论文部分内容阅读
本文根据克服数字通信中码间干扰（ISI）的最佳均衡解一般表达式，提出了一种新的自适应神经网络均衡器结构，然后导出了基于该结构的一种自适应算法和相应的学习规则，最后对提出的自适应神经网络均衡器性能进行了计算机模拟，模拟结果与分析表明：本文提出的神经网络均衡器用于实现最佳信道均衡非常有效，比传统线性均衡器和Gibson等人[1]提出的多层感知均衡器（MLPE）性能更优越，更具实用性．
In this paper, a new adaptive neural network equalizer structure is proposed based on the best general solution to overcome the intersymbol interference (ISI) in digital communications. Then an adaptive algorithm based on the structure and the corresponding Learning rules. Finally, the performance of the proposed adaptive neural network equalizer is simulated. The simulation results and analysis show that the proposed neural network equalizer is very effective in achieving the best channel equalization. Compared with the traditional linear equalizer and Gibson et al [1] proposed multi-layer sensor equalizer (MLPE) superior performance, more practical.