Sparseness-Aware Data-Selective LMS Algorithm

Data-selective adaptive algorithms are well-suited for reducing the complexity of weight updating in system identification problems. Nevertheless, impulse noise can obstruct the effectiveness of their data selection schemes. To address this issue, we introduce a sparsity-aware data-selective least mean square (DS-LMS) algorithm that enhances the data selection scheme for sparse system identification in the presence of impulse noise. Our approach was tested through numerical experiments, which confirmed its efficacy.