O Algoritmo LMS no Domínio da Transformada Wavelet Aplicado a Filtros de Volterra

This paper investigates the use of the Least-Mean Square (LMS) Algorithm in the Wavelet transform domain in the scenario of non-linear adaptive filters implemented with truncated Volterra series. Such transformation aims an increase in the speed of convergence of the LMS algorithm, very slow for this particular application, without a proportional increase in the computational complexity. The basic idea behind this approach is the Wavelet special feature of concentrating a higher amount of energy in a few coefficients, making possible a better weighting of the step size as well as a reduction of the coefficient vector order.