학술논문
A Novel Approach to Spectral Estimation and Moving Average Model Parameter Estimation
Document Type
Periodical
Author
Source
IEEE Signal Processing Letters IEEE Signal Process. Lett. Signal Processing Letters, IEEE. 30:1367-1371 2023
Subject
Language
ISSN
1070-9908
1558-2361
1558-2361
Abstract
We present a novel approach to spectral estimation, posing power spectrum density (PSD) estimation as a constrained optimization problem. More precisely, we optimize the coefficients of a finite impulse response filter such that the output variance is maximized. An analytical solution for this optimization problem is provided and we show that its solution establishes an unbiased estimator for the PSD from which the autocorrelation function and the moving average model coefficients can be derived. The performance of these estimators is demonstrated by comparison to a variety of established methods. Finally, our closed form solution allows interesting theoretical insights and an implementation based on the fast Fourier transform, leading to a fast and accurate estimator.