论文部分内容阅读
本文提出一种非线性码神经网络译码方案,在纠错能力范围内对满足码距特性的一般非线性码以零错误概率进行纠错译码,并在检错能力范围内检错。文中具体描述了神经网络模型构造、学习算法及其理论依据。最后通过非线性等重码的译码实例表明此方案的有效性及理论和应用价值。
This paper presents a non-linear code neural network decoding scheme, in the error correction capability to meet the code-distance characteristics of the general non-linear code with error probability of error-correction decoding and error detection within the error detection. The paper describes the neural network model construction, learning algorithm and its theoretical basis. Finally, the decoding example of non-linear isobar code shows the effectiveness of this scheme and its theoretical and practical value.