论文部分内容阅读
实际系统中存在大量的具有脉冲性质的非高斯噪声，本文介绍利用实数维低阶矩理论，根据最小离差准则设计自适应滤波器，去除一类服从稳定分布的噪声，并给出了基于LMAD算法的自适应滤波器．实验表明，这种滤波器去除SaS分布噪声的性能比基于LMS算法的滤波器优越。
There are a lot of impulsive non-Gaussian noise in the real system. In this paper, we introduce the theory of real dimension low order moments to design an adaptive filter according to the minimum deviation criterion to remove a class of noise obeying stable distribution. Algorithm of adaptive filter. Experiments show that the performance of this filter to remove the SaS distribution noise is superior to that based on the LMS algorithm.