Adaptive Doppler‐smoothed‐code bilateral kernel regression method for single‐frequency BeiDou receiver. Issue 4 (19th December 2022)

• Record Type: Journal Article
• Authors: Jiang, Yi; Shao, Han; Gao, Heng
• Is Part Of: IET radar, sonar & navigation. Volume 17:Issue 4(2023)
• Page Start: 628
• Page End: 641
• Publication Date: 2022-12-19
• Subjects: adaptive filters -- navigation -- receivers -- signal processing; Signal processing -- Periodicals; Radar -- Periodicals; Sonar -- Periodicals; Electronics in navigation -- Periodicals; Navigation -- Periodicals; 621.3848
• DOI: 10.1049/rsn2.12366
• Languages: English
• ISSNs: 1751-8784

Abstract

Abstract: Currently, single‐frequency Global Navigation Satellite System (GNSS) receivers dominate maritime navigation units due to their simple structure and low cost but usually cannot meet the positioning requirements of Maritime Autonomous Surface Ship (MASS). Herein, a novel adaptive Doppler‐smoothed‐code Bilateral Kernel Regression (DBKR) method is proposed, which improves pseudorange accuracy in the range domain and then reconstructs observations in the position domain. In the range domain, the Doppler observable is utilised to smooth the pseudorange with an optimal window smoothing width for the BeiDou Navigation Satellite System (BDS) receiver to alleviate ionosphere delay. In the position domain, we elaborate a bilateral kernel regression model to further reduce the positioning drift. The on‐line regression process starts with mapping the observation space to Euclidean space and subsequently fusing all observations of the same epoch using the Gaussian Radial Basis Function (RBF). Finally, the experiments under static and dynamic scenarios are carried out, which verify the validity and efficiency of the proposed DBKR method. Abstract : A novel adaptive Doppler‐smoothed‐code Bilateral Kernel Regression (DBKR) method for single‐frequency shipborne receivers is proposed, which improves positioning accuracy in both the range domain and position domain. In the range domain, the Doppler‐aid smoothing method is applied with an adaptive window smoothing width. In the position domain, we elaborate a bilateral kernel regression model to further reduce the positioning drift, which provides an on‐line regression process including spatial mapping, observation fusion and observation reconstruction.