Specific radar emitter recognition based on wavelet packet transform and probabilistic SVM

Radar emitter recognition plays an important role in military automated command and control system. It is a composite task that involves radar signal interception, modulation recognition, features extraction and classification. In this paper, a wavelet packet transform is used for feature extraction of unintentional modulation on pulse (UMOP). Then specific radar emitter recognition is achieved by a probabilistic SVM (PSVM). Experiments on simulation and actual radar data verify the correctness and validity of this method.