变步长LMS自适应滤波算法通过构造步长因子来进行权值调整, 使算法具有较快的收敛速度和较小的稳态误差。为了进一步改善算法的性能, 提出一种基于S函数的改进变步长LMS自适应算法。该算法基于S函数的曲线特点, 通过对函数的平移变换得到算法步长因子的表达式。为满足算法的可控性和抗干扰能力的要求, 通过引入可控参数和误差向量自相关值来调整步长因子, 得到算法的最终模型。详细分析了模型中各参数的取值对步长因子和滤波性能的影响。与现有算法的仿真结果对比表明,该算法在收敛速度、稳态误差及抗干扰能力方面的性能均有了很大的改善。
LMS自适应滤波 S函数 改进变步长 滤波性能 LMS adaptive filtering S function improved variable step-size filtering performance
Author Affiliations
Abstract
We investigate the nonvolatile holographic storage characteristics of near-stoichiometric LiNbO3:Fe:Mn crystals with different Li2O contents. Experimental results indicate that the optimal value of Li2O content is about 49.6 mol%. Nonvolatile sensitivity S′ considerably improved to 0.15 cm/J because of the use of near-stoichiometric LiNbO3:Fe:Mn with 49.6 mol% Li2O.
210.2860 Holographic and volume memories 190.5330 Photorefractive optics 090.7330 Volume gratings 160.5320 Photorefractive materials 160.3730 Lithium niobate Chinese Optics Letters
2012, 10(12): 122101
