基于压缩感知的ESPRIT波达方向定位OA

In order to solve the problem of low positioning accuracy of traditional algorithms in the direction of arrival under limited con-ditions,a CS-ESPRIT algorithm is proposed,which combines the compressed sensing algorithm of S-OMP with the traditional direction of arrival algorithm of TLS-ESPRIT.On the one hand,it solves the inherent problem that the compressed sensing algorithm cannot han-dle coherent signals,on the other hand,it improves the recovery accuracy under low snapshots.The compressed sensing S-OMP algo-rithm is used to calculate part of the input signal branch set,the noise space is established,the remaining branch set is calculated,the complete input signal matrix is obtained,and then the TLS-ESPRIT algorithm is used to obtain the source location.Experiments show that the proposed algorithm not only inherits the high-precision recovery of the compressed sensing algorithm in the case of a small num-ber of snapshots,but also applies to the case of coherent signal recovery.According to the simulation experiment,coherent signal pro-cessing RMSE below 10 snapshots is 20 percent lower than that of other coherent signal processing algorithms,filling in the shortcomings of the existing algorithm in the case of low sampling rate and coherent signal.