论文部分内容阅读
分析了一般变换域中的传统LMS算法,利用信号的相关矩阵经过子波变换后的稀疏带状结构以及区间子波变换能有效地抑制边界效应的能力,讨论了一种新的子波变换域的Newton-LMS类算法.仿真结果表明,对于高低频信号,该算法能给出矩阵的较准确的模块化估计,且收敛速度明显快于传统的DFT-LMS和DCT-LMS等算法.
The traditional LMS algorithm in general transform domain is analyzed. By using the sparse band structure of signal correlation matrix after wavelet transform and the ability of interval wavelet transform to effectively suppress the boundary effect, a new wavelet transform domain Newton-LMS class algorithm. The simulation results show that the algorithm can give a more accurate and accurate estimation of the matrix for high and low frequency signals, and the convergence speed is faster than the traditional algorithms such as DFT-LMS and DCT-LMS.