LDPC码构造方法和编码算法研究
发布时间：2018-11-03 09:22
【摘要】：自从Turbo码被发明,LDPC码被重新发现以来,LDPC码在世界各国学术界和.IT界引起了广泛关注,成为信道编码领域的研究热点。LDPC码是性能最接近香农限的一种信道编码方案,同时它具有描述简单,错误平层低,可以采用迭代算法进行译码等优点。具有准循环结构的LDPC码可以实现快速编码,并行译码,有利于硬件实现。这些优点使得LDPC码具有巨大的应用潜能和良好的应用前景。Turbo码已经被3G标准采用,LDPC码具有比Turbo码更好的性能,且实现复杂度更低,是第四代移动通信系统信道编码方案强有力的竞争者。最新的无线通信协议802.Had已采用LDPC码作为信道编码方案,实现60GHz高速率数据传输。 本文的研究工作主要围绕LDPC码构造方法和编码方法两个核心主题展开。同时,针对卫星通信和地震勘探数据传输两种不同应用环境与具体要求进行了应用研究。 在构造方法方面,主要研究LDPC准循环构造法,原模图构造法和基于原模图的终止LDPC卷积码构造法。其中第三种构造方法是本文的研究重点和创新点所在。对于第三种构造方法,解决了矩阵构造过程中最关键的边扩展方式问题和基矩阵扩展算法问题。国外学者对边扩展技术并未给出统一可行的方法和充足的理论分析与证明,也没有明确提出易于编译码实现的基础矩阵扩展算法。本文分析了不同边扩展方式对BP译码性能的影响；将有条件约束的PEG算法应用于基于原模图的终止LDPC卷积码构造。最后将所得成果应用于R4JA原模图,构造出了一种性能良好的新型空间耦合码。在编码方法方面,主要研究并实现了G矩阵编码算法,准循环编码算法,802.11n编码算法和LDPC卷积码编码算法。 通信环境不同,对误码率,编码效率,编码时延等要求也不尽相同,本文针对两种应用环境提供合适的LDPC码型与编码算法。最后,基于windows平台,应用QT软件开发技术,将所研究的构造方法和编码方法集成了一个LDPC工具包。
[Abstract]:Since the invention of Turbo codes and the rediscovery of LDPC codes, LDPC codes have attracted wide attention in academic circles and. IT circles all over the world, and have become a research hotspot in the field of channel coding. LDPC codes are a channel coding scheme whose performance is closest to Shannon's limit. At the same time, it has the advantages of simple description, low error level, and can be decoded by iterative algorithm. The LDPC code with quasi-cyclic structure can realize fast coding and parallel decoding, which is beneficial to hardware implementation. These advantages make LDPC codes have great application potential and good application prospects. Turbo codes have been adopted by 3G standards, LDPC codes have better performance than Turbo codes, and the implementation complexity is lower. Is the fourth generation mobile communication system channel coding scheme strong competitor. The latest wireless communication protocol 802.Had has adopted LDPC code as channel coding scheme to realize 60GHz high rate data transmission. The research work of this paper mainly focuses on two core topics of LDPC code construction method and coding method. At the same time, two different application environments and specific requirements of satellite communication and seismic exploration data transmission are studied. In terms of construction methods, the LDPC quasi-cyclic construction method, the primitive pattern construction method and the terminating LDPC convolutional code construction method based on primitive mode graph are mainly studied. The third construction method is the focus and innovation of this paper. For the third method, the most critical problem of edge expansion and the algorithm of base matrix expansion are solved in the process of matrix construction. Foreign scholars have not given a unified and feasible method and sufficient theoretical analysis and proof to the edge expansion technology, nor have they clearly put forward the basic matrix expansion algorithm which is easy to implement by encoding and decoding. In this paper, the influence of different edge expansion modes on the decoding performance of BP is analyzed, and the conditional constrained PEG algorithm is applied to the construction of terminated LDPC convolutional codes based on primitive mode graph. Finally, a new spatial coupling code with good performance is constructed by applying the obtained results to the R4JA primitive mode diagram. In terms of coding methods, the G matrix coding algorithm, the quasi-cyclic coding algorithm, the 802.11n coding algorithm and the LDPC convolutional code coding algorithm are mainly studied and implemented. Different communication environments have different requirements for bit error rate, coding efficiency, coding delay and so on. This paper provides appropriate LDPC code types and coding algorithms for two application environments. Finally, based on windows platform and QT software development technology, a LDPC toolkit is integrated with the methods of construction and coding.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.22
