Sparsity-aware normalized subband adaptive filters with jointly optimized parameters

• Published in: Journal of the Franklin Institute 2020-11, Vol.357 (17), p.13144-13157
• Main Authors: Ji, Lipeng, Ni, Jingen
• Format: Article
• Language: English
• Subjects: Adaptive filters; Adaptive technology; Algorithms; Computer simulation; Convergence; Misalignment; Optimization; Parameters; Sparsity; Steady state; Stochastic models
• ISSN: 0016-0032; 1879-2693; 0016-0032
• DOI: 10.1016/j.jfranklin.2020.09.015

•This paper addresses the problem of tradeoff between fast convergence rate and small misalignment existing in the sparsity-aware NSAF.•We jointly optimize the step-size and zero-attractor intensity factor in the mean-square deviation sense.•Simulation results are provided to verify the effectiveness of the proposed algorithms. The normalized subband adaptive filter (NSAF) has a faster convergence rate than the NLMS adaptive filter when the input signal is correlated. Recently some sparsity-aware NSAFs (SA-NSAFs) were presented, which make use of the sparsity of the unknown system to accelerate convergence rate or reduce the steady-state misalignment. However, like the NSAF they also need to take a tradeoff between fast convergence rate and small steady-state misalignment. To address this problem, this paper proposes to jointly optimize the step-size and intensity factor of the SA-NSAFs. Another advantage of the presented method is that it can solve the problem of selecting the optimal intensity factor for the SA-NSAFs with repeated manual attempts. The parameter optimization is achieved by minimizing the mean-square deviation (MSD). Simulation results are provided to show the superiority of the proposed algorithms.