COMPARISON OF THE CONVERGENCE OF 2 ALGORITHMS FOR ADAPTIVE FIR DIGITAL-FILTERS - COMMENTS

被引：30

作者：

BERSHAD, NJ

机构：

[1] Univ of California, Irvine, CA, USA, Univ of California, Irvine, CA, USA

来源：

IEEE TRANSACTIONS ON ACOUSTICS SPEECH AND SIGNAL PROCESSING | 1985年 / 33卷 / 06期

关键词：

MATHEMATICAL TECHNIQUES - Algorithms - SIGNAL FILTERING AND PREDICTION;

D O I：

10.1109/TASSP.1985.1164743

中图分类号：

O42 [声学];

学科分类号：

070206 ; 082403 ;

摘要：

In a recent paper Claasen and Mecklenbraeuker compared the convergence behavior of two adaptive filter algorithms, the stochastic iteration algorithm (SIA) and the sign algorithm (SA). The present author: 1) comments on the discarding of the algorithm that clipped the input samples instead of the error, 2) indicates an alternative way to obtain g( epsilon ) for the Gaussian PDF in Table I, 3) corrects an error in (34), and 4) points out the availability of some recent results for the SIA for other than white input data.

引用

页码：1604 / 1606

页数：3