本文编号：2307368
[Abstract]:Since the invention of Turbo codes and the rediscovery of LDPC codes, LDPC codes have attracted wide attention in academic circles and. IT circles all over the world, and have become a research hotspot in the field of channel coding. LDPC codes are a channel coding scheme whose performance is closest to Shannon's limit. At the same time, it has the advantages of simple description, low error level, and can be decoded by iterative algorithm. The LDPC code with quasi-cyclic structure can realize fast coding and parallel decoding, which is beneficial to hardware implementation. These advantages make LDPC codes have great application potential and good application prospects. Turbo codes have been adopted by 3G standards, LDPC codes have better performance than Turbo codes, and the implementation complexity is lower. Is the fourth generation mobile communication system channel coding scheme strong competitor. The latest wireless communication protocol 802.Had has adopted LDPC code as channel coding scheme to realize 60GHz high rate data transmission. The research work of this paper mainly focuses on two core topics of LDPC code construction method and coding method. At the same time, two different application environments and specific requirements of satellite communication and seismic exploration data transmission are studied. In terms of construction methods, the LDPC quasi-cyclic construction method, the primitive pattern construction method and the terminating LDPC convolutional code construction method based on primitive mode graph are mainly studied. The third construction method is the focus and innovation of this paper. For the third method, the most critical problem of edge expansion and the algorithm of base matrix expansion are solved in the process of matrix construction. Foreign scholars have not given a unified and feasible method and sufficient theoretical analysis and proof to the edge expansion technology, nor have they clearly put forward the basic matrix expansion algorithm which is easy to implement by encoding and decoding. In this paper, the influence of different edge expansion modes on the decoding performance of BP is analyzed, and the conditional constrained PEG algorithm is applied to the construction of terminated LDPC convolutional codes based on primitive mode graph. Finally, a new spatial coupling code with good performance is constructed by applying the obtained results to the R4JA primitive mode diagram. In terms of coding methods, the G matrix coding algorithm, the quasi-cyclic coding algorithm, the 802.11n coding algorithm and the LDPC convolutional code coding algorithm are mainly studied and implemented. Different communication environments have different requirements for bit error rate, coding efficiency, coding delay and so on. This paper provides appropriate LDPC code types and coding algorithms for two application environments. Finally, based on windows platform and QT software development technology, a LDPC toolkit is integrated with the methods of construction and coding.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.22
【参考文献】
相关期刊论文 前6条
1 郭建;刘光鼎;;无缆存储式数字地震仪的现状及展望[J];地球物理学进展;2009年05期
2 董树文;李廷栋;高锐;吕庆田;吴珍汉;陈宣华;周琦;刘刚;刘志强;梅琳;;地球深部探测国际发展与我国现状综述[J];地质学报;2010年06期
3 乔华;管武;董明科;项海格;;一种基于循环移位矩阵的LDPC码构造方法[J];电子与信息学报;2008年10期
4 张建国;LDPC码的应用研究[J];通信技术;2003年11期
5 杨建平;陈庆春;;IEEE802.16e标准LDPC译码器设计与实现[J];通信技术;2010年05期
6 龚险峰;陶孝锋;邱乐德;;基于原模图扩展的QC-LDPC构造方法[J];中国空间科学技术;2012年02期
，本文编号：2307368
本文链接：https://www.wllwen.com/kejilunwen/wltx/2307368.html
