The interference classification and recognition based on SF-SVM algorithm
Guisheng Wang, Qinghua Ren, Yuze Su · 2017
For the classification and recognition of interference signals, a interference classification and recognition algorithm is proposed based on the signal feature space and SVM. The basic principle of the proposed algorithm includes that: the signal feature space is established by feature extraction for the interference signals with the signal models expressions and signal space theory. The classification and recognition algorithm is proposed for binary and multi-class classification based on the signal feature space and SVM (SF-SVM). Simulation experiments and analysis show that SF-SVM can lead to better performance in speed of training and classification accuracy, and it can get 2 dB performance gains compared to the traditional algorithm. The advantage performances for interference classification and recognition can be produced with the proposed algorithm.