Performance study of receiver based on independent component analysis in CDMA systems

To improve the performances of the wireless mobile communication system, a statistical method is illustrated in this paper. It consists in separating the signals at the reception of communication systems based on code division multiple access technology. The idea is to optimize the separation of the various users sharing the same frequency and temporal resources using the emergent statistical method of independent component analysis (ICA). ICA makes it possible to extract emitted signals that are as statistically independent as possible. The bit error rate and the signal to noise ratio are used as criteria for evaluating the performances of the ICA receiver. Adding white Gaussian noise to the input signal channel and the Rayleigh channel (fading channel) cases has been considered. A comparative study with conventional receivers such as the RAKE, the matched filter (MF), and the minimum mean-squared error is carried out. The obtained results show the superiority of an ICA receiver compared to an MF receiver.