Efficient radar target classification using modular neural networks
Jiang Wenli, Zhang Huoju, Lu Qizhong, Zhou Yiyu · 2002
A radar target classifier based on modular neural networks is presented and its performance compared with that of a classifier based on non-modular neural networks. In this classifier, the response from an unknown target is sent to several waveform predictors that are BP neural networks trained by responses from known targets. The predictor errors are then sent to a classifier using the rule of maximum a posteriori or the rule of modified minimum squared errors. The simulation shows that the new classifier has a good performance on radar target recognition. The method also has other advantages such as easy realization, clear structure and easy expansion.