Simulation of Semi-Supervised Recognition Algorithm for ILS Communication and Navigation Signal Modulation
The reception and recognition of ILS communication signals have important practical significance,but with the increasing complexity and diversity of spatial electromagnetic fields,traditional template matching algorithms are unable to effectively identify distorted radar signals.To improve the stability and accuracy of signal recognition,an intelligent recognition model for SCLS-SVM modulated signals is constructed based on an improved dual feature ex-traction fusion algorithm.Firstly,the CWD algorithm was used to modulate five kinds of single signals and three kinds of combined signals,and the image processing method was used to enhance the feature extraction effect;Then the time-frequency feature of the modulated signal was extracted by the improved D-CNN,and the time-domain information feature of the modulated signal was extracted by the optimized SA-LSTM algorithm;Next the data of feature fusion was divided into training set and test set,and the input of signal recognition was established.Finally,based on the support vector machine classifier,the SCLS-SVM signal recognition model was constructed by using the ten-fold cross-validation optimization method.The ablation experiment results show that the superposition of SLS,DCN and CVO optimization strategies is helpful to improve the performance of the signal recognition model;the comparative ex-periment results show that on the MSD data set,compared with other baseline recognition algorithms,the P and R e-valuation indexes of SCLS-SVM algorithm are increased by 6.75%and 9.97%on average,and it has the highest sta-bility and accuracy.To sum up,the SCLS-SVM algorithm improves the identifiability of modulated signals by impro-ving the fusion algorithm of double feature extraction,and the SVM algorithm increases the classification and recognition effect of features,which has important research value in the field of ILS communication signal simulation and recognition.
Communication signal modulationFeature extractionSupport vector machine
万方数据
维普
