Automatic Radar Target Identification Using Radar Cross Section Fluctuations and Recurrent Neural Networks

Automatic radar target identification (ARTI) using recurrent neural networks (RNN) is being proposed in this work. Illumination of targets with transmitted electromagnetic waves from radar results in production of surface currents causing scattering of the incident energy from the target. The back scattered signal carries much information about the target which can be utilised for solving the target identification problem. The radar cross section (RCS) has been used as a feature in this work for target identification. The dependence of RCS on orientation of the target with respect to radar with a mono-static configuration has been exploited here. The dataset contains a sequence of RCS values for varying aspect angles which simulates the target. motion. This dataset is fed to the recurrent neural network (RNN)/long short-term memory (LSTM) model for classification. The classification accuracy of 98 percent was achieved using the RNN/LSTM model.