非高斯脉冲噪声下的非线性自适应最小p范数迭代方法

NONLINEAR ADAPTIVE LEAST P-NORM FILTER IN NON-GAUSSIAN IMPULSIVE NOISE ENVIRONMENTS

稳定分布可以更好地描述实际应用中所遇到的具有显著脉冲特性的随机信号和噪声。与其它统计模型不同,α稳定分布没有统一闭式的概率密度函数,其二阶及二阶以上统计量均不存在。针对系统中存在独立SαS噪声与高斯噪声,本文提出了一种稳定分布环境下的非线性自适应最小p范数迭代算法。计算机模拟和分析表明,这种算法是一种在SαSG分布背景噪声条件下具有良好韧性的滤波方法。

Alpha stable distribution is better for modeling impulsive noises than Gaussian distribution in signal processing.This class of process has no close form of probability density function and finite second order moments.Authous proposed a new nonlinear adaptive least p-norm filter.The simulation experiments Showed that the proposed new algorithm was more robust than the conventional filtering algorithm.