Discrete Wavelet Design for Target Classification in Pulse-Doppler Surveillance Radar

Wavelets are powerful signal classification tools that show promise for classification in Pulse-Doppler radar applications. A wavelet decomposition of match filtered radar data is performed for a surveillance radar in the Czech Republic. A custom discrete wavelet is designed based on the range profiles of the dataset with and without targets and clutter present. The entropy and energy for different levels of the DWT decomposition are used to determine the presence of various targets. Targets are differentiated from static clutter by thresholding the DWT decomposition levels.