Multipath Error Correction for Smartphones and Its Impact on Single Point Positioning

With the release of multi-constellation and multi-frequency GNSS chips, the positioning capabilities of smartphones have been improved. However, smartphones are commonly equipped with low-gain microstrip antennas, which make GNSS observations suffer from severe multipath effects. The results of short baseline single difference (SD) show that the multipath error of smartphones is almost 5-10 times higher than that of geodetic receivers, which greatly limits the positioning accuracy of smartphones. The multipath effect is affected by various factors such as equipment, environment, and signal, and it is usually difficult to establish an effective mathematical model for correction in complex urban scenarios. Considering that multipath errors have low-frequency characteristics, this paper proposes a method based on wavelet transform, which can effectively detect and extract multipath errors. First, the short baseline SD residual between the smartphone and the reference station is decomposed into low-frequency error and high-frequency noise. Then, the low-frequency error is compensated to GNSS pseudorange observations after doing wavelet inverse transform reconstruction, thus weakening the influence of multipath errors on the positioning. The experimental results show that different wavelet bases and decomposition layers can effectively improve the positioning accuracy, and the improvement is more obvious when the number of decomposition layers is smaller. When dB1 wavelet base and one layer of decomposition coefficients are selected, the single point positioning (SPP) accuracy of Huawei P40 and Xiaomi 8 in U, N, and E directions reaches 1.07 m, 1.31 m, 0.59 m, and 1.32 m, 0.87 m, 0.47 m respectively after multipath error compensation. Compared with the uncompensated multipath, the positioning accuracy of the two phones is improved by 53.5%, 48.6%, 60.7%, and 25.4%, 32%, 55.2%, respectively.