Singular Value Decomposition-Based ARMA Model Parameter Estimation of Non-Gaussian Processes

SUMMARY

Autoregressive moving average (ARMA) modeling has been used in many fields. This paper presents an approach to time series analysis of a general ARMA model parameters estimation. The proposedtechnique is based on the singular value decomposition (SVD) of a covariance matrix of a third order cumulants from only the output sequence. The observed data sequence is corrupted by additive Gaussian noise. The system is driven by a zero-mean independent and identically distributed (i.i.d.) non-Gaussian sequence. Simulationsverify the performance of the proposed method.