Gabor-Atom networks based radar target identification

A Gabor-Atom network (GAN) approach for application in radar target recognition is proposed. The Gabor atoms selected by a multilayer feedforward neural network extract discriminant features among different classes of radar target returns. The self-learning mechanism is used not only for the network but for the feature parameters. Results on the classification of microwave anechoic chamber data of three different scaled airplane models are presented.