一种基于Hopfield神经网络的伪Gray编码方法

A Method of Pseudo-Gray Coding Based on Hopfield Neural Network

伪Gray编码属信源—信道联合编码!可提供矢量量化器以零冗余的信道误码保护。本文运用Hopfield神经网络,构造适当能量函数,以求得伪Gray编码的次最优解;该Hopfield神经网络具有较简洁的形式,稍作修改,可求解简化信道模型下的伪Gray编码,且不要求知道信道的确切特性参数。仿真实验表明,本方法可以得到较好的结果。

A kind of Hopfield neural network is used by constructing suitable energy function to get sub-optimal solutions to pseudo-Gray coding. This kind of Hopfield neural network is easily modified to solve simplified version of pseudo-Gray coding for single-bit-error channel mod el. Simulating experiment results show that this method could offer good performances.