Antenna scanning type classification in electronic warfare support systems

In this study, antenna scanning types used in Electronic Warfare Support Systems were determined by using the feature vectors obtained by different feature extraction methods in supervised and unsupervised learning methods. The amplitude patterns of the separated radar pulses feed into the patterns we propose and the corresponding scan type is automatically obtained. The feature vector that provides input to supervised and unsupervised learning models was created using different methods. Experiments on synthetic datasets show that the proposed feature extraction technique with Long-Short Term Memory based Autoencoder gives the best results for the problem under consideration.