基于Hopfield神经网络多用户检测技术的研究

最优多用户检测器的优化问题可以映射为Hopfield神经网络的能量函数的最小化问题。本文对基于典型Hopfield神经网络多用户检测器和基于随机Hopfield神经网络多用户检测器进行了性能分析,计算机仿真表明这两种多用户检测器均具有抗多址干扰和抗远近效应的能力强、运算量小、实时性好等优点。

The Research on the Multi-User Detection Technology Based on Hopfield Neural Network

The optimization problem of the optimal multi-user detection can be transformed into the minimum of the Hopfield neural network energy function. This paper analyses the performance of the multi-user detection based on the model Hopfield neural network and the multi-user detection based on the stochastic Hopfield neural network. The computer simulation results show that they have good performances on multiple-access interference resistance and near-far effect resistance,much lower computational complexity, good real-time performance.