Evaluation of multiple reference active noise control algorithms on Dornier 328 aircraft data

This correspondence presents an evaluation of multiple-reference adaptive a lgorithms. Two least mean squares-types (LMS-types) and a Newton-type algor ithm are considered, The special structure of the adaptive filtering proble m implies that the Newton-type algorithm can be implemented with the same n umerical complexity as LMS-type algorithms. The concept of a fast filtered- x Newton algorithm is thus introduced.