改进的自适应小波变换在信号去噪中的应用

讨论了传统的LNS在信号处理中的不足。基于离散小波变换提出一种自适应去噪方法。该方法利用多分辨率分析理论，把信号和噪声分解在不同的频率范围之内，使信号和噪声的频谱得到简化，从而减少了自适应滤波器的级数，使输入自相关矩阵阶数降低。仿真结果表明，与传统的LMS算法相比，WLNS算法的收敛速度和稳定性得到了显著的提高。

The Application of Modified Self- adaptive Wavelet Transform in the Noise Elimination of Signal

This paper discusses on the weakness of conventional Least-Mean-Square(LMS) algorithm in real-time signal processing, presents a new self-adaptive noise eliminating method based on discrete wavelet transform, which analyzes the signal and noise into different frequency ranges with multi-resolution analysis and makes the spectrum of the signal and noise be simplified and therefore reduce the self-adaptive filter's orders and the self-correlative matrix's orders. The simulation results show that WLMS has higher convergence rate and better robustness than conventional LMS