遗传优化神经网络算法在信道盲均衡中的应用

摘要

在信道盲均衡问题的研究中,根据BP神经网络的信道肓均衡算法存在收敛速度慢,易陷入局部极小值的缺陷,导致信道肓均衡效果差,信道误码率高.为克服BP神经网络的缺陷,提高均衡道肓均衡效果和降低误码率,利用遗传算法全局搜索能力强的优点对BP神经网络的缺陷进行改进,提出一种基于遗传神经网络的信道盲均衡算法.采用BP神经网络构建信道分类器,通过遗传算法优化神经网络权值,最终实现盲均衡.仿真结果表明,相对于传统BP神经网络盲均衡算法,遗传神经网络算法收敛速度快,误码率降低,能获得更好的收敛特性和均衡效果.%The traditional BP neural networks have the characteristic of easily falling into local minimum point. In order to overcome the defect, a blind equalization algorithm based on genetic algorithm and BP neural network is put forward. Firstly, the global search ability of genetic algorithm is used to optimize the network weights, then the local search speed of BP algorithm is used to obtain the optimal weights of the network. Finally, the blind equalization is realized. Computer simulation results show that; compared with the blind equalization algorithm of traditional BP neural network, the new algorithm has a fast convergence speed, the steady residual error and low error rate is, and a-chieves better convergence properties and equilibrium effect.