Research On Key Technology Of Array Signal Processing For Low Altitude Target Detection Radar

There are problems of high false alarm rate and being difficult for tracking for low-altitude target detection due to the influence of strong clutter such as ground objects.Spatial signals are received by sensor array in array signal processing,which enjoys the advantages of flexible beam steering and strong interference suppression ability.Array signal processing has been widely used in radar,wireless communication,sonar,radio astronomy,and so on.With the development in decades,the basic theoretical framework and algorithms of array signal processing have been very mature.However,there are still many problems to be solved.Various beamforming techniques under robustness constraints and target tracking techniques at low elevations in array signal processing are studied in this thesis.The research content include transmit beam control（TBC）under the constraints of unimodular or low Peak-to-Average Power Ratio（PAPR）of weighting vecotors,robust transmit adaptive beamforming（RTAB）with the consideration of steering error,direction of arrival（DOA）estimation for non orthogonal waveform Multi-Input Multi-Output（MIMO）radar,and low elevation tracking algorithm based on DOA.The main work is summarized as follows.1.TBC with weighted vector constant modulus or low PAPRThe radiation energy in unneccessary areas with strong reflection can be reduced by TBC effectively,while the effective radiation power of the antenna would be wasted.To solve this problem,the performance of the traditional TBC algorithm has been analyzed firstly.The minimizing of the ₂ norm of the weighting vector is taken to optimize the robustness of the beamforming.Secondly,the objective function is established with the constraints on the energy in the ground side of the beam.The second-order cone programming has been used to solve the opitimzing problem for TBC.Thirdly,the amplitude normalization has been then applied to the optimal vector.The beam can be formed with the ability to suprpress clutter as well as the full use of the antenna radiation power.Afterwards,the constant modulus constraint has been relaxed and replaced by Low PAPR constraint.The TBC algorithm for low altitude detection under low PAPR constraints has been presented,which improves the performance of the sidelobe level in the ground side and interference suppression.2.RATB in the presence of steering vector errorsSteering vector erros would lead to errors in the array signal processing results,such as the deviation of the mainlobe direction from the ideal direction.Against this problem,the analysis on the robustness of beam has been given firstly.The maximum value of the deviation of the mainlobe direction with steering vector erros has been derived secondly.And then the tolerable steering vector error value in the worst case is derived.The beamforming algorithm based on the analysis of the mainlobe error has been proposed thirdly.The proposed method can effectively solve the problem of degraded performance of beamforming algorithms in the presence of errors in the steering vector.It is considered that existing RTAB algorithms suffers from robustness problems,such as expected signal cancellation and high sidelobe level,as well as significant reductions in antenna effective radiation power due to the amplitude weighting of the weighting vector.To this end,the beam robustness has been optimized,with the constraint of low PAPR.At the same time,the radiation energy in strongly reflective areas has also been limited.Finally,the low PAPR RTAB algorithm has been proposed.With certain errors in the steering vector,the proposed method can suppress the radiation energy in the strongly reflective region,enjoy interference suppression capabilities,and improve the utilization rate of transmission power.3.DOA estimation for non-orthogonal waveform monostatic MIMO radarThe DOA performance in monostatic multiple-input multiple-output（MIMO）radar would be degraded duo to the non-orthogonal waveforms.To solve this problem,the echo signal model of the non-orthogonal waveforms MIMO radar is presented firstly.Then,the problem of fast DOA estimation for non-orthogonal waveforms monostatic MIMO radars is discussed.The noise subspace is estimated through generalized eigendecomposition.And then the DOA estimation is obtained through multiple signal classification algorithms.Subsequently,the joint estimation of DOA and Doppler for nonorthogonal waveforms monostatic MIMO radar is studied.The Khatri-Rao product property of the left singular matrix and the Vandermonde structure of the measurement signal matrix are utilized for DOA estimation.Then,using the properties of Kronecker product,the analytical expressions of the receiving flow pattern matrix,the transmitting flow pattern matrix corrupted by the waveform correlation matrix,and the Doppler flow pattern matrix are derived successively.Finally,the closed form solution of the Doppler estimation is obtained using the least square principle.The prior information about the waveform correlation matrix is unnecessary.There is no search or iteration in the estimation process.Moreover,DOA estimation results and Doppler estimation results can be paired automatically.4.Target tracking at low elevationIn low altitude environment,due to the influence of multipath effect on target signal echo,there is a large error in target angle estimation.A general signal model for target detection at low elevation is still absent.To solve this problem,firstly,a multipath signal echo model is established.The characteristics and signal composition of multipath echo are analyzed.The relationship model between target height and horizontal distance and array received signal is established.Then forward/backward spatial smoothing algorithm is used for decoherence.Finally,the estimation accuracy of target height is improved through the high precision DOA estimation.