Abstract: Blind multi-user detection methods in Code-Division Multiple-Access（CDMA） system are used to eliminate the Multiple Access Interference（MAI） and the Near-Far effect. For microprocessor or DSP has a limit word length, blind multi-user detection based on Kalman filter must face the problem of numeric stability. A new method based on QR factor is used in the blind multi-user detection based on Kalman filter. Simulation results show that this method is efficient in MAI and Near-Far effect elimination.

Key words: Code-Division Multiple-Access（CDMA）, Multiple Access Interference（MAI）, blind multi-user detection, QR Kalman filter

摘要： 码分多址（Code-Division Multiple-Access，CDMA）系统中，采用盲多用户检测算法抑制多址干扰（Multiple Access Interference，MAI）和“远-近”效应（Near-Far effect）的影响，是CDMA技术研究热点之一。为避免基于卡尔曼（Kalman）滤波的多用户检测算法中，协方差矩阵失去非负定性而面临的数值稳定性问题，提出利用QR分解方法重新构造系统模型协方差阵，建立一种收敛速度更快、数值鲁棒性更好的卡尔曼盲自适应多用户检测算法。仿真结果表明，该算法具有较好的抗多址干扰和抗“远-近”效应能力。

关键词: 码分多址, 多址干扰, 盲多用户检测, QR分解卡尔曼滤波