Set-membership estimation for adaptive signal processing

The authors give a brief overview of the problem of set-membership estimation in adaptive signal processing. Recently, there seems to be a resurgence of interest in these types of estimation schemes as they feature 100% confidence regions for parameter estimates. Some of these recursive schemes also feature selective updating. More recently, the systolic implementation and lattice structure of some particular schemes have also been investigated. Exact polytope bounding, orthotope bounding, and the OBE algorithms are discussed.<
>