A performance evaluation of autoregressive clutter mitigation methods for over-the-horizon radar

Wide-area surveillance over-the-horizon radar (OTHR) requires minimizing the dwell time in each search region without compromising Doppler resolution of targets from clutter. Low-order autoregressive (AR) modeling of the clutter spectrum has been proposed to unmask targets from clutter using shorter coherent integration times (CITs). We compare the performance of three methods using AR parameters to maximize signal-to-clutter-to-noise ratio (SCNR) with shorter CITs. In particular, we consider AR-based data extrapolation (DATEX), minimum variance clutter pre-whitening (MV-CPW), and a method called HOPPLER, which consists of pre-whitening, conventional Doppler processing, and re-insertion of the clutter spectrum. We compare the methods using real data from a mid-latitude OTHR with a target beacon embedded in sea-clutter.