一种新的盲多用户检测最大似然方法

A Maximum Likelihood Approach to Blind Multiuser Detection

线性受限最大似然盲多用户检测是盲多用户检测中的一类重要方法,现有的限制方法不能保证目标用户的最优解落在限制的搜索空间中,因此在理论上不能达到最优解。提出了一种新的限制方法,能保证目标用户的最优解落在限制的搜索空间中,因此保证算法收敛到目标用户。与现有方法相比,这里给出的算法结构简单,降低了运算复杂度。仿真试验表明其具有较低的误码率和较高的系统容量。

Maximum likelihood linear(MLL) blind multiuser detection is novel in blind multiuser detection. To MLL detector, a main problem is how to add a linear constraint on the coefficients of detector which ensures that the desired user is extracted. A new linear constraint with simple construction is proposed. The computation complexity is reduced because RAKE receiver is not needed in the approach. Simulation shows that the proposed approach has lower symbol error rate and higher system capacity compared to existing approaches.