The method to suppress radar clutter by selecting a filter based on production rule using profile of clutter categorized by neural network

摘要

A method to improve the performance of radar clutter suppression is presented. Generally, a band rejection filter is used for the improvement of SCR (the ratio of signal to clutter). If a fixed band filter is used, the band rejection width and the center frequency of the filter do not match the spectrum of the clutter. A method to solve the mismatch by selecting a filter based on the production rule using the profile of the radar clutter categorized by the neural network is described. The authors' method is validated by computer simulation. A 10-dB improvement factor of SCR is obtained by the method as compared with the conventional pulsed Doppler radar applying MTI (moving target indicator) or adaptive MTI to reject the clutter leakage in coherent integration.