LEAST MEAN MIXED-NORM ADAPTIVE FILTERING

A new family of stochastic gradient adaptive filter algorithms is prop osed which is based on mixed error norms. These algorithms combine the advantages of different error norms, for example the conventional, re latively well-behaved, least mean square algorithm and the more sensit ive, but better converging, least mean fourth algorithm. A mixing para meter is included which controls the proportions of the error norms an d offers an extra degree of freedom within the adaptation. A system id entification simulation is used to demonstrate the performance of a le ast mean mixed-norm (square and fourth) algorithm.