Title :
Error detection for fast Toeplitz eigensolvers
Author :
Wallace, Tom ; Taylor, Fred
Author_Institution :
Adv. Power Technol. Inc., Washington, DC, USA
fDate :
5/1/1996 12:00:00 AM
Abstract :
Several authors have proposed fast methods for Toeplitz eigendecomposition based on fast Yule-Walker solvers. The most efficient of these fast methods are numerically unstable. This work presents an efficient test to detect errors resulting from instability and ensure the accuracy of eigenvalues and eigenvectors computed using fast methods
Keywords :
Hermitian matrices; Toeplitz matrices; array signal processing; eigenvalues and eigenfunctions; error analysis; error detection; estimation theory; Hermitian matrices; Toeplitz eigendecomposition; array-bearing estimation; eigenvalues; eigenvectors; error detection; fast Toeplitz eigensolvers; fast Yule-Walker solvers; instability; spectral estimation; Covariance matrix; Eigenvalues and eigenfunctions; Equations; Integrated circuit noise; Power engineering computing; Signal processing; Signal processing algorithms; Testing;
Journal_Title :
Signal Processing, IEEE Transactions on
