A novel recognition approach for radar emitter signals based on on-line independent support vector machines

Radar emitter signal recognition is one of the key procedures in signal processing of Electronic Intelligence. To enhance the ability of online recognition to meet the requirement of modern electronic warfare, A novel recognition approach for radar emitter signals based on on-line independent support vector machines is presented in this paper. Based on the cascade feature extraction of radar emitter signals, the on-line independent SVM is applied to effectively construct the basis vectors of SVM by incremental mode and make full use of the signal data information, then, an automatic online recognition system based on data driven are implemented. Experiment results shows that the method can achieve high accurate recognition rate even at lower SNR, and has good characteristics of rapid identification and less memory.