基于神经网络的(23,12)Golay码译码新算法

Neural Network Decoding to (23, 12) Golay Codes

对Hopfield神经网络（H.N.N）在纠错编译码技术中的具体应用进行了研究，分析指出了该网络与循环码，尤其是Golay码的关系，提出了一种采用高阶互连网络实现循环码最大似然译码的方法。并且实现了一套有效的算法，非常完满地解决了（23．12）Golay码的完全译码。最后指明了硬件实现的思路。

In this paper, it studies Error-Correct becoding technique based on Hopfield NeuralNetwork (H. N.V ). The relationship between H. N. N and cyclic codes (mainly Golay Codes) is stud-ied. Then we Put forward a method that can make maxlimum-likely decoding by use of high-ordermutual connecting networds. Moreover. a new effective algothim is designed which successfully com-plets the decnding to (23, 12)Golay codes sets. At the end of paper, the train of thought for hard-ware completion will be provided.