改进的NLMS算法在回声消除系统中的应用

发布时间：2018-03-11 03:07

本文选题：回声消除系统　切入点：滤波器阶数　出处：《计算机工程与应用》2017年10期 　论文类型：期刊论文

【摘要】：回声消除系统使用自适应滤波器模拟回声信道,自适应滤波器的阶数对回声消除效果具有重要的影响,合适的阶数可以提高回声消除效果。根据滤波器活跃区域权值可以模拟出基本的回声信道的原理,改进的NLMS算法首先将滤波器的初始权值阶数分帧,并求取每帧内权值的均值与抖动状态,然后通过比较阈值结果调整滤波器权值的活跃区域,摒弃非活跃权值区域,最终根据活跃权值的总帧数调整滤波器的阶数。改进的NLMS算法不仅保留了NLMS算法的结构简单、快速收敛的优点,而且实时调整滤波器阶数,降低稳态误差。通过回声消除仿真对比实验显示,算法的滤波器失准系数低于NLMS、PNLMS算法。
[Abstract]:The echo cancellation system uses adaptive filter to simulate the echo channel. The order of the adaptive filter has an important influence on the echo cancellation effect. According to the principle of the filter active region weight, the improved NLMS algorithm firstly frames the initial weight order of the filter. The mean and jitter states of the weights in each frame are obtained, and then the active region of the filter weights is adjusted by comparing the threshold results, and the inactive weights are discarded. Finally, the order of filter is adjusted according to the total frame number of active weight. The improved NLMS algorithm not only preserves the advantages of simple structure and fast convergence of NLMS algorithm, but also adjusts the order of filter in real time. The simulation results of echo cancellation show that the filter mismatch coefficient of the algorithm is lower than that of NLMS / PNLMS algorithm.
【作者单位】：华侨大学计算机科学与技术学院;
【基金】：国家自然科学基金(No.61271383) 2013年福建省战略性新兴产业发展专项(闽发改投资[2013]866号)
【分类号】：TN912.3

【相似文献】