Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
567326 | Signal Processing | 2006 | 9 Pages |
Abstract
For least mean-square (LMS) algorithm applications, it is important to improve the speed of convergence vs the residual error trade-off imposed by the selection of a certain value for the step size. In this paper, we propose to use a mixture approach, adaptively combining two independent LMS filters with large and small step sizes to obtain fast convergence with low misadjustment during stationary periods. Some plant identification simulation examples show the effectiveness of our method when compared to previous variable step size approaches. This combination approach can be straightforwardly extended to other kinds of filters, as it is illustrated with a convex combination of recursive least-squares (RLS) filters.
Related Topics
Physical Sciences and Engineering
Computer Science
Signal Processing
Authors
Jerónimo Arenas-García, Manel Martínez-Ramón, Ángel Navia-Vázquez, Aníbal R. Figueiras-Vidal,