LMS and RLS algorithms comparative study in system identification

In this paper the two most basic adaptive algorithms: LMS and RLS are discussed. Which were compared to the search path, RLS algorithm through theoretical analysis shows that the convergence rate is faster than the LMS algorithm. Finally, RLS and LMS algorithms are applied to system identification, the experimental data further illustrate the convergence rate of RLS algorithm is better than LMS algorithm.

[1]  Alfred O. Hero,et al.  Sparse LMS for system identification , 2009, 2009 IEEE International Conference on Acoustics, Speech and Signal Processing.

[2]  Anthony G. Constantinides,et al.  A novel kurtosis driven variable step-size adaptive algorithm , 1999, IEEE Trans. Signal Process..

[3]  Ali H. Sayed,et al.  Variable step-size NLMS and affine projection algorithms , 2004, IEEE Signal Processing Letters.

[4]  Shin'ichi Koike,et al.  A class of adaptive step-size control algorithms for adaptive filters , 2002, IEEE Trans. Signal Process..

[5]  Ender M. Eksioglu,et al.  RLS Algorithm With Convex Regularization , 2011, IEEE Signal Processing Letters.