Class identification of aircraft by means of artificial neural networks trained with simulated radar signatures

Abstract

Non-Cooperative Target Identification (NCTI) of aircraft from radar measurements is a formidable problem that has drawn the attention of engineers and scientists over the last years. NCTI techniques typically involve a database with a huge amount of information from different known targets and a reliable identification algorithm able to highlight the likeness between measured and stored data. This paper uses High Resolution Range Profiles produced with a high-frequency software tool to train Artificial Neural Networks for distinguishing between different classes of aircraft. Actual data from the ORFEO measurement campaign are used to assess the performance of the trained networks.