摘要
高速数据传输中信道均衡算法大致分为三类：即基于LMS准则、LS准则和人工神经网络的自适应算法。本文研究了这三类算法，并对部分算法作了改进，以加快收敛速度，减小稳态失调。文中的算法均用28．8KModem芯片设计的仿真系统产生的数据进行了性能测试，并给出了仿真曲线及实验结果。
Algorithms currently used for high speed data transmission over voice-band channels can be based on following three principles: on LMS criterion, on LS criterion or on artificial neural network. This paper investigates these algorithms, improves some algorithms to make them converge faster and with higher accuracy. The performance of these algorithms are also examined using the data generated by a simulation system which is configured with 28.8 K Modem chips.
出处
《电路与系统学报》
CSCD
1999年第2期10-16,共7页
Journal of Circuits and Systems
关键词
自适应均衡算法
BP神经网络
高速数据传输
Adaptive equalization algorithms, BP neural network, Convergence speed, Steady-state misadjustment
