Title :
Colored input-signal analysis of normalized data-reusing LMS algorithms
Author :
de Campos, M.L.R. ; Apolinário, J.A., Jr. ; Diniz, P.S.R.
Author_Institution :
Dept. de Engenharia Eletrica, Inst. Militar de Engenharia, Rio de Janeiro, Brazil
Abstract :
In this paper, a simplified model is used to qualitatively and quantitatively analyze two normalized data-reusing LMS algorithms when the input sequence is colored. The proposed approach relates the eigenvalue distribution of the input-signal autocorrelation matrix with probabilities of occurrence of subsequent parallel or orthogonal input-signal vectors. Closed-form formulas that describe steady-state mean-squared error (MSE) are provided and confronted with simulation results
Keywords :
adaptive filters; correlation theory; eigenvalues and eigenfunctions; filtering theory; least mean squares methods; closed-form formulas; colored input-signal analysis; eigenvalue distribution; input sequence; input-signal autocorrelation matrix; normalized data-reusing LMS algorithms; orthogonal input-signal vectors; parallel input-signal vectors; steady-state mean-squared error; Adaptive filters; Algorithm design and analysis; Computational modeling; Convergence; Data analysis; Eigenvalues and eigenfunctions; Filtering algorithms; Least squares approximation; Steady-state; Throughput;
Conference_Titel :
Electronics, Circuits and Systems, 1998 IEEE International Conference on
Conference_Location :
Lisboa
Print_ISBN :
0-7803-5008-1
DOI :
10.1109/ICECS.1998.813306
