A modified Least Squares Lattice filter to identify non stationary process
Data Analysis, Statistics and Probability
2007-05-23 v1 Instrumentation and Detectors
Abstract
In this paper the author proposes to use the Least Squares Lattice filter with forgetting factor to estimate time-varying parameters of the model for noise processes. We simulated an Auto-Regressive (AR) noise process in which we let the parameters of the AR vary in time. We investigate a new way of implementation of Least Squares Lattice filter in following the non stationary time series for stochastic process. Moreover we introduce a modified Least Squares Lattice filter to whiten the time-series and to remove the non stationarity. We apply this algorithm to the identification of real times series data produced by recorded voice.
Keywords
Cite
@article{arxiv.physics/0211077,
title = {A modified Least Squares Lattice filter to identify non stationary process},
author = {Elena Cuoco},
journal= {arXiv preprint arXiv:physics/0211077},
year = {2007}
}
Comments
19 pages, 15 figures, uses elsart.cls submitted to Signal Processing