摘要
基于RBF网络的信道均衡器,其性能取决于神经网络的隐层参数。文中采用判决反馈均衡器的结构,RBF网络隐层的结构和参数则由基于小生境技术的混合递阶遗传算法来确定。仿真结果表明,这种RBF网络均衡器性能优良。
Performance of the RBFN-based channel equalizer depends on the RBF nonlinear hidden layer. This paper uses genetic algorithm to detemine the hidden layer＇s structure and parameters. The proposed genetic algorithm is based on niche and hybrid hierarchy. In addition, the equalizer is constructed with decision feedback configuration. Simulation results show that this equalizer has good performance.
出处
《信息技术》
2008年第8期96-98,112,共4页
Information Technology
