摘要
重叠结构的迭代均衡器中运用固定步长LMS算法在静态信道下性能较好,但在动态信道下往往难于同时获得较快的收敛速度和较小的失调,而采用一些已有的基于误差信号来变步长的算法时,会因步长变化较大难以控制而导致性能变差。本文针对这一问题,利用迭代均衡的特点,提出一种步长因子随迭代次数而变化的自适应滤波算法,该算法复杂度低,步长控制简单。计算机仿真结果表明本文算法在Rician信道下SNR性能比LMS算法和MVSSLMS算法提升了2dB。
The fixed step size LMS algorithm perform well in static channel when used in a kind of iterative equalization with overlapping structure,but in dynamic channel it is difficult to exquisite the fast convergence speed and small misadjustment.When using the existing algorithms which use error signal to adjust the step size in dynamic channel,step size change greatly and is hard to control,which lead to a worse performance.Aimed at this problem and using the feature of iterative equalization,This paper present a new adaptive filter algorithm by changing the step-size according to the iterative times,which have a low complexity and insures the step-size easy to control.Computer simulations show that the proposed algorithm improves SNR by 2-3dB than LMS and MVSSLMS algorithm.
出处
《微计算机信息》
2010年第24期150-151,158,共3页
Control & Automation
关键词
变步长
LMS算法
迭代均衡
variable step size
LMS algorithm
iterative equalization
