Artificial neural networks for ESM emitter identification-an initial study

Over the past year research has been conducted in the Electronic Support Measures (ESM) Department at the Admiralty Research Establishment to assess the suitability of an artificial neural network (ANN) technique for ESM radar emitter identification. This is part of a larger study whose objective is to investigate new techniques to improve current ESM processing algorithms. Because of an ANN´s ability to be fault tolerant, generalise about class features and perform complex pattern classification, it is a possible solution to improving identification in the cases above. ANN´s immense parallelism also provide the potential for real time data computation which is important for emitter identification