Title :
Power normalization sensitivity of adaptive lattice structures
Author :
Hodgkiss, W.S. ; Alexandrou, D.
Author_Institution :
Scripps Institution of Oceanography, San Diego, CA, USA
fDate :
8/1/1984 12:00:00 AM
Abstract :
The behavior of the gradient lattice (GL) and the least-squares lattice (LSL) when used to track multiple sinusoidal components whose power levels are widely separated is investigated. The sensitivity of the instantaneous frequency estimates to different time constants of the power estimation loops in the gradient method is of particular interest. Simulations are presented for the specific case of dual steps.
Keywords :
Adaptive signal processing; Cepstrum; Frequency estimation; Lattices; Least squares methods; Nonlinear filters; Signal processing; Signal processing algorithms; Speech processing; Transversal filters;
Journal_Title :
Acoustics, Speech and Signal Processing, IEEE Transactions on
DOI :
10.1109/TASSP.1984.1164383
