An Off-Grid DOA Estimation Method Based on Sparsity Adaptive
The quantization error generated by grid segmentation is an important drawback affecting the performance of source localiza-tion estimation.Aiming at the problems of the computational intensity of current Lp-norm off-grid algorithms and the need to predict sparsity in advance,a sparsity adaptive variable-step off-grid DOA localization method is proposed.Firstly,an angle-optimised off-grid parametric model is constructed based on first-order Taylor expansion,the change in residual energy is used as a condition for predicting sparsity K.The noise subspace and signal subspace orthogonality is then used as the basis for atomic error selection,and accurate solu-tions under off-grid models are realized using alternating iterative optimization methods.The proposed method combines a greedy algo-rithm support set selection strategy with valid information from the array covariance matrix.Simulation results reveal that the method not only reduces the operation time significantly,but also enables accurate estimation of any angle in the airspace angle range under the sparsity condition.
off-griddirection of arrival(DOA)sparse reconstructiongreedy algorithm
万方数据
维普
