Steady-state mean-square error analysis of regularized normalized subband adaptive filters

• The MSE performance of regularized NSAFs is analyzed.
• The analysis shows the impact of parameters of the regularized NSAF on the MSE.
• The analysis explains the reason why the VR-NSAF obtains small misalignment.

The normalized subband adaptive filter (NSAF) has faster convergence rate than the normalized least-mean-square (NLMS) algorithm for colored input signals. Regularization of the NSAF is of importance in practical applications. In this paper, we analyze the steady-state mean-square error (MSE) of regularized NSAFs. The analysis is carried out based on the derivation of a variable regularization matrix NSAF (VRM-NSAF). Theoretical expressions for the steady-state MSE of two regularized NSAFs are derived under some assumptions. Simulation results are given to support the theoretical analysis.