Transient Performance Analysis of Zero-Attracting LMS

Jie Chen (1) , Cédric Richard (2) , Yingying Song (3) , David Brie (3)

1 Centre of Intelligent Acoustics and Immersive Com- munications at School of Marine Science and Technology, Northwestern Polytechinical University
2 LAGRANGE - Joseph Louis LAGRANGE
3 CRAN - Centre de Recherche en Automatique de Nancy

Zero-attracting least-mean-square (ZA-LMS) algorithm has been widely used for online sparse system identification. It combines the LMS framework and 1-norm regularization to promote sparsity, and relies on subgradient iterations. Despite the significant interest in ZA-LMS, few works analyzed its transient behavior. The main difficulty lies in the nonlinearity of the update rule. In this work, a detailed analysis in the mean and mean-square sense is carried out in order to examine the behavior of the algorithm. Simulation results illustrate the accuracy of the model and highlight its performance through comparisons with an existing model.

Mots clés: zero-attracting LMS; Sparse system identification; performance analysis; transient behavior

Domaines: Traitement du signal et de l'image [eess.SP]

Jie Chen, Cédric Richard, Yingying Song, David Brie. Transient Performance Analysis of Zero-Attracting LMS. IEEE Signal Processing Letters, 2016, 23 (12), pp.1786-1790. ⟨10.1109/LSP.2016.2616890⟩. ⟨hal-01370271⟩