Set-membership adaptive filtering

Based on a bounded-error assumption, set-membership adaptive filtering (SMAF) offers a viable alternative to traditional adaptive filtering techniques that are aimed to minimize an ensemble (or a time) average of the errors. This chapter presents an overview of the principles of SMAF, its features, and some applications. Highlighting the novel features of SMAF includes data-dependent selective update of the parameter estimates. Simulation experiences have shown that the SMAF algorithms normally use less than 20% of the data for updating parameter estimates while achieving performance comparable to traditional algorithms.