实现CDMA最佳多用户检测的神经网络方法

提出并讨论了两种实现码分多址系统中最佳多用户检测的神经网络方法。一种基于离散Ｈｏｐｆｉｅｌｄ神经网络，另一种基于采用反向传播算法的多层感知器神经网络。理论分析和计算机模拟都证实了这两种神经网络方法的可行性，优越性和实用性。前者适用“固定”用户情况；后者既可用于“固定”用户吼适用于移动用户的情况，因而在未来的ＣＤＭＡ个人通信网中有较好的应用前景。

Neural Network Implementation of Optimum Multiuser Detectors in Code-division Multiple Access Channels

Two neural networks for implementing optimum multiuser detector (MUD) in code division multiple access(CDMA) channels are discussed. One is based on discrete Hopfield neural network (DHNN),andthe other is based on multilayer perceptron network using back propagation algorithms (BPNN), which areverified by theoretical analysis and computer simulation. The former is suitable for the case that the users inthe system are not mobile, while the later iS suitable for the case that the users might be mobile, such as in CDMA personal communication networks (PCN).