1. Variable Step Size Norm-Constrained Adaptive Filtering Algorithms.
- Author
-
Shi, Long and Zhao, Haiquan
- Subjects
- *
ADAPTIVE filters , *ALGORITHMS , *SPARSE approximations , *SPARSE matrices , *COMPUTATIONAL complexity , *CONSTRAINED optimization - Abstract
Variable step size norm-constrained adaptive filtering algorithms are proposed in the paper. A variable step size is derived by minimizing the variance of the noise-free a posterior error. Thus, the update equation can obtain a reasonable step size at each iteration. Due to the introduction of variable step size, the proposed algorithms based on the constrained conditions of $$L_{1}$$ and $$L_{0}$$ norm have a significant advantage that the convergence rate is faster than some well-known algorithms in the sparse system. The simulation results illustrate the good performance of the proposed algorithms. [ABSTRACT FROM AUTHOR]
- Published
- 2017
- Full Text
- View/download PDF