Adaptive filters employing partial updates

In some adaptive filtering applications, the least-mean-square (LMS) algorithm may be too computationally- and memory-intensive to implement. In this paper, we analyze two adaptive algorithms that update only a portion of the coefficients of the adaptive filter per iteration. These algorithms use decimated versions of the error and regressor signals, respectively. Simulations verify the accuracy of the analyzes, and the robustness of the algorithms is also explored.