基于集群型纠错的两比特BF译码算法研究

发布时间：2018-02-16 17:02

本文关键词： 陷阱集陷阱集配置两比特BF算法置信传播译码集群型纠错　出处：《西安电子科技大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着现代信息技术的快速发展,对数字通信系统的要求越来越高。在通信编码领域出现了一种低密度奇偶校验(Low-Density Parity-Check,简称LDPC)码,该码字具有低复杂度、纠错能力强等特点,且经过置信传播(Belief Propagation)迭代译码,它的信道容量能够逼近香农限。在现代数字通信系统中,纠错码的译码性能高低对整个通信系统影响极大,在很多实际应用中LDPC码的误码率要求在1210-左右甚至更低,因此如何有效提高LDPC码的译码性能成为当前研究的一大热点。通常提高译码性能主要从两个方面进行研究,一是提高译码速率,二是降低译码的误码率。本文主要对LDPC码的比特翻转译码算法和码字Tanner图中的错误类型进行了研究,给出了两比特“比特翻转”(Two-bit Bit Flipping,简称TBF)译码算法、陷阱集配置(Trapping Set Profile)的定义以及陷阱集配置的构造,针对不同的陷阱集配置对TBF译码算法的选择方案进行了研究。以下是本文的主要工作内容概括:1.系统地介绍了数字通信系统的发展以及LDPC码的研究现状;详细描述了LDPC码的Tanner图表示和陷阱集(Trapping Set)的相关概念;概括分析了LDPC码的两种典型的译码算法。2.在比特翻转(Bit Flipping)译码算法的基础上,针对该算法对某些错误类型译码不能够成功收敛的问题,在变量节点处额外添加一比特位信息,进而给出两比特变量节点的BF译码算法。实例验证该译码算法在若干次迭代后,译码成功收敛。3.在两比特变量节点的BF译码算法基础上,分析另一类错误类型,发现TBFA1(Two-bit Bit Flipping Algorithm 1)对该错误类型译码失败的问题,在校验节点处额外添加一比特位信息,进而给出两比特校验节点的BF译码算法(TBFA2),同时实例验证了TBFA2在若干次迭代后,译码成功收敛。再次结合算法分析给出了算法译码成功的收敛条件,实现了在BSC信道下TBF(Two-bit Bit Flipping)译码算法的仿真,结果表明该译码算法的性能优于BP译码算法的性能。4.针对不同的错误类型,结合集群型纠错(Collective Error Correction)给出了陷阱集和陷阱集配置的定义及构造,该构造是通过在子图中添加变量节点进行扩展的过程,进而得到一个不含陷阱集的图集合序列,通过依次递归可得到一个码字Tanner图的陷阱集配置。5.针对不同的错误类型,给出消除码字Tanner图中的陷阱集配置的TBF算法选择方案。分析选择单个TBF算法和选择多重TBF算法异同,给出有效的TBF算法选择方案。最后基于不同的陷阱集配置,实现了选择多重TBF译码算法的仿真,结果验证了该方法的有效性。
[Abstract]:With the rapid development of modern information technology, the requirement of digital communication system is becoming more and more high. In the field of communication coding, a low-density parity check (LDPC) code has emerged, which has the characteristics of low complexity and strong error-correcting ability. After iterative decoding with confidence propagation propagation, its channel capacity can approach Shannon limit. In modern digital communication systems, the decoding performance of error-correcting codes has a great impact on the whole communication system. In many practical applications, the BER of LDPC codes is about 1210- or even lower, so how to effectively improve the decoding performance of LDPC codes has become a hot research topic. Usually, the improvement of decoding performance is mainly studied from two aspects. One is to improve the decoding speed, the other is to reduce the bit error rate of decoding. This paper mainly studies the bit flipping decoding algorithm of LDPC code and the error types in the codeword Tanner diagram, and gives a two-bit "bit flipping" two-bit Bit coding algorithm. The definition of trap set configuration trapping Set profile and the construction of trap set configuration, The selection scheme of TBF decoding algorithm for different trap set configurations is studied. The following is the main work of this paper: 1. The development of digital communication system and the research status of LDPC code are introduced systematically. The concepts of Tanner graph representation of LDPC code and trap set trapping set are described in detail, and two typical decoding algorithms of LDPC code. In order to solve the problem that the algorithm can not converge successfully for some error type decoding, an additional bit bit information is added to the variable node, and then the BF decoding algorithm of two bit variable nodes is given. An example is given to verify the algorithm after several iterations. On the basis of the BF decoding algorithm of two-bit variable nodes, another kind of error type is analyzed. It is found that TBFA1(Two-bit Bit Flipping Algorithm 1) fails to decode the error type, and adds an extra bit of bit information at the check node. Furthermore, the BF decoding algorithm of two-bit check node is given, and an example is given to verify the successful convergence of TBFA2 decoding after several iterations. Again, the convergence conditions of the algorithm are given based on the analysis of the algorithm. The simulation of TBF(Two-bit Bit coding algorithm in BSC channel is carried out. The results show that the performance of this decoding algorithm is better than that of BP decoding algorithm. In this paper, the definition and construction of trap set and trap set configuration are given in combination with collective Error correction of cluster error correction. The structure is extended by adding variable nodes to the subgraph, and a graph set sequence without trap set is obtained. The trap set configuration of a codeword Tanner graph can be obtained by recursion in turn. For different error types, a TBF algorithm selection scheme for eliminating trap set configuration in codeword Tanner graph is presented. The similarities and differences between selecting single TBF algorithm and selecting multiple TBF algorithm are analyzed. An effective TBF algorithm selection scheme is presented. Finally, the simulation of selecting multiple TBF decoding algorithm is implemented based on different trap set configurations, and the effectiveness of the method is verified.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.22

【相似文献】