.png){kind=icon}
.png){kind=icon}
Return
Cluster-sparsity-induced affine projection algorithm and its variable step-size version
.png){kind=icon}
PRE
AI
DOI：10.1016/j.sigpro.2022.108490.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The affine projection algorithm (APA) is widely used in various applications of adaptive filtering due to its decorrelating property. However, when the unknown system to be estimated is a cluster-sparse one, the APA cannot make full use of its cluster-sparsity characteristic to accelerate convergence rate. In this paper we propose a cluster-sparsity-induced APA (CSI-APA) to promote the performance of the adaptive filter for estimating cluster-sparse systems. When the number of elements in each cluster of the adaptive filter weights is set to one, the CSI-APA reduces to a sparsity-induced APA (SI-APA). Like other fixed step size adaptive filtering algorithms, the proposed CSI-APA needs to make a trade-off between convergence rate and steady-state misalignment. To address this problem, the step-size of the CSI-APA is optimized by minimizing the mean-square deviation (MSD) at each iteration and a variable step-size CSI-APA (VSSCSI-APA) is developed. Simulation results are provided to show the superior performance of the proposed algorithms.(c) 2022 Elsevier B.V. All rights reserved.
Keywords:
Cluster-sparse system
Affine projection algorithm (APA)
Variable step-size (VSS)
System identification
Zero-attraction
Mixed-norm regularization