论文部分内容阅读
本文提出了一种源于汉明类多层前向神经网络分组码译码器。它不需要改变网络结构和参数,根据输入可完成硬判决、软判决和最小距离译码。该网络不存在漫长的学习过程。计算机模拟表明,在加性高斯噪声下,使用该神经网络可以达到最大似然译码。
This paper presents a Hamming-based multi-layer forward neural network block code decoder. It does not need to change the network structure and parameters, according to the input to complete the hard decision, soft decision and minimum distance decoding. The network does not have a long learning process. Computer simulation shows that the maximum likelihood decoding can be achieved using this neural network under additive Gaussian noise.