Robust Weight-Constraint Decorrelation Normalized Maximum Versoria Algorithm

Abstract

A robust weight-constraint decorrelation normalized maximum versoria algorithm is proposed in this paper. The proposed algorithm is designed by maximizing the normalized versoria cost function of the decorrelation error, and thus is robust against the impulsive noise and exhibits fast convergence in the case of highly correlated signals. The stability and computational complexity also be analyzed. Finally, simulation results demonstrate that the proposed algorithm achieves faster convergence speed than the MVC algorithm for colored input signal under the impulsive noise environment.