Research On Target Angle Estimation Methods For Bistatic MIMO Radar In Complex Background

Multiple-input multiple-output(MIMO)radar is a new radar system,which utilizes waveform diversity to obtain higher degrees of freedom and flexibility in the airspace,and has received much more attention in recent years.The target angle estimation of MIMO radar is an important issue in radar signal processing,especially the joint estimation of direction of departure(DOD)and direction of arrival(DOA)of bistatic MIMO radar,which plays a significant role in target tracking,positioning,identification,etc.Because the target angle estimation of bistatic MIMO radar involves two-dimensional parameters,the processing is more complicated.Traditional methods can only obtain better estimation results under ideal backgrounds such as large snapshots,Gaussian white noise and uniform linear array.However,the operating environment of MIMO radar is becoming more and more complicated,leading to the decrease of the estimation accuracy of DOD and DOA by traditional methods.Therefore,based on gridless and off-grid compressed sensing theory,and the working principle and research progress of MIMO radar,this research focuses on the problem of DOD and DOA estimation of bistatic MIMO radar in complex backgrounds such as few snapshots,impulsive noise,array element mutual coupling and non-uniform noise,and non-uniform linear array.The main research results are as follows:(1)In the case of few snapshots,a joint estimation method of DOD and DOA for bistatic MIMO radar based on two dimensional atomic norm minimization(2D-ANM)is proposed.The proposed method combines gridless compressed sensing theory and does not need to obtain the signal covariance matrix,which achieves higher angle estimation accuracy than that of the traditional two-dimensional subspace based algorithm.Moreover,since the two-dimensional angle atom set in the continuous domain is adopted,this method overcomes the basis mismatch problem existing in the grid-based compressive sensing algorithm,and improves the estimation accuracy of DOD and DOA.In order to further reduce the complexity of the 2D-ANM algorithm,two low-complexity 2D-ANM algorithms are proposed based on the singular value decomposition method and the alternating direction method of multipliers(ADMM),reducing the running time and improving the efficiency.(2)In the case of impulsive noise,a joint estimation method of DOD and DOA for bistatic MIMO radar based on iterative reweighted l_p norm minimization is proposed.This method takes advantage of the characteristic that the l_p norm reduces extreme value sensitivity,and solves the objective function through block successive upper-bound minimization(BSUM),which effectively improves the estimation accuracy of DOD and DOA under impulsive noise.In addition,since the objective function is non-convex and nonlinear,the initial value selection of DOD and DOA is very important.Thus a two-dimensional Newtonized orthogonal matching pursuit(2D-NOMP)method is proposed,which uses the two-dimensional off-grid parameter estimation technology to correct the grid estimated value of the target angle through the Newton method,improving the accuracy of the initial value.(3)In the case of array element mutual coupling and non-uniform noise,a joint estimation method of DOD and DOA for bistatic MIMO radar based on truncated nuclear norm minimization(TNNM)and mutual array element elimination is proposed.This method takes advantage of the low-rank characteristics of the signal covariance matrix and the sparsity of the noise covariance matrix,and uses TNNM method to eliminate non-uniform noise,which can effectively overcome the mutual leakage between signal subspace and noise subspace caused by non-uniform noise in traditional subspace based algorithms.Furthermore,in order to effectively solve the mutual coupling problem of bistatic MIMO radar,a mutual coupling element elimination algorithm is proposed.By constructing two selection matrices according to the structural characteristics of the mutual coupling matrix,a new covariance matrix is formed.This method can avoid discarding the elements of the transmitting and receiving array at the same time compared with that of traditional methods,so it can obtain greater degrees of freedom and higher estimation accuracy in DOD and DOA estimation.(4)In the case of non-uniform linear array,a joint estimation method of DOD and DOA for bistatic coprime MIMO radar based on decoupled atomic norm minimization(D-ANM)and virtual difference coarray interpolation is proposed.The coprime array has the advantage of large array aperture and high resolution,but its elements are discontinuous,resulting in the degradation of the estimation performance of traditional methods based on uniform linear array.Considering the bistatic MIMO radar where the transmitting and receiving arrays are both coprime arrays,the D-ANM method is used to interpolate the virtual difference coarray in the coprime array to overcome the influence of the"hole"phenomenon and make the virtual differential array uniform and linear,thus fully utilizing the entire virtual difference coarray aperture,increasing the number of effective array elements and further improving the resolution of DOD and DOA estimation.In addition,this method converts D-ANM into a semidefinite programming(SDP)problem with residuals,and derives an efficient solution algorithm based on a specific ADMM,which effectively reduces the algorithm complexity.