Wavelet and independent component analysis of Doppler spectrum for higher validity of sea sensing by HF radar

Doppler spectrum (DS), comprising two first order peaks (Bragg peaks) and a second order continuum, is generated by twice Fourier transform on the radar data received by a high frequency surface wave radar to monitor the ocean surface state. Oceanic parameters i.e. wave, current, and wind, can be estimated from a valid DS by inversion. The primary task in the inversion algorithms is to identify the position of first order peaks correctly. Otherwise, the DS will either be taken as invalid or misused. Signals from other sources, referred as clutters, sometimes mix into sea echoes and are confused with Bragg peaks. Based on the different micro-Doppler feature, a combination of wavelet and independent component analysis is thus recommended to distinguish real out of pseudo peaks.