Modeling And Simulation Of Ground-based Array Radar Anti-jamming Based On Jamming Perception

With the development of advanced radar electronic countermeasure technology,the radar needs to know the jamming type imposed by the enemy efficiently and accurately,and adopt the anti-jamming method for specific jamming type.The jamming perception module can distinguish the jamming types by using the difference of the characteristics between different jamming,so as to identify the enemy jamming signals.In view of the increasingly complex electromagnetic environment,the training and optimization of jamming perception module is more practical.Aiming at the problem that the traditional radar system can not accurately detect the jamming signal type of the enemy,and can not take corresponding anti-jamming measures based on the specific jamming type,a ground-based array radar anti-jamming system based on jamming perception modules is built.The radar system can realize the identification of the enemy jamming signal through the well-trained jamming perception module,and adopt corresponding anti-jamming measures for the specific jamming signal type,which is more efficient and accurate than the conventional radar anti-jamming process.Firstly,the modeling and Simulation of multiple jamming signals and target echo signals are completed,and the corresponding features are extracted based on the time domain,entropy theory and gray level co-occurrence matrix texture parameters;Secondly,the original feature space is established by using the extracted features,and the training matrix is generated based on the original feature space to train three classifiers: support vector machine,random forest and BP neural network,that is,the jamming perception module is obtained to realize the recognition of jamming signals;Thirdly,SVM-RFE,ReliefF and Q-learning are used to generate the optimal feature subset based on the original feature space,and the training matrix is generated to train the optimal SVM classifier,and the trained optimal jamming perception module is obtained.The performance of the feature selection algorithm based on Q-Learning is optimized by comparison;Fourthly,the optimal jamming perception module and ground-based array radar anti-jamming system are combined to generate jamming signals to test the module and complete the subsequent anti-interference processing.The main idea of this thesis is the generation of the optimal jamming perception module.The main process includes jamming signal feature extraction,classifier training and feature selection optimization.Compared with other classifiers,the SVM classifier based on Qlearning algorithm that designed in this thesis improves the recognition rate of various jamming signals and requires less features.It can further improve the performance of jamming perception module by extracting better distinguishing features,optimizing classifier training algorithm and studying new feature selection optimization algorithm,so that jamming perception module has greater engineering value.