A Semidefinite Programming Approach to ARMA Estimation

Abstract A method for estimation of scalar ARMA models is proposed. In a first step the AR polynomial is estimated using a novel subspace fitting method. By using the estimated AR polynomial to filter the original data the MA polynomial is determined via an MA-covariance fitting method. Both estimation steps are formulated as quadratic optimization problems with LMI constraints which guarantee valid ARMA solutions and are solved as semidefinite programs.

[1]  Mats Viberg,et al.  Subspace-based methods for the identification of linear time-invariant systems , 1995, Autom..

[2]  James Durbin,et al.  The fitting of time series models , 1960 .

[3]  Petre Stoica,et al.  Vector ARMA estimation: an enhanced subspace approach , 1999, Proceedings of the 38th IEEE Conference on Decision and Control (Cat. No.99CH36304).

[4]  P. Stoica,et al.  MA estimation in polynomial time , 1999, Proceedings of the 38th IEEE Conference on Decision and Control (Cat. No.99CH36304).

[5]  Bjorn Ottersten,et al.  A subspace fitting method for identification of linear state-space models , 1995, IEEE Trans. Autom. Control..

[6]  T. McKelvey.  Identification of State-Space Models from Time and Frequency Data , 1995 .

[7]  H. Akaike.  Maximum likelihood identification of Gaussian autoregressive moving average models , 1973 .

[8]  Stephen P. Boyd,et al.  FIR filter design via semidefinite programming and spectral factorization , 1996, Proceedings of 35th IEEE Conference on Decision and Control.

[9]  Bo Wahlberg,et al.  Analysis of state space system identification methods based on instrumental variables and subspace fitting , 1997, Autom..

[10]  G. Wilson.  Factorization of the Covariance Generating Function of a Pure Moving Average Process , 1969 .

[11]  Manfred Deistler,et al.  Consistency and relative efficiency of subspace methods , 1994, Autom..

[12]  Jan M. Maciejowski,et al.  Realization of stable models with subspace methods , 1996, Autom..