Improved Constraint for Multipath Mitigation in Constrained Moe Multiuser Detection

ZY Xu
DOI: https://doi.org/10.1109/jcn.2001.6596796
2001-01-01
Journal of Communications and Networks
Abstract:Constrained optimization method has been shown to be applicable for multiuser detection in the presence of multipath distortion. The performance of the constrained minimum output energy (MOE) detector is close to the minimum mean-square-error (MMSE) detector. However, due to additive noise, optimal constraints are biased estimates for channel parameters, resulting in some performance loss in multiuser detection. This loss becomes more significant when the communication channel has larger background noise (low SNRs). It is revealed in this paper that constrained optimization is closely related to subspace method which provides good estimate for the channel. Motivated by this result, the constrained cost function is modified to improve the channel estimate and thus better detection performance than the previously proposed constrained method. Detailed performance analysis and simulation results are also presented.
What problem does this paper attempt to address?