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

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.