Cumulant Based Identification of Multichannel Moving-Average Models

Georgios B. Giannakis, Yujiro Inouye, Jerry M. Mendel

Research output: Contribution to journalArticlepeer-review

68 Scopus citations

Abstract

Given cumulants of a stationary, perhaps noisy, non- Gaussian r-variate moving-average, MA(q) process, we study identifiability conditions, under which the MA coefficient matrices, the input statistics, and the order q can be uniquely determined. The selection of a unique representative from the equivalence class corresponding to a given cumualant structure involves less restrictions than that corresponding to a given covariance structure. We derive two algorithms for estimating the (possibly) nonminimum-phase MA coefficient matrices.

Original languageEnglish (US)
Pages (from-to)783-787
Number of pages5
JournalIEEE Transactions on Automatic Control
Volume34
Issue number7
DOIs
StatePublished - Jul 1989

Fingerprint

Dive into the research topics of 'Cumulant Based Identification of Multichannel Moving-Average Models'. Together they form a unique fingerprint.

Cite this