Research on Multipath Suppression Method of Satellite Navigation Signal Based on Sparse Representation in The Background of Artificial Intelligent

Abstract

In radar signal processing, radar target parameter estimation is one of the important tasks of radar target detection. Multipath effect is one of the main error sources of satellite navigation system. In the existing research, the method based on maximum likelihood estimation criterion is considered as the best multipath suppression technology, but its calculation is particularly complicated when there are multiple signals and the number of paths is unknown. This paper mainly studies the application of sparse representation theory in radar target parameter estimation, taking Doppler frequency estimation as an example. The sparse model of Doppler frequency estimation is established, and the frequency estimation is carried out by the classical greedy iterative reconstruction algorithm, and high resolution is obtained. In this paper, a variable step-size fitting algorithm is adopted, which is simple and practical, has little computation, is independent of multipath number, is easy to realize and is easy for real-time signal processing.

Enhancing BDS-3 PPP-AR with observable-specific signal biases

“…BDS-3 comprises satellites in three distinct orbits, namely the medium The BDS-3 multi-frequency signals, coupled with the three kinds of satellites in MEO, IGSO and GEO orbits, are increasing the research dimension in the GNSS community. In both BDS-only and multi-GNSS constellations, a considerable volume of research has followed in different fields such as multipath [2,3], signal-to-noise ratio (SNR) [4], GNSS interferometric reflectometry [5,6], quality of BDS products [7,8], precise time transfer [9,10], and ambiguity resolution (AR) [11]. While multi-frequency and multi-constellation GNSS may be applied in such broad areas, the increase in signals itself complicates the treatment of biases in the formulated observables.…”