Multiweight optimization in optimal bounding ellipsoid algorithms

Optimal Bounding Ellipsoid (OBE) algorithms offer an attractive alternative to traditional least-squares methods for identification and filtering problems involving affine-in-parameters signal and system models. The benefits-including low computational efficiency, superior tracking ability, and sele...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on signal processing 2006-02, Vol.54 (2), p.679-690
Main Authors: Joachim, D., Deller, J.R.
Format: Article
Language:English
Subjects:
Algorithms
Applied sciences
Computational efficiency
Detection, estimation, filtering, equalization, prediction
Ellipsoids
Exact sciences and technology
Filtering
Filtering algorithms
Information, signal and communications theory
Iterative algorithms
Least squares method
Mathematical models
Noise robustness
Optimization
Recursive estimation
Samarium
Set-membership identification
Signal and communications theory
Signal processing
Signal processing algorithms
Signal, noise
system identification
Technological innovation
Telecommunications and information theory
Tracking
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:Optimal Bounding Ellipsoid (OBE) algorithms offer an attractive alternative to traditional least-squares methods for identification and filtering problems involving affine-in-parameters signal and system models. The benefits-including low computational efficiency, superior tracking ability, and selective updating that permits processor multi-tasking-are enhanced by multiweight (MW) optimization in which the data history is considered in determining update times and optimal weights on the observations. MW optimization for OBE algorithms is introduced, and an example MW-OBE algorithm implementation is developed around the recent quasi-OBE algorithm. Optimality of the solution is discussed, and simulation studies are used to illustrate performance benefits.
ISSN:1053-587X
1941-0476
DOI:10.1109/TSP.2005.861893