An application of independent component analysis to DS-CDMA communications systems

In this work we present the application of independent component analysis (ICA) to the problem of blind detection of the transmitted users in DS-CDMA communication systems. Our main contribution is a novel criterion (and the corresponding algorithm) that allows to jointly exploit the information of several higher order statistics to blindly detect the users in DS-CDMA systems. In order to corroborate the theoretical behaviour of the algorithms, we built a software radio platform at 5 GHz aimed at the development of radio interfaces for the fourth generation of mobile communication systems. We made real measurements of a WCDMA-3GPP signal transmitted at 3.84 Mchip/s.