Research On Adaptive Waveform And Game Waveform Design Algorithm For Cognitive Radar

With the continuous innovation of scientific theory and the vigorous development of technology,the electromagnetic environment faced by radar in practical combat has become increasingly complex,and in the face of the current combat situation,traditional radar systems have become somewhat difficult to meet the needs.In recent years,with the continuous breakthroughs in artificial intelligence technology,cognitive radar,a new system of radar systems,is gradually transforming from a concept to a reality.The cognitive radar system endows the radar with the ability to learn,analyze,and infer information about the external environment.Moreover,the radar can choose a reasonable combat strategy suitable for the current environment based on the mastered changes in the external environment,effectively improving the radar’s combat and survival ability to cope with complex environments.In modern battlefield environments,radar is often not limited to a single task and operating mode.In addition,the real confrontation between radar and clutter and interference cannot be ignored.In order to comprehensively improve the performance of cognitive radar in various aspects and effectively enhance its dynamic game adversarial ability in the face of intelligent interference,the main research content of this thesis is as follows:1.In the face of complex real-world combat environments,the prior information of radar may not be able to adapt to the current environment,making it difficult to meet performance needs.In practical combat,the detection performance of radar is being increasingly severely suppressed.A radar waveform design method based on CNNBi LSTM-A network is proposed to balance the multiple performance of electronic warfare radar.First,generate the waveform corresponding to the Signal to Noise Ratio(SCNR)criterion and mutual information(MI)criterion according to the prior environmental information,and construct a data set by one-to-one correspondence between the environmental information and the generated waveform,and then train the proposed neural network model using the training set separated from the data set,Finally,the trained neural network will be subjected to simulation experiments for performance testing and comparison.Simulation experiments have verified the effectiveness of the proposed method in balancing multi criteria performance and effectively improving radar comprehensive performance.2.In modern electronic warfare,the game between radar and jammers cannot be ignored.To improve the multifaceted performance of radar under game conditions,a radar game waveform design method based on weighted criteria is proposed.Firstly,a weighting criterion was established based on MI and Signal to Interference plus Noise Ratio(SINR).Subsequently,a corresponding radar and interference game model was designed.Finally,a maximum edge reassignment algorithm was proposed to solve the repeated game dilemma and refine the Nash equilibrium.