Filtragem adaptativa com restrições lineares por minimos quadrados

AUTOR(ES)
DATA DE PUBLICAÇÃO

1991

RESUMO

The basic problem of linear constrained filtering consists In searching ao optimal filter whose parameters are subject to a set of linear equations. For transversal structures, the optimal coefficients In the mean-square sense are given by the Wlener-Hopf equations. The introduction of constraints leads to a modified solution, obtained by the Lagrange multipliers method. In many cases, it is interesting to obtain the eonstrained filter by ao adaptive procedure in order to provide useful methods for real-time operations and for non-stationary environment. In this sense, we propose in this work a new recursive algorithm which updates the constrained filter by least-squares techniques. The algorlthm is exact, stable and accurate, with a eomputational complexity proportional to NK; N being the number of parameters and K the number of constraints. The technique may be applied in several problems of signal processing, as: spatial filtering, adaptive arrays, spectral analysis and channel equalization. Finally, we employ the proposed algorithm in order to implement ao adaptive filter with linear-phase property

ASSUNTO(S)

telecomunicações engenharia eletrica

ACESSO AO ARTIGO

Documentos Relacionados