BIFURCATION-ANALYSIS OF THE CONTROL OF AN ADAPTIVE EQUALIZER

In digital radio systems, adaptive equalization is often necessary to combat time-varying signal distortion that is introduced by transmission through the atmosphere. One possible adaptation technique is investigated, particularly in relation to the novel 'series' equalizer architecture. The dynamics of the adaptation algorithm are modelled by a system of nonlinear coupled first-order differential equations, which. for the simplest case of two adaptive parameters and scalar signals, may be transformed into a generalization of the Takens-Bogdanov normal form for a codimension-two double-zero bifurcation. Techniques from dynamical systems theory are used to analyse the model, and show that with sensible choices of physical parameters, there is no possibility of periodic behaviour.