基于陷阱集本体论的 LDPC 译码算法研究-计算机系统结构专业论文.docxVIP

摘 要 LDPC 码(Low Density Parity Check Code，LDPC)是由 Robert G.Gallager 博士于 1963 年提出的一类具有稀疏校验矩阵的线性分组码，不仅有逼近 Shannon 限的良 好性能，而且译码复杂度较低, 结构灵活，是近年信道编码领域的研究热点，目前 已广泛应用于深空通信、光纤通信、卫星数字视频和音频广播等领域。 本文对 LDPC 码中的陷阱集本体论进行了研究，并在陷阱集本体论相关理论 的基础上，给出了一种较 BP 译码算法更简单更有效的新译码方法。本文的主要工 作概括如下： 1. 系统地总结了现代编码理论的发展历程；概述了 LDPC 码及其 Tanner 图模 型的理论基础；详细分析了消息传播机制及置信传播译码算法。 2. 详细介绍了陷阱集以及陷阱集本体论的概念；介绍了陷阱集本体论在编译 码方面的应用；依据陷阱集本体论，给出了一个陷阱集数据库生成算法，该数据 库可以用来评估一个陷阱集的危害性大小。 3. 设计实现了多比特译码算法，该算法在校验节点和变量节点之间传递的信 息上添加了一定比特信息位，这些添加的比特位能够传递节点的局部邻居信息， 从而使译码算法跳出相应的陷阱集，译码成功。仿真结果表明这种算法简单有效。 关键词：低密度校验码 迭代译码 陷阱集 陷阱集本体论 Abstract LDPC (Low Density Parity Check Code, LDPC) is a class of sparse check matrix of a linear block code proposed by Dr. Robert G. Gallager in 1963. Because of its near Shannon limit performance, low decoding complexity and flexible structure, LDPC has become a research hotspot in the field of channel coding in recent years. It has been widely used in deep space communication, optical fiber communication, satellite digital video and audio broadcasting. This paper conducts a study of trapping set ontology. On the basis of the trapping set ontology theory, this paper gives a simpler and more effective new decoding method than BP decoding algorithm. The main work of this paper is summarized as follows: The course of development of modern coding theory has been systematically summarized. LDPC code Tanner graph model and its theoretical basis are briefly illuminated. Message passing mechanism and belief propagation decoding algorithm are analyzed in detail. The concept of trapping set and trapping set ontology are thoroughly analyzed. The application of trapping set ontology in the encoding and decoding is introduced. Based on trapping set ontology, a trapping set database generation algorithm is given. The database can be used to evaluate the size of the danger of a trapping set. A multi-bit decoding algorithm is design and realized. This algorithm adds some bits to the information passed through check node