Blind Equalization Using a Hybrid Algorithm of Multilayer Neural Network  

作　　者：梁启联 

机构地区：[1]Radio Engineering Department,Beijing University of Posts & Telecom.Beijing 100088,P.R.China

出　　处：《High Technology Letters》1996年第1期47-50,共4页高技术通讯（英文版）

基　　金：Supported by the National Natural Science Foundation of China;the High Technology Research and Development Programme of China

摘　　要：A novel blind equalization scheme based on multilayer neural network and Higher OrderCumulants(HOC)is proposed in the paper.The training of the neural network uses a newhybrid algorithm which has strict convex character(after a threshold)and converges muchfaster than the CMA algorithm.The inverse channel is built on the basis of the estimatedchannel and the training of neural network.The scheme can be used in nonlinear and timevarying channel and to deal with PAM or QAM signals.Simulation results Show that it per-forms well for blind equalization.A novel blind equalization scheme based on multilayer neural network and Higher Order Cumulants (HOC) is proposed in the paper. The training of the neural network uses a new hybrid algorithm which has strict convex character (after a threshold) and converges much faster than the CMA algorithm. The inverse channel is built on the basis of the estimated channel and the training of neural network. The scheme can be used in nonlinear and time varying channel and to deal with PAM or QAM signals. Simulation results show that it performs well for blind equalization.

关 键 词：BLIND EQUALIZATION Multilayer neural network HIGHER order CUMULANTS Hybrid algorithm CONVEX CHARACTER 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]