Title :
Non-stationary AR model identification by batch estimation
Author_Institution :
Istituto Comunicazioni Elettriche, Roma, Italy
Abstract :
A batch procedure is proposed for identifying a non-stationary AR process. The procedure is based on a suitable model for the process and is characterized by optimal performance in terms of computational cost and identification. The procedure is illustrated by examples which compare it to the usual adaptive methods.
Keywords :
Computational efficiency; Minimization methods; Optimization methods; Power measurement; Predictive models; Reflection; Tin;
Conference_Titel :
Acoustics, Speech, and Signal Processing, IEEE International Conference on ICASSP '83.
DOI :
10.1109/ICASSP.1983.1172158
