error and the step error. The windows size of MA-filter ranges from 1 to 5. The detection times of our method are reduced in both types of errors when compared to those of conventional approach. However, those are some interesting results in the simulation results. Under ramp errors, the MA-filter of the larger windows size is better for detecting small errors. Under step errors, the MA-filter of the larger windows size is better for detecting all kinds of errors.
Published in:	Proceedings of the 2002 National Technical Meeting of The Institute of Navigation January 28 - 30, 2002 The Catamaran Resort Hotel San Diego, CA
Pages:	666 - 675
Cite this article:	Tsai, Yi-Hsueh, Chang, Fan-Ren, Yang, Wen-Chieh, "Moving Average Filters for Faster GPS Receiver Autonomous Integrity Monitoring," Proceedings of the 2002 National Technical Meeting of The Institute of Navigation, San Diego, CA, January 2002, pp. 666-675.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In