Adaptive soft-input soft-output multiuser detection for asynchronous coded DS-CDMA systems

In this paper, we apply adaptive algorithms in the SISO multiuser detector in order to avoid the need for a priori information. First, we present the optimum SISO parallel decision feedback detector for asynchronous DS-CDMA systems, which requires many system parameters. Then, we propose two adaptive versions of this SISO detector which are based on the normalized least mean square (NLMS) and recursive least squares (RLS) algorithms, respectively. The signature waveforms and relative delays of multiple users are unknown to the adaptive SISO detector. Only a training sequence is required for each user. Our SISO adaptive detectors effectively exploit the a priori information of coded symbols, which is based on soft inputs from a bank of single user decoders, to further improve their convergence performance. Furthermore, we consider how to select practical finite feedforward and feedback filter lengths to obtain good tradeoff between the performance and computational complexity of the receiver. Finally, Monte-Carlo simulation results are presented and compared.