Title :
Sparse-befap: A fast implementation of fast affine projection avoiding explicit regularisation
Author :
Rombouts, Geert ; Moonen, Marc
Author_Institution :
KULeuven/ESAT-SISTA, Kardinaal Mercierlaan 94, 3001 Heverlee, Belgium
Abstract :
It is well known that régularisation of the covariance matrix is necessary in Affine Projection Algorithm (APA) based adaptive filters. A technique to achieve régularisation is using non-successive equations in the system of equations that is solved in the APA-algorithm, leading to the Sparse-APA algorithm [5] [4]. In this paper, a fast exact frequency domain implementation of Sparse-APA is derived which does not rely on any of the approximations which are found in the classical fast affine projection algorithms, while only a very small number of extra computations is needed.
Keywords :
Complexity theory; Covariance matrices; Equations; Frequency-domain analysis; Mathematical model; Projection algorithms; Vectors;
Conference_Titel :
Signal Processing Conference, 2000 10th European
Print_ISBN :
978-952-1504-43-3