Research On Virtual-array Approach For Airborne Radar Forward-looking Super-Resolution Imaging

Airborne radar forward-looking(ARFL)imaging has increased great interest in fields of self-landing,supply dropping and terrain awareness.Currently,the ARFL images is obtained by scanning the real beam over the field of view.However,the resolution of real-beam radar images is constrained by the aperture size,and can not satisfy the resolution requirement for practical applications.The synthetic aperture radar(SAR)and Doppler beam sharpening(DBS)techniques can provide high resolution in side-looking and squint-looking direction,but they are incapable of ARFL imaging.The reason is that terrain points situated symmetrically about the flight path have the same Doppler history,and the gradient of the Doppler frequency is small in the ARFL flight direction.The super-resolution technique can achieve high angular resolution,by adjusting the scanning and waveform parameters and transcends the Rayleigh limit in a way of signal processing,which has been a cutting-edge technique for ARFL imaging.Aiming at the fundamental problems of imaging model,imaging performance and imaging approach,this dissertation emphatically develops researches on the virtual-array signal model,super-resolution performance bound and iterative adaptive approach.The main innovations include:1.This dissertation establishes the virtual-array parallel signal model for ARFL sequential echoes.This reveals that the physical nature of angular super-resolution lies in extending the virtual array in the direction of tangential track by jointly processing the azimuthal scanning sequential echoes.Therefore,the established virtual-array signal model breaks through the mechanism constraint of ARFL imaging,and provides theoretical basis for researches on super-resolution performance and approaches.2.This dissertation derives the super-resolution performance bound of ARFL imaging for the first time.This establishes the quantitative mathematical relationship between the radar system parameters,signal-to-noise ratio(SNR)and the super-resolution performance,which provides the rationale for ARFL super-resolution imaging system designing and performance improvement.3.This dissertation proposes an improved iterative adaptive approach,which solving the problem of noise amplification caused by eigenvalue spread when extending the virtual array.The proposed approach notably enhances the multiple-target angular resolution.4.This dissertation proposes a kind of frequency-domain fast iterative adaptive approach,which converts the matrix inversion into the operation of ciculant matrixvector fast multiplication.The proposed approach solves the conundrum of efficiently implementing the iterative adaptive approach,which can realize real-time ARFL super-resolution imaging.The above-mentioned signal model and super-resolution methods have been validated by simulations and real data processing.The results demonstrate that the proposed method can improve the angular resolution of ARFL images.Meanwhile,the proposed method can realize real-time signal processing,and can be utilized to fill in the dead zone of SAR and DBS techniques in ARFL direction.