Doppler Radar-Based Target Localization Algorithm Combining Adaptive Linear Chirplet Transform and Frequency Compensation

When using Doppler through-wall radar (TWR) for target localization, it is necessary to extract the target's instantaneous frequency (IF) from the time-frequency distribution (TFD) of the echoes. However, the TFD of the echo often has time-frequency aliasing, such as intersecting or overlapping time-frequency ridges, thereby rendering the extraction of the IF a highly challenging task. This paper presents a target localization algorithm that combines adaptive linear chirplet transform (ALCT) and frequency compensation, which can accurately extract the IF curve of target signal. This method first enhances the time-frequency energy concentration of the tracking target through ALCT and extracts its IF from the enhanced TFD. Then, the target's IF value is compensated using the redundant Fourier model, resulting in obtaining the corrected IF curve. Finally, the target trajectory is synthesized based on the corrected IF curve, enabling precise target positioning. The experimental results show that the proposed method not only enhances the time-frequency energy concentration of the target, but also accurately extracts the target's IF curve, especially for signals with time-frequency aliasing, compared with the traditional localization methods.