MIMO系统自适应均衡算法研究

MIMO信道为频率选择性信道,由于时延扩展而存在色散,因此研究MIMO系统的自适应均衡技术显得尤为重要.通过对自适应均衡技术的两种主要的算法最小均方(LMS)算法和递归最小二乘(RLS)算法的研究可以看出,在不同的步长因子及遗忘因子等参数变化情况下,LMS算法的收敛速度较慢,但均衡简单易实现,RLS算法收敛速度较快,但迭代运算较复杂.结合二者的特点提出了在MIMO系统中引入改进的最小均方算法,即归一化最小均方(NLMS)算法.仿真实验对比表明NLMS算法的计算量与LMS相当,但收敛条件简单,易实现,收敛速度较快,有很实际的应用价值.

An Adaptive Equalization Algorithm for MIMO System

MIMO channel is the frequency-selective channel,which has dispersion caused by delay spread. Therefore,it is particularly important to study adaptive equalization of MIMO system. Through the study of the two main adaptive equalization algorithms,least mean square algorithm(LMS)and recursive least squares algorithm(RLS),it is indicated that LMS algorithm is slow convergence,but simple and easy to relize while RLS algorithm is faster convergence,but more complex iterative calculations in circumstances such as different step-size factor and the forgetting factor and other changing parameters. With the characteristics of the two algorithms,improved least mean square algorithm is introduced in MIMO system,which is normalized least mean square algorithm(NLMS).Through simulation,it is indicated that NLMS algorithm and the LMS algorithm have the same computational complexity, but the former is faster convergence,whose convergence condition is simple and easy to achieve and is practically useful.