Title :
HRLS: a more efficient RLS algorithm for adaptive FIR filtering
Author_Institution :
Dept. of Inf. Manage., Nat. Defense Manage. Coll., Taipei, Taiwan
fDate :
3/1/2001 12:00:00 AM
Abstract :
The fast convergence rate and its immunity to the eigenvalue spread of the input correlation matrix make the RLS algorithm particularly attractive. However, the computational complexity is high. We propose using a hierarchical approach to reduce the computational complexity and further increase the convergence rate. The results of simulation runs and theoretical justifications confirm our claims.
Keywords :
FIR filters; adaptive Kalman filters; adaptive signal processing; computational complexity; filtering theory; least squares approximations; recursive estimation; recursive filters; HRLS; Kalman RLS algorithm; adaptive FIR filtering; computational complexity reduction; efficient RLS algorithm; eigenvalue spread immunity; fast convergence rate; hierarchical RLS; hierarchical approach; input correlation matrix; simulation results; Adaptive filters; Computational complexity; Computational modeling; Convergence; Eigenvalues and eigenfunctions; Filtering algorithms; Finite impulse response filter; Kalman filters; Mean square error methods; Resonance light scattering;
Journal_Title :
Communications Letters, IEEE
DOI :
10.1109/4234.913147
