Title :
Frequency estimation derived from a dynamical system analysis
Author :
Hory, Cyril ; Ting, Michael ; Hero, Alfred
Author_Institution :
Dept. of Electr. Eng. & Comput. Sci., Michigan Univ., Ann Arbor, MI, USA
fDate :
28 Sept.-1 Oct. 2003
Abstract :
The main concern of this paper is to build a method of characterization and identification of a dynamical system with several degrees of freedom when only one observation record is available. The analysis of the number of degrees of freedom provided by performing a principal component analysis leads to an eigendecomposition-based method. The link between the proposed method and the MUSIC algorithm is then drawn. This approach gives a new insight on the use of this type of methods.
Keywords :
covariance matrices; eigenvalues and eigenfunctions; frequency estimation; principal component analysis; signal processing; time-varying systems; MUSIC algorithm; dynamical system analysis; eigendecomposition-based method; frequency estimation; principal component analysis; Couplings; Delay; Electrons; Frequency estimation; Magnetic fields; Magnetic force microscopy; Magnetic resonance; Multiple signal classification; Phase estimation; Principal component analysis;
Conference_Titel :
Statistical Signal Processing, 2003 IEEE Workshop on
Print_ISBN :
0-7803-7997-7
DOI :
10.1109/SSP.2003.1289378
