Title :
Blind system identification based on the complex cepstrum of the cyclic autocorrelation
Author :
Hatzinakos, Dimitrios
Author_Institution :
Dept. of Electr. & Comput. Eng., Toronto Univ., Ont.
Abstract :
A new discrete-time method is proposed for blind identification of nonminimum phase systems. It is assumed that the system is driven by a cyclostationary independent and identically distributed (i.i.d.) process. The complex cepstrum of the cyclic autocorrelation of the system output is derived, and its properties are investigated. It is shown that this complex cepstrum can be utilized to correctly identify the magnitude and phase characteristics of the system. The analytical work is supported by computer simulations
Keywords :
cepstral analysis; correlation theory; identification; blind system identification; complex cepstrum; cyclic autocorrelation; discrete-time method; magnitude characteristics; nonminimum phase systems; phase characteristics; Additive noise; Autocorrelation; Cepstral analysis; Cepstrum; Deconvolution; Higher order statistics; Linear systems; Signal processing; Speech processing; System identification;
Conference_Titel :
Circuits and Systems, 1993., ISCAS '93, 1993 IEEE International Symposium on
Conference_Location :
Chicago, IL
Print_ISBN :
0-7803-1281-3
DOI :
10.1109/ISCAS.1993.393824
