论文信息 - A roundoff error analysis of the normalized LMS algorithm

A roundoff error analysis of the normalized LMS algorithm

This paper describes an analysis of the normalized LMS algorithm under finite word-length effects. It is shown that, using the averaging principle, it is possible to derive a good approximation for the total MSE in steady-state. Implementation issues are considered. It is shown that, in general, the quantization of the division operation will not cause the performance of the algorithm to deteriorate. An example using a look-up table for the division operation is analyzed and simulations are shown to support the analysis.

A.N. Willson | P.S. Chang

[1] Dirk T. M. Slock,et al. On the convergence behavior of the LMS and the normalized LMS algorithms , 1993, IEEE Trans. Signal Process..

[2] S. Haykin,et al. Adaptive Filter Theory , 1986 .

[3] Neil J. Bershad,et al. Analysis of the normalized LMS algorithm with Gaussian inputs , 1986, IEEE Trans. Acoust. Speech Signal Process..

[4] Andreas Antoniou,et al. Analysis of LMS-Newton adaptive filtering algorithms with variable convergence factor , 1995, IEEE Trans. Signal Process..

[5] N. Bershad. Nonlinear quantization effects in the LMS and block LMS adaptive algorithms-a comparison , 1988, ICASSP-88., International Conference on Acoustics, Speech, and Signal Processing.

[6] Arie Feuer,et al. Convergence and performance analysis of the normalized LMS algorithm with uncorrelated Gaussian data , 1988, IEEE Trans. Inf. Theory.

[7] C. Caraiscos,et al. A roundoff error analysis of the LMS adaptive algorithm , 1984 .

[8] Claude Samson,et al. Fixed point error analysis of the normalized ladder algorithm , 1983 .

[9] V. Umapathi Reddy,et al. Fixed point error analysis of the normalized ladder algorithm , 1982, ICASSP.