In 2015 23rd European Signal Processing Conference (EUSIPCO), pages 2686-2690, Aug, 2015. Paper doi abstract bibtex
In this paper, we propose a novel structure for implementing a kernel adaptive filter as an add-on component for a linear adaptive filter. The kernel adaptive filter has been proposed as a solution to non-linear adaptive problems and their effectiveness has been demonstrated. However, it is not intended for replacing the linear adaptive filters at all, rather, we expect it to complement the performance of linear ones in nonlinear environments. We, therefore, consider a novel structure which enables us to implement a kernel adaptive filter as an add-on for a linear adaptive filter. The proposed structure performs as a linear adaptive filter in the linear-dominant environments, however, in non-linear environments, we can add a kernel adaptive filter without any modification on the operation of the linear one. The effectiveness of the proposed method is confirmed through the computer simulations.