Blind identification and array processing applications of generalized higher-order statistics

Generalized higher-order statistics (GHOS) has been proposed as a possible approach to handle signals which may involve /spl alpha/-stable processes for which traditional higher-order statistics (HOS) can not be applied. Initial research results on GHOS revealed that GHOS may actually be a valuable tool for the analysis and processing of signals which do not contain /spl alpha/-stable processes, as well. In this paper, we will describe some possible applications of GHOS in the areas of blind system identification and array signal processing.