Adaptive Semiblind Multiuser Detection for CDMA Systems

grantor: University of TorontoA novel adaptive semiblind multiuser detector based on subspace tracking is proposed for CDMA systems. A comparison of blind and semiblind multiuser detectors is also presented. Blind multiuser detectors consider only the in-cell users and neglect the inter-cell interference from out-of-cell users while semiblind multiuser detectors take into account both, intra-cell and inter-cell interference. Semiblind multiuser detectors are superior to blind multiuser detectors when the codes of some users are known to a base station e.g., codes of in-cell users Vs. out-of-cell users. It is shown that using the proposed semiblind detector, a considerable improvement in the capacity of the system is achieved. Compared to the existing blind and semiblind multiuser detectors, the proposed semiblind detector exhibits faster convergence, and is more robust in the case when the system is heavily loaded. It is also computationally less expensive.M.A.Sc