1984IEEE Transactions on Acoustics Speech and Signal ProcessingOpen access

An improved maximum likelihood method for power spectral density estimation

M. Lagunas-Hernandez, A. Gasull-Llampallas

Open full text 52 citations

Abstract

In this paper, we present a new procedure for spectral analysis; it represents a tradeoff between the high resolution provided by the maximum entropy method (MEM) and the low sidelobe characteristic of the maximum likelihood method (MLM). The approach that we follow is to introduce a slight modification of the second technique to obtain from the ML filter the true spectral density, not the power level as the original procedure does. The proposed method can be used both in one-dimensional and two-dimensional problems without a remarkable increase of computational load relative to the original MLM.

Open-access reader

About this research paper

What this paper is about

In this paper, we present a new procedure for spectral analysis; it represents a tradeoff between the high resolution provided by the maximum entropy method (MEM) and the low sidelobe characteristic of the maximum likelihood method (MLM). The approach that we follow is to introduce a slight modification of the second technique to obtain from the ML filter the true spectral density, not the power level as the original procedure does. The proposed method can be used both in one-dimensional and two-dimensional problems without a remarkable increase of computational load relative to the original MLM.

Why it matters

OpenAlex reports 52 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

In this paper, we present a new procedure for spectral analysis; it represents a tradeoff between the high resolution provided by the maximum entropy method (MEM) and the low sidelobe characteristic of the maximum likelihood method (MLM). The approach that we follow is to introduce a slight modification of the second technique to obtain from the ML filter the true spectral density, not the power level as the original procedure does. The proposed method can be used both in one-dimensional and two-dimensional problems without a remarkable increase of computational load relative to the original MLM.

Key concepts: Maximum entropy spectral estimation, Maximum likelihood, Maximum entropy method, Spectral density, Principle of maximum entropy, Spectral density estimation, Maximum likelihood sequence estimation, Mathematics

Related papers

Back to paper searchBrowse research topicsOriginal source
An improved maximum likelihood method for power spectral density estimation — Research Paper | ScholarLens