J. Idier et Jf. Giovannelli, STRUCTURAL STABILITY OF LEAST-SQUARES PREDICTION METHODS, IEEE transactions on signal processing, 46(11), 1998, pp. 3109-3111
A structural stability condition is sought for least squares linear pr
ediction methods in the given data case. Save the Toeplitz case, the s
tructure of the normal equation matrix yields no acknowledged guarante
e of stability. Here, a new sufficient condition is provided, and seve
ral least squares prediction methods are shown to be structurally stab
le.