Study On Some Issues Of Low-angle Altitude Measurement For VHF Array Radar

Radar is confronted with the threat of stealth aircraft and anti-radiation missile.Very high frequency(VHF)radar attracts extensive attention because of the advantages and potential in anti-stealth and anti-arm.Because the wavelength of VHF radar is long,VHF radar limits by short aperture and has bad angular resolution,which leads to multipath phenomenon.Multipath echo and direct echo are coherent sources,which are in the same distance cell,have the same Doppler frequency and are in the closed elevation angle.It is hard to distinguish the multipath and direct signal from spatial space,Doppler space or distance space.Especially in complex terrain,the reflection surface is time-varied,unknown and complex.The echo includes complex multipath signal and clutter,which degrades the performance of the altitude measurement of low angle target for VHF radar.Only if the performance of the altitude measurement of low angle target for VHF radar is improved,the advantage and potential of the VHF radar in anti-stealth and anti-arm will be made full use of.With the VHF radar as the background,on the basis of the achievements obtained by now,this dissertation starts from the signal model of altitude measurement of low angle targets,mainly analyzes the direction of arrival(DOA)estimation method of spatially adjacent coherent sources,inversion method for terrain parameter,altitude measurement method of low angle target in complex terrain and the utilization of polarization information.The main content of this dissertation is summarized as follows.1.The signal model of altitude measurement for low angle target is studied.To solve the problem of model mismatch for multipath,firstly,three progressive layers of the signal model of altitude measurement for low angle target are established: classical model,distributed-source model and perturbation model.In the classical model,the multipath is modeled as the correlated source with direct signal.The determinant factor of the complex reflection coefficient and the relationship between the radar,target and the reflection surface are analyzed.In distributed source model,the multipath is modeled as correlated distributed source,which emphasizes the expansibility in the spatial domain.The perturbation model is proposed to describe the perturbation of the multipath by complex terrain.Secondly,three altitude measurement techniques of low angle target are studied,which are spatial smoothing super-resolution method,synthesized vector maximum likelihood method and method based on compressive sensing.The advantages and disadvantages of the three methods are discussed.Lastly,the Cramer-Rao Bound(CRB)of the DOA for low angle targets are derived and the CRB are compared with the performance of the DOA estimation method.Computer simulation results indicate that the change of phase of the fading coefficient causes the fluctuation of the CRB and performance of the DOA estimation method.The prior knowledge of complex reflection coefficient and the difference of wave path are the key point to improve the performance of low angle altitude measurement.2.The DOA estimation method of spatially adjacent coherent source is studied.In low angle scenario,the direct signal and multipath are a pair of spatially adjacent coherent sources,whose performance of DOA estimation remains to be improved.Firstly,the decorrelation performance of the forward-backward spatial smoothing algorithm is analyzed,which indicates that the phase of the correlation coefficient influences the decorrelation performance of the forward-backward spatial smoothing algorithm.In some scenario,the forward-backward spatial smoothing algorithm could not decorrelate the spatially adjacent coherent sources.Secondly,a Root-MUSIC DOA estimation method based on spatial filtering is proposed.The method could separate the coherent sources utilizing spatial filtering.In theory the interference between the coherent sources could be alleviated or even eliminated.Then the DOA of the separated sources are estimated respectively.The performance of the method is improved by iteratively optimizing the parameter of the filters and the DOAs of the coherent sources.For non-uniform linear array,the virtual array technique is utilized to expand the application.Because the Root-MUSIC method is utilized for DOA estimation,there is no need of spatial searching and the computation burden is small.Computer simulation results indicate that SNR threshold of the method is low and the performance of the DOA estimation is close to CRB.3.Terrain parameter inversion method of low angle targets for VHF radar is studied.The prior knowledge of the terrain parameter can improve the accuracy of the altitude measurement method for low angle target,and the existing open literature has no effective terrain parameter inversion method.Especially for the perturbation model mentioned above,the terrain parameters are difficult to obtain.Therefore,inversion method of terrain parameter of low angle target is studied.Firstly,the terrain parameter inversion method is given for the classical model.The terrain parameters in the classical model include the height of the reflection surface and the complex reflection coefficient.The ambiguous characteristics of the reflection surface are analyzed,and the determined height search range is given.Then,based on the classical model,the terrain parameter inversion method of the perturbation parameters are given for the perturbation model.The exact model and synthesized vector maximum likelihood method is combined to calculate the low elevation angle.The validity of the method is verified by the computer simulation results.The real data verifies the validity of the model and the inversion method in the complex case.4.The altitude measurement method for low angle targets in complex terrain is studied.In complex terrain,the perturbation model is used to describe the multipath,and the existing algorithms do not consider the influence of the perturbational parameters.A new method based on sparse Bayesian learning is proposed.Firstly,using the sparseness of the target in the spatial domain and the flexibility of the sparse Bayesian learning to solve the special structure,the target elevation angle and the perturbation reflection coefficient are modeled as the super-parameters in the sparse Bayesian learning signal model.Then the low angle signal model for sparse Bayesian learning is established and the super-parameters are iteratively optimized by the EM(Expectation Maximization)algorithm,and then the target elevation and height are obtained.Finally,the computer simulation and real data processing result indicate that the proposed method can accurately estimate the perturbational parameters and target height.Because the effect of the ground perturbation of the multipath and the spatial information such as the spatial sparseness of the target and the information of the radar,reflection surface and the target radar are taken into account,performance of the proposed method to estimate the elevation and height of the target is better than the existing algorithms.5.The low angle DOA estimation method of polarized array is studied.Because the polarization information could significantly enhance the performance of environment sensing for array,the low angle DOA estimation method of double orthogonally oriented electromagnetic vector-sensor array is studied.Firstly,the multipath model of low angle target for double orthogonally oriented electromagnetic vector-sensor array is proposed.There is no significant difference between the horizontal polarization and the vertical polarization reflection coefficient at low elevation angle,and it is impossible to achieve better coherence performance from the polarized domain or the spatial space,respectively.It is necessary to use both the polarized domain and the spatial space information to decorrelate the direct and multipath signal.Finally the CRB of DOA estimation for the double orthogonally oriented electromagnetic vector-sensor array for the low angle signal model is analyzed.The comparison of the CRB for the scalar array and the polarized array illustrates the advantages of the use of polarization information for the DOA estimation of the coherent source of the polarized array.The results of computer simulation show that the polarization auxiliary angle and multipath phase difference are the main factors that affect the fluctuation of CRB and the performance of the DOA estimation method.Because of the use of spatial domain information and polarization information,the proposed method has the advantages of better performance for low angle DOA estimation when the polarization difference between direct signal and multipath is small.To some extent,the proposed method overcomes the poor decorrelation performance problem that the existing algorithms have in low elevation angle.