Step-Size Control for the Normalized Subband Adaptive Filter
归一化子带自适应滤波器步长控制

Fixed step-size Subband Adaptive Filters (SAFs) must carry out a trade-off between fast convergence rate and low steady-state misadjustment. According to the functional relationship between the Mean-Square Deviation (MSD) of the coefficient vector of the adaptive filter and step-size, this paper proposes a step-size control algorithm to address the problem above, which is derived by the largest decrease method of the MSD of the coefficient vector of the adaptive filter for each iterative update. This algorithm can obtain both fast convergence rate and low steady-state misadjustment. Experimental results verify the validity of the proposed method.