Alpha稳定分布噪声下改进的频域预滤波方法

Improved prefiltering method in frequency domain under Alpha stable distribution noise

针对α稳定分布噪声和谐波信号在频域均表现为异常值这一特性,提出了一种新的频域预滤波估计方法。通过分数低阶共变将信号转换到频域,在频域利用自适应加权Myriad滤波器滤除尖锐脉冲,提取稳定分布噪声的共变谱,将稳定分布有色噪声转化为稳定白噪声,然后利用基于分数阶共变的Music方法估计信号的共变谱。仿真结果表明,在不需要知道稳定噪声任何统计特性的情况下,该方法在1<α<2的非线性度量空间,该算法取得了理想的效果。

A new method of prefiltering in frequency domain is adopted in this paper.The main idea of this method is the observation that harmonics in α stable noise show themselves as outliers in frequency domain.The signal is transferred to frequency domain through fractional lower order eonvariation.In frequency domain,using adaptive weighted Myriad filter to remove outliers impulse from data,the convariation spectrum of stable noise can be gotten,stable color noise can be converted to stable white noise in farther.Last,the convariation spectrum of signal is estimated with music method based on fractional lower order convariation.The simulation results show that this method gets relatively perfect effect in 1〈α〈2 nonlinear metric space,without knowing any of statistical characteristics of α stable noise.