A new block adaptive algorithm using order recursive UD factorization method

Abstract

Adaptive algorithm especially plays the important role in adaptive signal processing, and much work has been proposed with various techniques until up to now. Among them, the algorithms using orthogonal projection are well known, which can be expressed with Moore-Penrose type generalized inverse matrix, with respect to the convergence characteristics. This paper presents a new block adaptive algorithm in which Moore-Penrose type generalized matrix can be efficiently calculated with order-update UD factorization. The proposed algorithm is expected to have stable property, since it includes less divisions in the procedures compared with the traditional last block adaptive algorithms.<>