Research On Active Interference Separation And Intelligent Waveform Design Method For Bulletborne Radar

The seeker will face a complex electromagnetic environment for the missile penetration process on the modern battlefield.Currently,new radar jamming methods spring up,and various jamming methods are complex and changeable,which poses a significant challenge to the normal detection,tracking and identification of radar.The target self-screening jamming,the stand-off jamming around the target,clutter and noise often exist simultaneously,forming a complex jamming environment.In such an environment,in order to effectively achieve anti-jamming,the radar should necessarily cognize the environment.In recent years,the concept of "cognitive radar" has attracted much attention.The key technology that needs to be broken through is the radar's cognition of the environment.The radar can adaptively transmit the next anti-jamming waveform on the premise of fully cognizing the environment.The radar's cognition of the environment involves detecting and recognizing jamming and the interference separation.After decades of development,the theory of jamming detection and recognition has become mature.The thesis focuses on the missile-borne radar platform and studies the interference separation and intelligent waveform design.The main contents are as follows:1.The active radar jamming scenarios faced by the missile-borne radar in flight were analysed.When there are multiple jamming signals in different directions in the space,the jamming is separated in the spatial domain.The difficulty of DOA estimation of strong coherent jamming in space was discussed,and the adaptive space separation technology based on the maximum SINR criterion,MMSE criterion and LCMV criterion was analysed.The changes in similarity of the target signal before and after interference separation under different signal-to-noise ratios through simulation was obtained.2.Conv-Tas Net-based self-screening jamming separation is proposed to address the jamming separation problem that arose from the combination of the target echo and selfscreening jamming.The influence of network parameters on the network model size,training time and separation performance is discussed.The commonly-seen active deception radar jamming and echo talker were mixed according to the jam-to-signal ratio of 0-10 d B as the input data of the network,and meanwhile,appropriate network parameters were selected to train and verify it.Eventually,the simulation verified the efficient separation performance of the method.3.The production model and features of the commonly-seen repeater pulse jamming were analysed.Constrained by the transmitted waveform energy condition,an intelligent anti-jamming waveform design based on the maximum SINR criterion was proposed,which combined the standard clutter model to make the algorithm application scenario closer to the actual situation.Through simulation experiments,it could be found that the echo signal-to-jam ratio of most of the transmitted waveforms designed for jamming can be improved above 15 d B.Thus,the method's effectiveness against partial relayed shift jamming was verified.4.The DQN network was connected with the cognitive radar,and the flight state change scene of the missile-borne radar was established,and accordingly,the intelligent anti-repeater deception jamming waveform design based on the DQN network was proposed.This end-to-end waveform design method can better fit the cognitive radar.In the meantime,the simulation verified that the waveform designed based on the DQN network has the ability to resist jamming effectively.For common repeater jamming,the echo signal-to-jam ratio of the designed waveform was improved above 15dB.