360 likes | 618 Views
DSP-CIS Chapter-12: Least Mean Squares (LMS) Algorithm. Marc Moonen Dept. E.E./ESAT, KU Leuven marc.moonen@esat.kuleuven.be www.esat.kuleuven.be / scd /. Part-III : Optimal & Adaptive Filters. : Optimal & Adaptive Filters - Intro General Set-Up Applications
E N D
DSP-CISChapter-12: Least Mean Squares (LMS) Algorithm Marc Moonen Dept. E.E./ESAT, KU Leuven marc.moonen@esat.kuleuven.be www.esat.kuleuven.be/scd/
Part-III : Optimal & Adaptive Filters : Optimal & Adaptive Filters - Intro • General Set-Up • Applications • Optimal (Wiener) Filters • : Least Squares & Recursive Least Squares Estimation • Least Squares Estimation • Recursive Least Squares (RLS) Estimation • Square-Root Algorithms • : Least Means Squares (LMS) Algorithm • LMS/NLMS : Stochastic Gradient Algorithms • LMS analysis • LMS Family • : Fast Recursive Least Squares Algorithms : Kalman Filtering Chapter-11 Chapter-12 Chapter-13 Chapter-14 Chapter-15
Least Mean Squares (LMS) Algorithm (Widrow 1965 !!)
Least Mean Squares (LMS) Algorithm Bernard Widrow
Least Mean Squares (LMS) Algorithm large λ_max implies a small stepsize
Least Mean Squares (LMS) Algorithm error vector projected onto eigenvectors initial error vector projected onto eigenvectors (=projection on i-th eigenvector) • small λ_i implies slow convergence • λ_min <<λ_max (hence small μ) implies *very* slow convergence