An estimation of non-minimum phase ARMA model using 2nd and 3rd order cumulants
Tomoaki Nitta, N. Hamada · 1991
The identification problem of non-minimum phase (NMP) systems by 2nd and 3rd order cumulants calculated from output sequences is addressed. The authors model the NMP system as an ARMA rational function whose orders are known, and the input of the system is unobservable, non-Gaussian white noise. The approach is simulated when zero-mean Gaussian white noise is added to the observation signal and the model order is mismatched.>