Study On KA-STAP Of The Airborne Cognitive Radar

Cognitive radars are knowledge-aided cycle radar systems which are able to adjust the transmit signal and receiver signal processing algorithms according to the real-time environmental obtained information,a priori knowledge as well as the task features.By introducing the idea of cognitive radar into the air-borne radar system,the detection performance of the traditional space-time adaptive processing algorithms can be improved.Knowledge-aided space-time adaptive processing(KA-STAP)is one of the critical functions in a cognitive air-borne radar system.This work studies the air-borne radar KA-STAP algorithms,and the main contributions of this paper are as follows:(1)Air-borne radar's clutter modeling studyModeling of the environment clutter is a crucial auxiliary work to the study of airborne radars.This work introduces the heterogenous clutter model of air-borne radars based on the integrating clutter model.Aiming at the problem that the fitting performance of the existing sea clutter chaotic modeling methods is not stable enough,this work studies the difference between the real measured sea clutter and chaotic signals,and develops a sea clutter decomposition(SCD)model.Then a sea clutter constituent synthesis(SCCS)approach is proposed based on the SCD model.Comparing with the existing sea clutter chaotic modeling methods,the SCCS method has the ability to fit sea clutter series under different sea states,and provides a new angle to the sea clutter chaotic modeling problem.(2)Study the taper matrix model and its improvement strategyThe KA-STAP algorithm in air-borne cognitive radar systems uses the a priori clutter covariance matrix to modify the sample estimated covariance matrix in order to enhance the space-time filter's performance under heterogenous clutter environment.As a component of the a priori clutter covariance matrix,a taper matrix is utilized to adjust the a priori two-dimensional clutter spectrum in the angel-doppler domain.The value of the taper matrix can influence the filter notch directly.The widely used CMT(covariance matrix taper)model has only one adjustable parameter,which makes it lacking in the optimization degree of freedom and not suitable to the sea clutter fitting problems.Based on the CMT model,this paper proposes a generalized CMT model(GCMT).Aiming at the characteristics of the sea clutter spectrum,the GCMT model introduces a band-pass distributed phase dithered steering vector random process to adjust the spectrum width at the middle and high doppler frequency area.The undetermined parameter changes from the width parameter of the singer function to multiple weighting coefficients.Comparing to the CMT model,the GCMT model increases the optimization degree of freedom of the taper matrix model,and enhances the fitting ability of the a priori covariance matrix.(3)Taper matrix parameter adaptively optimization approach studyThis work develops two adaptively optimization approaches for the parametric taper matrix model.The first one is based on the existing PW colored-loading factor optimization approach.By involving the colored-loading matrix pre-whitenging ability maximizing optimization strategy,a taper matrix parameter adaptively optimization approach(TPO)based on a priori covariance matrix pre-whitening ability maximization is developed.Considering the feature of the taper matrix optimization problem,a pre-whitening ability evaluation quantity is newly proposed and utilized in the TPO optimization function designation.The TPO approach is applicable to both the space domain and time domain taper matrix optimization problems.The second approach(MS algorithm)is proposed by estimating the space domain minimum variance spectrum with the help of space domain snap obtained in both the slow time dimension and the fast time dimension.Then the space domain taper matrix parameter optimization function is designed to maximize the spectrum similarity between the a priori data and the real measured data.Since the real-time echo signal is introduced into the a priori information optimization,these two approaches enhance the adaptability of the knowledge-aided clutter suppression algorithm.(4)Robustness designing of the colored-loading factor optimizationThe colored-loading factor in the KA-STAP weighting vector describes the emphasis we put onto the a priori information.The value of it should be determined by the accuracy degree of the a priori information as well as the fitting performance of the sample estimated covariance matrix.The existing PW colored-loading factor optimization approaches cannot evaluate the performance of the CUT a priori information and are vulnerable to the a priori information heterogeneity among range domain.Aiming at the robustness lacking problem of the PW approach,this work develops an improved algorithm(CPW),which can evaluate the accuracy degree of the a priori information of the CUT,and obtain the optimized colored-loading factor for each under test cell independently.Combined with the rank-reduced KA-STAP,CPW is robust to the a priori information heterogenous in both the range domain and doppler domain.(5)Study the application of low-rank matrix completion algorithms in KA-STAPThe CUT echo signal is usually not utilized in KA-STAP methods due to the possibly existing target signal.A target signal elimination approach is proposed in this work to obtain the real-time clutter information of the CUT.Since the space-time domain spectrum of the air-borne radar clutter is a low-rank matrix,and the spot target signal only spread in few channels within the angle-doppler domain,this work introduces the low rank matrix completion algorithm into the CUT clutter recovery problem,and develops a CUT target signal elimination approach realized in the transmit domain.As an example,the cleaned CUT echo signal is used to optimize the undetermined parameter in the taper matrix.Utilization of the recovered CUT clutter signal reliefs the dependency of the KA-STAP to secondary samples,thus increases the robustness of the KA-STAP to the failure of the sample selected algorithms.(6)KA-STAP effectiveness test based on measured dataVerify the effectiveness of KA-STAP algorithms on air-borne phased array radar real measured data.Develops a fullly-adaptive KA-STAP flow for measured data processing,including measured data based environment information extraction,a priori clutter covariance matrix optimization,and colored-loading factor optimization.Test results show that,the KA-STAP filter outperforms the traditional STAP when the sample support is insufficient,and the effectiveness of the algorithms discussed in this work is demonstrated.