Modeling And Simulation Of Digital Array Radar Intelligent Anti-jamming Based On Jamming Perception

With the rapid development of electromagnetic interference technology,more and more new kinds of interference appear in the modern war,which seriously damage the radar’s ability to detect and identify targets,and pose a great threat to the survival of radar in the battlefield.This threat forces radars to seek more advanced methods to improve their ability to perceive the external environment and to suppress electronic interference.With the rapid development of machine learning and reinforcement learning,intelligent anti-jamming has become an important development direction of radar.In this thesis,the overall scheme of intelligent anti-jamming radar simulation system is designed in detail,its overall framework and workflow are studied,and its key components are briefly introduced.Secondly,in view of the problem that traditional radar can not effectively identify the specific types of enemy jamming,the intelligent jamming sensing technology is studied.This thesis uses the differences in the performance of different types of active interference features for interference identification,carries out joint feature extraction in multiple domains such as the time domain and frequency domain,and constructs a feature space to obtain a sample feature matrix and a sample label matrix.Using the above data,an active jamming classifier based on BP neural network and support vector machine is designed and trained to establish the mapping relationship between jamming features and jamming types to achieve accurate identification of active jamming.Aiming at the actual combat scenario of cooperative jamming of multiple jammers,an intelligent jamming sensing method based on nulling and shape keeping multi-beam is studied.Experiments show that the method can effectively identify the interference with lower interference noise ratio in the combined interference,and establish the corresponding relationship between the interference type and the angle,which provides important prior information for the next intelligent anti-interference decision.Then,the interference parameters measurement and active-passive multi-domain anti-jamming methods of digital array radars are modelled and simulated.Aiming at the problem that traditional radar cannot select the optimal anti-jamming method based on the interference scene,this thesis studies the intelligent anti-jamming decision method based on Q-learning in conjunction with the interference perception results.The method takes the state transition of the jammer as the feedback reward,obtains the iteratively convergent value function through training,and establishes the mapping relationship between the jammer state and the optimal anti-jamming measure.Finally,the software development and interactive interface design of the radar simulation system are completed.On the basis of constructing typical jamming scenarios,its key functions are tested.The test results verify the effectiveness and correctness of the jamming perception and intelligent anti-jamming modules.Based on a scientific research project of the group,this thesis designs and builds an intelligent anti-jamming radar simulation system,and uses methods such as machine learning and reinforcement learning to study intelligent jamming perception and intelligent anti-jamming decision-making techniques,which improve the jamming recognition capability and anti-jamming decision-making capability of radars,with certain engineering significance and reference value.