EPFL | Biomedical Imaging Group | Autoregressive Identification

On the Unique Identification of Continuous-Time Autoregressive Models from Sampled Data

H. Kirshner, M. Unser, J.P. Ward

IEEE Transactions on Signal Processing, vol. 62, no. 6, pp. 1361-1376, March 15, 2014.

In this work, we investigate the relationship between continuous-time autoregressive (AR) models and their sampled version. We consider uniform sampling and derive criteria for uniquely determining the continuous-time parameters from sampled data; the model order is assumed to be known. We achieve this by removing a set of measure zero from the collection of all AR models and by investigating the asymptotic behavior of the remaining set of autocorrelation functions. We provide necessary and sufficient conditions for uniqueness of general AR models, and we demonstrate the usefulness of this result by considering particular examples. We further exploit our theory and introduce an estimation algorithm that recovers continuous-time AR parameters from sampled data, regardless of the sampling interval. We demonstrate the usefulness of our algorithm for various Gaussian and non-Gaussian AR processes.

@ARTICLE(http://bigwww.epfl.ch/publications/kirshner1402.html,
AUTHOR="Kirshner, H. and Unser, M. and Ward, J.P.",
TITLE="On the Unique Identification of Continuous-Time Autoregressive
	Models from Sampled Data",
JOURNAL="{IEEE} Transactions on Signal Processing",
YEAR="2014",
volume="62",
number="6",
pages="1361--1376",
month="March 15,",
note="")

© 2014 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from IEEE. This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.