A novel reduced-rank approach for implementing Volterra filters

This paper presents a novel reduced-rank approach for implementing Volterra filters with reduced complexity. Such an approach is based on the application of the singular value decomposition to a new form of coefficient matrix obtained by exploiting the representation based on diagonal coordinates of the Volterra kernels. The result is a parallel structure of extended Hammerstein models in which each branch is related to one of the singular values of the coefficient matrix. Then, removing the branches related to the smallest singular values, an effective reduced-complexity Volterra implementation is obtained. Simulation results are presented to confirm the effectiveness of the proposed approach.