| In practical applications,due to the improvement of the radar angular resolution requirements and the superior estimation performance,the array size has been extended.However,especially for a large-scale antenna array,there is always a high probability of array antenna failure caused by the severe environment,human-made interference,hardware impairments,etc.The disabled transmit antennas are incapable of emitting the transmit signals,and the damaged receive antennas are not working for receiving targets echoes.Apparently,each damaged antenna in Multiple-Input Multiple-Output(MIMO)radar results in the absence of massive virtual elements in the virtual array after processing by the matched filtering.This inevitably corrupts the integrity of the virtual array,i.e.,thus reducing the maximum number of identifiable targets and greatly degrading the DOA estimation accuracy.This paper focuses on the direction of arrival(DOA)estimation in MIMO radar under array antenna failure,and the following research work mainly including:(1)In order to explore the impact of traditional DOA estimation under array antenna failure,subspace-based methods such as the Reduced Dimensional MUSIC(RD-MUSIC)algorithm and the Reduced Dimensional ESPRIT(RD-ESPRIT)algorithm are applied to DOA estimation.Firstly,the signal model of MIMO radar under array antenna failure is established.And then,the structure of the received signal matrix and virtual array covariance matrix under array antenna failure are analyzed in this paper.Constructing a tensor model allows the reliable recovery performance of the missing entries.The different position and num of the failed array elements are selected to compare their performance.Consequently,there is an important need in MIMO radar for reducing the impacts of faulty antennas via recovering the missing entries.(2)To alleviate this negative effect that the entire-row and entire-column missing data in the covariance matrix,a complete covariance matrix reconstruction method via jointly utilizing nuclear norm and smoothly clipped absolute deviation(SCAD)penalty is proposed.Firstly,a double priors’ constrained model is established,in which nuclear norm and SCAD penalty is utilized.A very good trade-off between the coarse sampling grid and the modeling accuracy can be achieved,and thus gaining a considerably reduced computational cost.Thereafter,this optimization problem is tackled properly under the alternating direction method of multipliers(ADMM)framework with the help of augmented Lagrange multipliers(ALM).Finally,based on the restored covariance matrix,the RD-ESPRIT algorithm is applied to DOA estimation.The proposed method can recover the missing entries of covariance matrix in a computationally inexpensive way and thus provide superior DOA estimation performance.(3)An improved tensor completion method based on tensor truncated convolution nuclear norm minimization is proposed to explore the inherent multidimensional structures in MIMO radar and handle faulty antennas in direction-of-arrival(DOA)estimation.First,we formulate a new tensor by convoluting the third-order tensor of MIMO radar with a kernel tensor in a truncated manner,and establish a low-rank tensor completion model equipped with the truncated convolution nuclear norm minimization.Then,we introduce a tractable relaxation of our minimization function by deducing the equality relation between the truncated convolution nuclear norm of a tensor and the nuclear norm of the truncated convolution matrix of the same tensor.Finally,we derive an efficient algorithm implementation based on ALM-ADMM to tackle our problem.By transposing of the mode-3 matrix unfolding of the recovery tensor,we can obtain the data matrix,and then the target DOA can be estimated via the RD-ESPRIT algorithm.The proposed method is quite effective in completing tensors with structurally missing entries and facilitating more accurate data recovery by exploiting the inherent multidimensional nature.The measured data in the actual environment is collected,and then the effectiveness of the proposed method in this paper is verified.This provides superior DOA estimation performance in MIMO radar. |
