论文部分内容阅读
改进型的变步长LMS算法在有效抑制瞬时噪声对经典的变步长LMS算法影响的同时,也增加了算法的计算复杂度,提高了其硬件实现的难度。为降低变步长LMS算法的计算复杂度,提出了一种步长改变因子与前后两个时刻误差的乘积成正比的新的变步长LMS改进算法,在不增加计算复杂度的条件下,有效地抑制了瞬时噪声对迭代步长的影响。仿真结果表明,提出的算法和现有的变步长LMS算法收敛速度相当,但其稳态误差更小,计算复杂度也更低,有利于算法的硬件实现。
The improved variable step size LMS algorithm can effectively reduce the influence of instantaneous noise on the classical variable step size LMS algorithm, and also increases the computational complexity of the algorithm and increases the difficulty of implementing the hardware. In order to reduce the computation complexity of variable length LMS algorithm, a new improved algorithm of variable step length LMS is proposed, which is proportional to the product of the step change factor and the error between the two previous moments. Without increasing the computational complexity, The effect of instantaneous noise on the iteration step size is effectively suppressed. The simulation results show that the proposed algorithm has the same convergence speed as the existing variable-length LMS algorithm, but its steady-state error is smaller and its computational complexity is lower, which is in favor of the hardware implementation of the algorithm.