Best Integer Equivariant Estimation With Quality Control in GNSS RTK for Canyon Environments

Abstract

Global Navigation Satellite System (GNSS) has been widely used for navigation and positioning in many areas, from professional applications to the mass market. However, in the real-time kinematic (RTK) positioning, the ambiguities usually cannot be fixed in natural and urban canyon environments due to the high occlusion, strong reflection, and frequent maneuvering. In this paper, the best integer equivariant (BIE) estimation with quality control is proposed. Specifically, the quality control issues based on the observation and state domains are added to the BIE estimator. First, an improved multipath processing approach is used to mitigate the effects of multipath and diffraction, and a modified detection, identification, and adaptation procedure is adopted for processing the outliers and non-line-of-sight reception. For state-domain quality control, a two-step selection and segment estimation are used to refine the ambiguity candidates. Finally, the BIE estimator is obtained. To validate the effectiveness of the proposed method, two experiments, including natural and urban canyon environments, are conducted. Compared with the float, fixed, and traditional BIE solutions, the BIE solution with quality control has the best positioning performance, including precision and reliability. Moreover, the problem of ambiguity resolution is also alleviated to a great extent. Specifically, in real-time monitoring, the real-time single-epoch positioning accuracy in east, north, and up directions are approximately 0.023, 0.014, and 0.044 m, respectively. For the low-cost urban vehicle data, the proposed method performs best regarding availability, precision, and reliability.