Semi-Blind Adaptive Multiuser Detection For Asynchronous CDMA

Abstract

In this paper, we propose a semiblind multiuser detection framework for asynchronous CDMA. Compared with most existing semiblind/blind detectors, the proposed framework requires a minimum number of previously received signals, which is about the number of interfering signals, and no detection filter converging or subspace separation procedure. The computational complexity and detection delay are therefore much lower. In this framework, a semiblind multiuser signal model is used instead of the widely-discussed conventional multiuser model or subspace-based parametric multiuser signal model. Following this framework, two optimal semiblind linear detectors are developed using the minimum variance unbiased estimation (MVU) and minimum mean squared error (MMSE) estimation criteria. Meanwhile, a multi-window scheme is proposed for simultaneously detecting several bits and a recursively adaptive procedure is developed for further lowering the complexity. After these, the asymptotic multiuser efficiency (AME) of the proposed framework, the comparison between the employed semiblind multiuser signal model and the conventional signal model, and several estimation bounds are discussed. Computer simulation results are presented to support the performance of the proposed semi-blind multiuser detection schemes.