Application of neural networks to radar signal detection in K-distributed clutter

The radar signal detection is a very complex task, which is generally based on conventional statistical methods. These methods require a lot of computing and they are optimal only for one type of clutter distribution. Recently, artificial neural networks (ANN) have been used as a means of signal detection. In this paper, we consider the problem of radar signal detection using ANN in a K-distributed environment. Two training algorithms are tested; namely, the back propagation (BP) and genetic algorithms (AG) for a MLP architecture. The simulation results have shown that the MLP architecture outperforms the classical CA-CFAR detector.