EM Algorithm State Matrix Estimation for Navigation

The convergence of an expectation-maximization (EM) algorithm for state matrix estimation is investigated. It is shown for the expectation step that the design and observed error covariances are monotonically dependent on the residual error variances. For the maximization step, it is established tha...

Full description

Saved in:
Bibliographic Details
Published in:IEEE signal processing letters 2010-05, Vol.17 (5), p.437-440
Main Authors: Einicke, G.A., Falco, G., Malos, J.T.
Format: Article
Language:English
Subjects:
Algorithms
Convergence
Covariance matrix
Errors
Estimates
Filtering
Filtration
Iterative algorithms
Kalman filtering
Maximization
Maximum likelihood estimation
Navigation
Noise measurement
Parameter estimation
Signal processing algorithms
State estimation
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The convergence of an expectation-maximization (EM) algorithm for state matrix estimation is investigated. It is shown for the expectation step that the design and observed error covariances are monotonically dependent on the residual error variances. For the maximization step, it is established that the residual error variances are monotonically dependent on the design and observed error covariances. The state matrix estimates are observed to be unbiased when the measurement noise is negligible. A navigation application is discussed in which the use of estimated parameters improves filtering performance.
ISSN:1070-9908
1558-2361
DOI:10.1109/LSP.2010.2043151