R. Shaghaghi-Kandovan (Iran)
HOS , Blind Identification , ARMA , SISO, AR, MA
In blind identifying Single Input / Single Output (SISO) systems, if we assume a minimum phase system, the possibility of identifying it by using autocorrelation function of the received signal does exist. For non minimum phase systems, owing to absence of phase information in autocorrelation function, its deployment alone is not sufficient. To identify the non-minimum phase systems, it is possible to use methods based on Higher Order Statistics (HOS). But due to high variance of prediction errors, these methods do not yield satisfactory results. To identify the non-minimum phase systems, a new method based on autocorrelation function and the fourth order comulant for the received signal is presented in this article. In this method, the same level of precision used in minimum phase system through autocorrelation function, the possibility of extracting phase information from HOS, and combining the two methods is used.
Important Links:
Go Back
