Analysis of measured radar data for Specific Emitter Identification
Mariette Conning, Ferdie Potgieter · 2010
Measured radar data assisted in the successful development and implementation of Specific Emitter Identification (SEI) signal processing algorithms. The aim of the algorithm is the identification of a specific emitter within a single class of emitters. The processes developed are pulse extraction, feature calculation, dimensionality reduction and classification. A pulse is detected whenever the phase changes from being random to being linear. Time domain features are then calculated from these extracted pulses to populate a feature vector. Some redundancy exists in the feature vector and a process of dimensionality reduction is followed to obtain the optimum set of features. A Fuzzy ARTMAP classifier completes the process of specific emitter identification to demonstrate the improvement in Correct Classification Decisions (CCD) using the reduced feature vector.