基于中值正交化准则的自适应滤波器

ADAPTIVE FILTERING ALGORITHM BASED ON MEDIAN ORTHOGONALITY CRITERION

α-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,本文提出了一种稳定分布环境下的基于中值正交化准则的自适应滤波算法。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 can better decrike encountered in the practical applications impulsive signal and noises of dislinct characteristics. This class of process has no close form of probability density function and finite second order moment different from other statislic models. In response existing independent Sets noise of system, authors proposed a new adaptive filtering algorithm based on median orthogonality criterion in the stable distribution environment. The simulation experiments showed that the proposed new algorithm is a filtering method with good toughness in the background of SαSG distribution noisy.