Neural Network-Guided Sparse Recovery for Interrupted-Sampling Repeater Jamming Suppression
Algorithm 1
The proposed method.
Input: The echo signal , the sliding window , the trained GRU-Net, the dictionary D.
Output: Target locating result rt, the reconstructed target echo signal
Begin
(1)
Perform pulse compression of ;
(2)
Implement STFT to transform the PC result into TF domain;
(3)
Normalize the TF data;
(4)
Input the normalized TF data into the GRU-Net;
(5)
Obtain the GRU-Net target locating result rt;
(6)
Carry out the sparse representation for ;
(7)
Construct an adaptive filter Filter according to rt to filter out the corresponding peaks of the false target groups in the sparse representation domain;
(8)
Reconstruct the target echo signal through D and amplitude compensation to obtain the jamming-free result;