Multipath Mitigation in GNSS Positioning by the Dual-Path Compression Estimation

摘要

The combinations of triple-frequency carrier phase have been widely used for precise positioning in the global navigation satellite system (GNSS). However, the position accuracy of this system is negatively affected by multipath (MP) interference and highly kinematic problems. To bridge this gap, in this paper, a dual-path compression estimation algorithm based on generalized least absolute shrinkage and selection operator (LASSO) method with penalty terms is proposed to mitigate MP. We establish a linear observation model, taking into account triple-frequency carrier phase and Doppler shift. In order to improve the accuracy of the compression estimation, a small penalty term is added after the re-weighted L1 regularization. Accordingly, the linear observation model is transformed into the MP estimation model, with the weight matrix consisting of the carrier-to-noise ratio function and satellite elevation. We use the dual-path algorithm to solve the LASSO problem in the MP estimation model. QR decomposition and the specialized implementation strategy are applied to improve the dual-path LASSO with penalty (DPLP) algorithm. The results on real-world dataset show that the sparse estimation method proposed in this study can effectively reduce the carrier phase MP errors at triple-frequency so long as the number of MP-affected satellites is less than five (the visible satellites is eight-ten).