Robust Target Localization in Distributed MIMO Radar With Nonconvex ℓp Minimization and Iterative Reweighting

This letter deals with the problem of robust target localization in distributed multiple-input multiple-output (MIMO) radar using bistatic range measurements contaminated by outliers. Motivated by the robustness of nonconvex $\ell _{p} $ -norm for outlier rejection, in this letter, we reformulate the target localization problem as a nonconvex $\ell _{p} $ -norm minimization of residual matrix with nonconvex quadratic constraints. However, the resulted problem is very challenging. We consider the use of iterative reweighting algorithms, which approximates the nonconvex problem by a sequence of tractable subproblems. In particular, a new weight update method is proposed to accommodate the solving algorithm of the subproblem and avoid the selection of a regularization parameter, leading to an improved iterative reweighting ( $\ell _{p} $ -IIRW) solution. Numerical results demonstrate substantially enhanced robustness and improved positioning accuracy of the proposed method in both cases of low signal-to-interference-plus-noise ratio (SINR) outliers and non-line-of-sight (NLOS) outliers.