A New Step-Size Control Algorithm Based on Noisy Data

In this paper, the VSS strategy shrinkage SM-NLMS (VSSM-NLMS) algorithm is proposed for noisy input data. In the presence of noisy input data, the step-size adaptation based on minimization of the squared residual can enhance the stability or convergence capability of the SM-NLMS algorithm. Moreover, the shrinkage method is employed for performance improvement. Simulation results demonstrate that the VSSM-NLMS algorithm is more robust and accurate than the state-of-the-art algorithms.

[1]  Lei Huang,et al.  Shrinkage Linear and Widely Linear Complex-Valued Least Mean Squares Algorithms for Adaptive Beamforming , 2015, IEEE Transactions on Signal Processing.

[2]  Andreas Antoniou,et al.  A Family of Shrinkage Adaptive-Filtering Algorithms , 2013, IEEE Transactions on Signal Processing.

[3]  Jacob Benesty,et al.  A Nonparametric VSS NLMS Algorithm , 2006, IEEE Signal Processing Letters.

[4]  Badong Chen,et al.  Time series prediction using kernel adaptive filter with least mean absolute third loss function , 2017 .

[5]  A. Antoniou,et al.  A set-membership nlms algorithm with adaptive error bound , 2008, 2008 Canadian Conference on Electrical and Computer Engineering.