MATLAB simulation of performance evaluation model of time-frequency analysis method based on SVM

Abstract

Time-frequency analysis is a prerequisite for intelligent recognition of radar signal types based on deep learning network. Deep learning uses Convolutional Neural Network (CNN) to automatically extract the time-frequency images (TFIs) features of radar signals to achieve intelligent recognition of radar signal modulation methods. However, the quality of TFI generated by different time-frequency conversions is usually different to a certain extent. At present, the selection of time-frequency analysis methods in many pieces of research is mainly done through the feature differences of TFIs. There are certain subjective factors, which cannot provide a reference for subsequent research. This paper proposes a time-frequency analysis method performance evaluation model based on Support Vector Machine (SVM). The simulation results show that the evaluation model can make an objective and optimal choice based on the data, and can provide a valuable reference for subsequent research.