Radar Signal Sorting System Based on Multi-domain Feature Extraction and Decision Tree Algorithm

With the rapid development of radar technology, radar signal types are becoming more and more complex and changeable. Moreover, the real-time and accuracy of traditional radar sorting systems are facing increasingly severe challenges. This paper proposes a radar classification method based on the classification and regression tree (CART). After multi-domain feature extraction of the radar signal, the decision tree model is trained and verified by the new data according to the Gini impurity minimum criterion. Finally, the radar signal can be effectively recognized. The simulation results demonstrate that the proposed radar signal sorting system's recognition accuracy can reach 98.9%, which is 18.7% higher than that of the decision tree model without feature extraction.