Intrinsically stable adaptive recursive filters

Linear recursive filters and also recurrent neural networks can be adapted on-line but sometimes with instability problems. Stability control techniques exist for the linear case but they are either computationally expensive or non-robust. For the nonlinear case, stability control is simply usually not performed in applications. This paper presents a new stability control method for IIR adaptive filters that makes it possible to continually adapt the coefficients with no need for stability test or poles projection. This method can be applied to various filter realizations (direct forms, cascade or parallel second order sections, lattice form) as well as to locally recurrent neural networks, such as the IIR-MLP, with improved performance over other techniques and over not controlling stability. In the paper this method is applied to normalized lattice filters with valuable results; an analysis of the stabilization effects is also presented both analytically and experimentally