Robust Fuzzy-based Data-Selective Affine Projection Algorithm

This paper presents a fuzzy-based approach to determine a threshold during the data selection process systematically. This makes the conventional data-selective affine projection algorithm (DSAP) robust to the impulse noise. In addition, an error-weighting vector is proposed to enhance our algorithm's robustness further. Simulation results confirmed that our proposed method ultimately improves stability, tracking ability, and robustness while reducing the weight-updating rate in the case of impulse noise presented.