Research On Target Detection Technology For Frequency Diverse Array Radar

Different from conventional Phased Array(PA),which transmits angle-dependent beampattern,Frequency Diverse Array(FDA)radar produces angle-dependent,rangedependent and even time-dependent beampattern by imposing different frequency increments to each array element in turn.Therefore,FDA radar has broad application prospects in the fields of transmitting beamforming,clutter and interference suppression,target detection and tracking,parameter estimation,and radio frequency(RF)stealth.Based on the fundamental characteristics of the FDA radar structure and signal,this dissertation systematically studies the key technologies involved in target detection,including the target Radar Cross Section(RCS)of the FDA radar,the clutter model and its characteristics,square law detection for Independent Identically Distributed(IID)and Non-Independent Identically Distributed(Non-IID)fluctuating targets,and Generalized Likelihood Ratio Test(GLRT)based on the reconstruction of the interference covariance matrix.The main contributions of this dissertation are summarized as follows:(1)Analysis and modeling of FDA radar target RCS.Aiming at the modeling problem of FDA radar target RCS,the statistical distribution model of the FDA radar target RCS is derived in this dissertation based on the principle of electromagnetic propagation and the central limit theorem.This model can more accurately describe distribution characteristics and correlation characteristics of the FDA radar target RCS in time and frequency dimension,solving the problem of mismatch with the traditional model effectively.In addition,the de-correlation effect of frequency increment on the relative RCS amplitude of FDA radar is analyzed,and the principle of frequency increment configuration in different application scenarios is discussed according to the frequency de-correlation criterion.The research on statistical model of target RCS lays a theoretical foundation for FDA radar clutter modeling and fluctuating targets detection.(2)Analysis and modeling of FDA radar clutter.For the range-ambiguous clutter problem of airborne radar,a pulse FDA radar clutter signal model based on the multichannel matched filtering(MCMF)receiving structure is proposed in this dissertation.This model is suitable for both coherent FDA and FDA-MIMO(Multi-Input and MultiOutput)radar.On the basis of clutter signal model,the clutter characteristics of the FDA is analyzed in this dissertation,which mainly includes the space-time power spectrum characteristics and eigenspectra characteristics of the clutter.Moreover,this dissertation utilizes the range-dependent beampattern of the FDA radar to distinguish the rangeambiguous clutter of the airborne forward-looking array radar,and analyzes the suppression performance of FDA radar on the range-ambiguous clutter.Compared with the corresponding indicators of conventional PA and MIMO radar,the result shows the application superiority of FDA-MIMO radar in range-ambiguous clutter suppression.(3)FDA radar fluctuating targets detection based on incoherent square-law detector.In view of the detection problem of IID fluctuating targets,this dissertation derives the square-law detection model of pulse FDA-MIMO radar,analyzes its theoretical detection performance,and compares it with coherent FDA,conventional PA and MIMO radar.Furthermore,for the detection of Non-IID fluctuating targets,the existing RCS models are difficult to accurately characterize the electromagnetic scattering characteristics of targets due to the influence of frequency increments and time variation of beampattern on the statistical characteristics of FDA RCS.Therefore,this dissertation respectively derives the FDA-MIMO radar incoherent square-law detection models of correlated identically distributed and independent non-identically distributed fluctuating targets RCS models.The analysis of their asymptotic detection performance and the upper bound of the truncation error shows that FDA-MIMO radar has the application advantages in Non-IID fluctuating targets detection.(4)FDA radar target detection based on interference covariance matrix reconstruction.Aiming at the problem of target detection with unknown interference covariance matrix,this dissertation proposes an algorithm about FDA radar target detection based on the reconstruction of interference covariance matrix,considering the lack of additional priori knowledge for parameter estimation of interference signals in practical applications.When the number of data samples is small or the training samples contain both target and interference signals,the sampling covariance matrix of interference will be discrepant with the real scenario,which will lead to a serious degradation in target detection performance.This algorithm combines FDA with MIMO radar,and uses Estimating Signal Parameters via Rotational Invariance Techniques(ESPRIT)to estimate the azimuth and range parameters of the interference signals step by step,thereby reconstructing the highprecision interference covariance matrix independent of the target signal.Based on the reconstructed covariance matrix,the asymptotic performance of FDA-MIMO GLRT detector is derived.Besides,this dissertation also analyzes the computational complexity of the reconstruction algorithm,the theoretical minimum mean square error,the Cramér–Rao bound of parameter estimation,and the asymptotic detection performance.