Skip to Main Content
A data extrapolation method is applied to improve the performance of a target recognition scheme based on the central moments of one-dimensional range profiles. We adopt the autoregressive (AR) model to extrapolate the radar cross section data of a target in order to expand the measured data window. It is shown that the resulting high resolution range profiles can enhance target recognition capability, providing a more accurate recognition performance than the multiple signal classification range profile for moderate signal-to-noise ratio SNR ranges. Furthermore, the effects of the AR model-based data extrapolation on target recognition accuracy are carefully analyzed and investigated.