Abstract: A blind adaptive recursive least squares multiuser detection algorithm based on Constrained Minimum Output Energy（CMOE） criterion is proposed to cope with the numerical instability and slow convergence rate of normal Recursive Least Squares（RLS） algorithm. This algorithm is a combination of CMOE criterion and normal RLS algorithm. Detailed simulation is carried out to compare the?proposed?algorithm with the normal RLS algorithm in the aspects of Signal Interference Ratio（SIR）, Excess Output Energy（EOE） and Bit Error Rate（BER） under conditions of both static and dynamic. Simulation results show that the algorithm is performed with more dynamic tracking ability, faster convergence speed and better stability.

Key words: Recursive Least Squares（RLS）, Constrained Minimum Output Energy（CMOE）, blind adaptive multiuser detection, Signal Interference Ratio（SIR）, Excess Output Energy（EOE）, Bit Error Rate（BER）, multiuser interference

摘要： 针对直接递归最小二乘（RLS）算法存在的检测数值不稳定和收敛速度较慢等问题，将约束最小输出能量（CMOE）准则与直接RLS算法结合，提出一种基于CMOE准则的盲自适应RLS多用户检测算法。将该算法与直接RLS算法进行动、静态环境下输出信干比（SIR）、剩余输出能量（EOE）和误码率（BER）等方面的仿真，对比得出该算法具有更好的动态跟踪能力，更快的收敛速度和更高的稳定性。

关键词: 递归最小二乘（RLS）, 约束最小输出能量（CMOE）, 盲自适应多用户检测, 信干比（SIR）, 剩余输出能量（EOE）, 误码率（BER）, 多用户干扰