一种分阶自适应Volterra滤波算法

A novel divided-order adaptive Volterra filtering algorithm

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噪声。这种噪声的二阶及二阶以上统计量均不存在,需要用分数低阶统计量描述。针对Volterra级数非线性项将稳定分布的尖峰脉冲特性更加放大,导致输入信号自相关矩阵的特征值扩展更大的问题,本文提出了一种对于二阶Volterra级数的线性项部分和非线性部分分别采用两个不同收敛因子的分阶Volterra滤波器最小平均p范数(DOVLMP)算法,并分析证明了该算法的收敛性能。仿真结果验证了本文方法较传统算法的优越性。

In many practical applications,α-stable distribution can model impulsive random noise more efficiently.This class of noises has no finite second and higher order statistics,so fractional lower order statistics are used to describe them.For the nonlinear components of the input signal amplify the damaging effects of the impulse noise and eigen values of the correlation matrices of input signals spread further,this paper proposes a novel divided-order adaptive Volterra least mean p norm（DOVLMP） filtering algorithm,which adopts two different convergence factors for linear items and nonlinear items respectively,and analyzes the convergence performance of this algorithm.Computer simulation results illustrate that the performance of the proposed algorithm is superior to conventional algorithm.