Exploiting independence for co-channel interference cancellation and symbol detection in multiuser digital communications

In the blind equalization of multi-input multi-output (MIMO) finite impulse response communication channels, co-channel interference (CCI) is typically cancelled by exploiting the properties of digital modulations, such as their finite alphabet (FA). This contribution takes advantage of the mutual independence of the users' signals through the application of independent component analysis (ICA). We demonstrate that ICA-based CCI suppression remarkably improves an FA-based approach. In addition, proposed ICA-assisted minimum mean square error receivers are shown to enhance the conventional detection capabilities of MIMO equalization methods relying on channel identification. The particular structure of the MIMO model yields a simplified detection scheme with improved performance at a reduced computational cost.