Variable length LMS adaptive filter for carrier phase multipath mitigation

Multipath on carrier phase measurements is among the major error sources for short baseline positioning. A new method is proposed to improve the accuracy of the positioning results by mitigating the multipath effects on carrier phase measurements using the variable length Least Mean Square (VLLMS) adaptive filter. The performance of the filter is analyzed as well as compared with that of the standard LMS adaptive filter using a set of carrier phase observation data of two consecutive days collected in a short baseline experiment. Because the two antennas are static, the multipath error is the only dominant component in the carrier phase double-differenced residuals and indicates a repeated pattern. The numerical results show that both filters can significantly mitigate the multipath effects in carrier phase double-differenced residuals, and hence improve the positioning results. Furthermore, the VLLMS filter that reduces up to about 47.4% of the multipath effects on 3D positioning performs better than the LMS filter that only reduces 22.0%. Both filters are suitable for real-time applications.