针对LMS—HB自适应时间延迟估计方法在分数低阶α稳定分布噪声环境下的退化现象,依据分数低阶统计量理论,提出了基于分散系数最小化的LMP—HB自适应时延估计方法,并进一步提出了不依赖于参数估计的基于非线性变换的HB加权自适应时延估计方法。理论分析和计算机仿真结果表明,新方法在高斯和分数低阶α稳定分布噪声环境下具有良好的韧性。
According to the degeneration of LMS-HB adaptive time delay estimation under fractional lower order α-stable noise conditions,this paper proposes a novel adaptive time delay estimation algorithm called LMP-HB based on the minimum dispersion criterion. Furthermore,in order to avoid the estimation of character parameter α, another novel adaptive time delay estimation algorithm is developed based on nonlinear transform. Theoretical analysis and computer simulations show that new algorithms perform robust under both Gaussian and fractional lower order α-stable noise conditions.