Study On Key Technologies Of Cognitive ISAR

The signal processing of traditional radar systems is mainly concentrated on the receiver,which lacks the transmitting-receiving feedback mechanism and makes it difficult to adaptively perceive and interact with the working environment.As a new type of radar system,cognitive radar constructs a closed-loop feedback structure which includes the transmitter,the receiver,the target,and the environment.This structure makes full use of the receiver to dynamically sense the target and the environment,and then designs the operating parameters of the transmitter according to their feedback,which can improve the performance in all aspects.The combination of cognitive theory and inverse synthetic aperture radar(ISAR)imaging theory sheds lights on the improvement of ISAR imaging performance under complex environments.This thesis focuses on the key issues and difficulties in cognitive ISAR,which mainly include waveform design,high-resolution imaging,and radar resource scheduling.In waveform design,high-resolution imaging methods for linear-frequency-modulated waveforms,stepped-frequency waveforms,sparse stepped-frequency waveforms and stepped-frequency chirp waveforms are systematically studied.Furthermore,aiming at the complex observation environments,waveform optimization methods based on the maximum mutual information criterion and the joint criterion are studied,and high-resolution imaging methods of the optimized waveforms are designed.In high-resolution ISAR imaging under low signal-to-noise ratio(SNR),the sparse observation model and the corresponding Bayesian probabilistic graphical model are constructed for the sparse stepped-frequency chirp waveform,and the imaging and motion parameter estimation method based on Bayesian nonparametrics is proposed for maneuvering targets.In radar resource scheduling,a new resource scheduling evaluation index is proposed,and a resource scheduling model based on cognitive ISAR imaging is established according to the time and energy constraints.Then,the genetic algorithm is applied to find the optimal solution.The related work of this thesis will provide theoretical and technical supports for the improvement of the space target detection and imaging capabilities of ISAR,and will intelligentize and systemize the development of radar imaging technologies.The main content of this thesis can be summarized as follows:The first part studies frequency modulated waveforms and the corresponding high-resolution imaging methods.First,echo models of the linear-frequency-modulated waveforms,stepped-frequency waveforms,sparse stepped-frequency waveforms and stepped-frequency chirp waveforms are established,respectively.Then,high-resolution range profile synthesis methods of the four waveforms are studied,and two-dimensional high-resolution ISAR imaging is discussed.Finally,the effectiveness of the related methods is proved by simulations.The second part studies cognitive ISAR waveform optimization and high-resolution imaging.Firstly,the waveform optimization models based on the maximum mutual information criterion and the joint criterion are established,respectively,based on which the model solving methods are proposed.Then,to deal with the mainlobe broadening and high sidelobes in ISAR imaging of the optimized waveform,sparse modeling is performed of the optimized waveforms,and sparse signal reconstruction methods such as the orthogonal matching pursuit(OMP)are applied to obtain well-focused high-resolution imaging.Finally,the effectiveness of the proposed method is verified by simulation.The third part studies high-resolution ISAR imaging of maneuvering targets under low SNR based on the sparse stepped-frequency chirp waveform.Firstly,the signal model of sparse stepped-frequency chirp waveforms is established.Then,the target motion parameter estimation is combined with the high-resolution range profile synthesis by constructing a parameterized dictionary.After that,a motion compensation and high-resolution ISAR imaging method based on the genetic algorithm and Bayesian nonparametric prior is proposed.Finally,Monte Carlo experiments and imaging results of measured data prove the validity of the proposed method.The fourth part studies resource scheduling method of cognitive ISAR.Firstly,a new resource scheduling evaluation index is proposed.Then,according to the resources required by each imaging task and the time and energy constraints in resource scheduling,a new cognitive ISAR resource scheduling model is constructed,from which the optimal scheduling is solved by genetic algorithm.Finally,simulation results verify the effectiveness of the proposed method.