A Robust Orthogonal Adaptive Approach to SISO Deconvolution

Abstract

This article presents a design procedure for an adaptive deconvolution processor with an orthogonal structure provided by a family of basis functions. The processor is designed to be robust to small parametric uncertainties in the single input single output (SISO) system model. Additionally, a simple gradient type of adaptive algorithm is applied to update the coefficients that linearly combine the fixed robust basis, so that larger changes in the system dynamics can be efficiently tracked. Performance comparisons are included between the optimal nominal, adaptive FIR and the proposed design for an equalization application on an asymetric digital subscriber loop (ADSL) channel. The main contribution is a structure requiring a low computational burden with good performance under a wide variety of conditions.