Assessment of the efficiency of the LMS algorithm based on spectral information

The LMS algorithm is used to find the optimal minimum mean-squared error (MMSE) solutions for a wide variety of problems. Unfortunately, its convergence speed depends heavily on its initial conditions when the autocorrelation matrix R of its input vector has a high eigenvalue spread. In many applications such as system identification or channel equalization, R is Toeplitz. In this paper we exploit the Toeplitz structure of R to show that when the weight vector is initialized to zero, the convergence speed of LMS is related to the similarity between the input PSD and the power spectrum of the optimum solution.

[1]  J. B. Foley,et al.  Stochastic convergence of the LMS algorithm in adaptive systems , 1987 .

[2]  H. J. Butterweck,et al.  A wave theory of long adaptive filters , 2001 .

[3]  Behrouz Farhang-Boroujeny.  On statistical efficiency of the LMS algorithm in system modeling , 1993, IEEE Trans. Signal Process..

[4]  B. Widrow,et al.  Stationary and nonstationary learning characteristics of the LMS adaptive filter , 1976, Proceedings of the IEEE.

[5]  Ali H. Sayed,et al.  H∞ optimality of the LMS algorithm , 1996, IEEE Trans. Signal Process..

[6]  Bernard Widrow,et al.  Least-mean-square adaptive filters , 2003 .

[7]  Scott C. Douglas,et al.  Exact expectation analysis of the LMS adaptive filter , 1995, IEEE Trans. Signal Process..

[8]  James R. Zeidler,et al.  A performance bound for the LMS estimator , 2000, IEEE Trans. Inf. Theory.

[9]  Bernard Widrow,et al.  On the statistical efficiency of the LMS algorithm with nonstationary inputs , 1984, IEEE Trans. Inf. Theory.

[10]  Dennis R. Morgan.  Slow asymptotic convergence of LMS acoustic echo cancelers , 1995, IEEE Trans. Speech Audio Process..

[11]  Ali H. Sayed,et al.  A unified approach to the steady-state and tracking analyses of adaptive filters , 2001, IEEE Trans. Signal Process..

[12]  James R. Zeidler,et al.  Nonlinear effects in LMS adaptive equalizers , 1999, IEEE Trans. Signal Process..

[13]  K. Senne,et al.  Performance advantage of complex LMS for controlling narrow-band adaptive arrays , 1981 .

[14]  Bernard Widrow,et al.  Least-Mean-Square Adaptive Filters: Haykin/Least-Mean-Square Adaptive Filters , 2005 .

[15]  Ehud Weinstein,et al.  Convergence analysis of LMS filters with uncorrelated Gaussian data , 1985, IEEE Trans. Acoust. Speech Signal Process..

[16]  S. Thomas Alexander,et al.  Adaptive Signal Processing , 1986, Texts and Monographs in Computer Science.

[17]  Onkar Dabeer,et al.  Analysis of mean-square error and transient speed of the LMS Adaptive algorithm , 2002, IEEE Trans. Inf. Theory.

[18]  Bernard Widrow,et al.  Statistical efficiency of adaptive algorithms , 2003, Neural Networks.