Aerial target classification by micro-Doppler signatures and bicoherence-based features

The possibility of aerial target classification by extraction of micro-Doppler contributions contained in radar returns is studied. Novel classification features based on time-frequency distribution estimation are proposed in order to increase the aspect-independence property of classifier. Classification probability rates are computed for three different types of aerial targets including helicopter AH-64, aircrafts An-26 and B-52. The benefits achieved by using the proposed classification features are demonstrated and discussed.