UM ALGORITMO RLS-LMS USANDO COMBINAÇÃO AFIM DE FILTROS ADAPTATIVOS PARA CONFORMAÇÃO DE FEIXE EM ANTENAS INTELIGENTES

Abstract

: In this paper, we show the convergence aspects of LMS (least mean-square) and RLS (least mean-square) algorithms using an affine combination of filters to adaptive beamforming in smart antennas. Specifically, we present the performance investigation of an affine combination of two individual adaptive filters of different classes, considering the mixture parameter of the combination calculated adaptively through the stochastic gradient algorithm, called of 𝝆𝜼 -LMS. The aim of the combination is to obtain, in stationary environment, an RLS-LMS adaptive algorithm that outperforms the classic algorithms in terms of convergence speed. The performance of the proposed affine RLS-LMS algorithm is assessed through computational experiments.