一种基于神经网络的自适应回波消除器  

A Adaptive Echo Canceller Based on Neural Network

作　　者：杨大力[1] 刘泽民[1] 

机构地区：[1]北京邮电学院145信箱,北京100088

出　　处：《通信学报》1993年第6期16-21,共6页Journal on Communications

基　　金：国家自然科学基金;教委博士基金

摘　　要：本文提出了一种基于多层查表神经网络的自适应回波消除器。这种新的回波消除器具有神经网络分布式并行信息处理、较好的容错特性和鲁棒效应等优点,且比传统的回波消除器有更好的收敛特性,并可方便地实现多级回波消除。A adaptive echo canceller based on the neural network with mutilayer lookup table and its algorithm are proposed in this paper. This novel echo canceller is of the advantages of neural network, such as the paralell distributed information processing, the fault-tolerance and the robustness. Compared with traditional structure, the echo canceller has better convergence property and may realize the multistage echo cancelleration for transmission system.

关 键 词：ISDN 回波消除 神经网络 通信网 

分 类 号：TN913.24[电子电信—通信与信息系统]