Global convergence analysis for the NIC flow

Author

Liu, Wanquan ; Yan, Wei-Yong ; Sreeram, V. ; Kok Lay Tao

Author_Institution

Dept. of Electr. & Comput. Eng., Curtin Univ. of Technol., Perth, WA, Australia

Volume

49

Issue

10

fYear

2001

fDate

10/1/2001 12:00:00 AM

Firstpage

2422

Lastpage

2430

Abstract

Previously, a family of fast subspace tracking algorithms based on a novel information criterion (NIC) has been proposed and investigated. It is known that these new algorithms are associated with a new kind of flow, which is called the NIC flow in this paper as in the case of the conventional Oja subspace algorithms with the Oja flow. In this paper, some fundamental questions about this new NIC flow such as its solution existence and convergence, are investigated. In addition, the convergence domain is characterized. Some important results on these issues are obtained via manifold theory

Keywords

convergence; information theory; nonlinear dynamical systems; signal processing; tracking; NIC flow; convergence domain; fast subspace tracking algorithms; global convergence analysis; manifold theory; novel information criterion; solution existence; Algorithm design and analysis; Convergence; Differential equations; Error correction; Manifolds; Neural networks; Reactive power; Riccati equations; Signal processing algorithms; Symmetric matrices;

fLanguage

English

Journal_Title

Signal Processing, IEEE Transactions on

Publisher

ieee

ISSN

1053-587X

Type

jour

DOI

10.1109/78.950796

Filename

950796