S1), constructing a real-world direct-multipath two-dimensional color image mode data set, specifically comprising the following steps:

S11), receiving satellite original baseband signals at a current position using a global navigation satellite system (GNSS) receiver, and resolving and demodulating the received satellite original baseband signals;

S12), capturing correlation peak values of the resolved and demodulated satellite baseband signal through a correlator to acquire characterization of a possible multipath signal; and

S13), converting correlation peaks characterized by the recorded satellite original baseband signals in a direct scene and a multipath scene into a two-dimensional color heatmap through correlation peak detection, to construct the direct-multipath two-dimensional color image mode data set;

S2), building a multipath suppression model of a deep learning network based on a long-short term memory (LSTM) network and a self-attention mechanism module, and training the model;

S3), inputting the data set in step S1) into the trained multipath suppression model, and extracting a time series feature of the two-dimensional color heatmap through the LSTM network; and extracting a spatial aggregation feature of the two-dimensional color heatmap through the self-attention mechanism module; and

S4), fusing the time series feature and the spatial aggregation feature to obtain a fusion feature, and outputting a suppression result of the multipath signal after the fusion feature is processed by a fully connected layer of the model.