ADAPTIVE VOLTERRA CHANNEL EQUALIZATION WITH LATTICE ORTHOGONALIZATION

The authors present a new recursive least-squares adaptive nonlinear e qualiser based on Volterra series expansion. The main approach is to t ransform the nonlinear equalisation problem into an equivalent multich annel. but linear, equalisation problem. Then the multichannel input s ignal is completely orthogonalised using sequential processing multich annel lattice stages. With the complete orthogonalisation of the input signal, only scalar operations are required, instability problems owi ng to matrix inversion are avoided, and good numerical properties are achieved. Avoidance of matrix inversion and vector operations reduce t he complexity considerably, and make the filter highly modular and sui table for VLSI implementation. Several experiments demonstrating the p erformance under different channel distortion and channel noise condit ions are also included in the paper.