Symbol separation in oversaturated CDMA system

Code division multiple access (CDMA) is a multiple access technique based on spread spectrum methods. In a downlink signal processing, only the code of the mobile phone user is known, while the codes of the interfering users are unknown. In addition, the number of sources (users and paths) is usually larger than the code length. In this case, linear methods such as the matched filter fail to estimate the parameters. Blind source separation or independent component analysis is an approach offering the solution to this problem. We introduce blind source separation methods for estimating the symbols in an AWGN channel. The performance of the methods are based on the facts that the symbols have finite alphabet and/or are non-Gaussian. We present an algorithm which can estimate the parameters in an oversaturated system, i.e. when the number of binary signals can be larger than the code length. Simulations show that one can estimate the symbols without any knowledge of the chip sequences