Robust super-exponential methods for deflationary blind equalization of static systems

Abstract

The so called “super-exponential” methods (SEM's) are attractive methods for solving blind signal processing problems. The conventional SEM's, however, have such a drawback that they are very affected by Gaussian noise. To overcome this drawback, we propose a new SEM. While the conventional SEM's use the second-and higher-order cumulants of observations, the proposed SEM uses only the higher-order cumulants of observations. Because higher-order cumulants are not affected by Gaussian noise, the proposed SEM is robust to Gaussian noise, which is referred to as a robust super-exponential method (RSEM). To show the validity of the proposed RSEM, some simulation results are presented.