论文部分内容阅读
综述了神经网络技术用于纠错码译码，设计好码与估计码的性能的研究现状，分析指出了已有工作中存在的突出问题．根据现代数字通信和信息存储系统的发展要求，展望了今后的研究重点与方向，包括研究高速有效的专用神经网络译码模型与算法并设计实用的神经网络译码器；研究神经网络译码器新思想并开展深入工作；探索神经网络ＴＣＭ技术与Ｔｕｒｂｏ码的神经网络迭代译码技术等．
The current research status of neural network technology for error correction code decoding, design of good code and estimated code performance is summarized. The outstanding problems existing in the existing work are pointed out. According to the development requirements of modern digital communication and information storage systems, the research priorities and directions are prospected in the future, including researching high-speed and efficient dedicated neural network coding models and algorithms and designing practical neural network coders; researching neural network coders New ideas and carry out in-depth work; to explore neural network TCM technology and Turbo code neural network iterative decoding technology.