On the identification of a third-order Volterra nonlinear system using a frequency-domain block RLS adaptive algorithm

A frequency-domain block recursive least-square adaptive algorithm is presented for the identification of nonlinear systems which can be modeled by a third-order Volterra series. The identification algorithm is based on the application of the Volterra theory and adaptive signal-processing techniques. The approach does not assume any particular statistics of the input. Moreover, the adaptive algorithm can be used, with a little modification, for the identification of a second-order Volterra system driven by stationary random inputs