TY - JOUR

T1 - Kronecker product formulation of the cumulant based realization of stochastic systems.

AU - Giannakis, Georgios B.

PY - 1988/12/1

Y1 - 1988/12/1

N2 - The author derives recursive equations and closed-form expressions relating the parameters of an autoregressive-moving-average (ARMA) model (which may be non-minimum-phase and/or noncausal) with the cumulants of its output, in response to excitation by a non-Gaussian i.i.d. process. Based on these relationships, cumulant-based stochastic realization algorithms are developed. The output noise can be colored Gaussian or i.i.d. non-Gaussian with unknown variance. When a state-space representation is adopted, the stochastic realization problem reduces to the realization of a appropriate Hankel matrix formed by cumulant statistics. Using a Kronecker product formulation, exact expressions are presented for identifying state-space quantities when output cumulants are provided, or for computing output cumulants when the state-space triple is known. Conditions for stationarity of a linear process, with respect to its cumulants, are also presented. If a transfer function approach is used, alternative formulations are given to cover the case of noncausal models.

AB - The author derives recursive equations and closed-form expressions relating the parameters of an autoregressive-moving-average (ARMA) model (which may be non-minimum-phase and/or noncausal) with the cumulants of its output, in response to excitation by a non-Gaussian i.i.d. process. Based on these relationships, cumulant-based stochastic realization algorithms are developed. The output noise can be colored Gaussian or i.i.d. non-Gaussian with unknown variance. When a state-space representation is adopted, the stochastic realization problem reduces to the realization of a appropriate Hankel matrix formed by cumulant statistics. Using a Kronecker product formulation, exact expressions are presented for identifying state-space quantities when output cumulants are provided, or for computing output cumulants when the state-space triple is known. Conditions for stationarity of a linear process, with respect to its cumulants, are also presented. If a transfer function approach is used, alternative formulations are given to cover the case of noncausal models.

UR - http://www.scopus.com/inward/record.url?scp=0024135023&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0024135023&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:0024135023

VL - 88 pt 1-3

SP - 2096

EP - 2101

JO - Proceedings of the American Control Conference

JF - Proceedings of the American Control Conference

SN - 0743-1619

ER -