Efficient Satellite Selection Method for Instantaneous RTK-GNSS in Challenging Environments

Abstract

The global navigation satellite system (GNSS) can potentially provide centimeter-level positioning using real-time kinematic (RTK) positioning. However, in static positioning, such as for surveying, receivers easily receive multipath signals continuously. Our goal was to improve the performance of instantaneous RTK-GNSS in multipath environments. Two conventional satellite selection methods based on the idea of correctly removing multipath signals, which allows more reliable solutions, were evaluated in this study. The first method is based on using signal-to-noise ratio (SNR) observations to mask measurements having degraded quality. In the second method, a mask of sky obstacles is generated using a fisheye view lens camera to detect non-line-of-sight (NLOS) signals. In this study, several static tests were performed to evaluate these conventional methods. The results show that both methods can efficiently improve availability. Furthermore, the performance when using a fisheye view mask was slightly better than that when using the SNR method, in particular for situations where a powerful reflected signal by NLOS was received. Based on these results, an improved SNR-based satellite selection method that uses the SNR fluctuation magnitude for a certain period is proposed. The results show that this method effectively improves the performance as compared with the conventional SNR mask.