Maximum likelihood blind multiuser detection

Code division multiple access (CDMA) is the most promising candidate to implement the wireless multiple access scheme of third-generation mobile communication systems. This paper introduces a new blind approach to multiuser detection in time-dispersive channels. Asymptotically optimum multiuser detectors described in the literature rely on the knowledge of the channel characteristics to carry out maximum likelihood (ML) estimation of the transmitted symbols. On the other hand, blind multiuser receivers proposed so far are linear and, therefore, exhibit suboptimal performance. The receiver presented herein is based on the maximum likelihood (ML) estimation of both the multiuser channel and the transmitted symbols and thus provides asymptotically optimum performance. It is blind because it only requires knowledge of the probability density function (PDF) of the transmitted symbols and an estimate of the additive white Gaussian noise (AWGN) power