一种改进的NLMS算法及其在回声消除中的应用

An Improved NLMS Algorithm and Its Application in Acoustic Echo Cancellation

针对回波抵消器中大量抽头系数需要更新的问题,提出了一种基于权系数局部迭代和集员滤波的NLMS算法。首先基于权系数瞬时梯度估计的Mmax系数局部迭代方法,在每次迭代中把幅度较大输入元素对应的权系数筛选出来进行更新。其次,为了进一步减少算法的运算量,引入了基于系数稀疏更新理论的集员滤波算法,该算法中只有当参数估计误差大于给定的误差门限时滤波器系数才进行迭代更新,从而有效地减少了滤波器系数的迭代次数。

In some adaptive filtering applications,the NLMS algorithm may be too computationally and memory intensive to implement.In this paper,a new NLMS algorithm based on set membership with partial-update is presented.The new algorithm allows the reduction of the frequency of updates of the filter coefficients,where the filter coefficients areupdated such that the output estimation error is upper bounded by a pre-determined threshold.Moreover,the work is the MMax tap-selection criterion in which,given a filter length N,only M coefficients are updated that correspond to the M largest magnitude elements of the regression vector.