On soft multiuser channel estimation of DS-CDMA uplink using different mapping strategies

The performance of multi-user direct-sequence code division multiple access (DS-CDMA) transmission over frequency selective, slowly fading channels strongly depends on the reliability of the channel parameter estimates: state-of-the-art detection algorithms that exploit multiple-access-interference and inter-symbol-interference require very powerful estimation algorithms. In this contribution we consider a system with bit-interleaved coded modulation (BICM) with an iterative detector consisting of minimum mean squared error (MMSE) interference cancellation, followed by a soft demapper and a MAP decoder. We derive an estimator that makes use of the a posteriori probabilities computed by the detector. This multi-user estimator is based on the iterative space alternating generalized expectation maximization (SAGE) algorithm. We show through computer simulation that the proposed receiver considerably outperforms conventional channel estimation schemes. Finally, the impact of different mappings in the BICM scheme is evaluated.