光通信系统中LDPC码的代数构造方法研究
发布时间：2018-07-13 16:16
【摘要】：随着社会信息化进程的不断推进，现有的各种视频业务以及刚刚商用化的4G对带宽需求剧增，进而对光通信系统的传输容量需求也不断增加。如何实现高可靠性、长距离、大容量、低成本的传输是高速光传输技术面临的最大挑战。信道编码技术是一种提高传输系统可靠性的有效方法。而低密度奇偶校验（Low-densityParity-check, LDPC）码作为信道编码技术中一种极其重要的参考码型，凭借其良好的纠错性能，已成为近年来信道编码技术研究的热点。准循环低密度奇偶校验（Quasi-cyclic Low-density Parity-check, QC-LDPC）码又是一种具有码长、码率选择更灵活、编码复杂度更低、在光通信系统中具有更好应用前景的LDPC码。因此，本文对光通信系统中LDPC码的代数构造方法进行了深入的研究。 本论文的选题来源于重庆市自然科学基金项目“高速光通信系统中SFEC码型的新颖构造机理研究”，（CSTC,2010BB2409）。主要完成了以下几项工作： （1）阐述了LDPC码的基础理论原理并对LDPC码的编译码理论进行了探讨，仿真验证了影响LDPC码纠错性能的主要因素。并对光通信系统中主要模块进行噪声分析，基于MATLAB搭建了光通信系统仿真模型，为接下来对LDPC码的代数构造方法研究以及性能仿真分析做好坚实的铺垫工作。 （2）在考虑光通信系统高码率要求的基础上，结合LDPC码代数构造方法中有限域乘法群以及循环子群的基本特性，提出了QC-LDPC码的一种新颖构造方法。并运用该新颖构造方法，构造出一种适用于光通信系统高码率要求的规则QC-LDPC（5334,4955）码。仿真分析表明：在码率为0.93时，利用此构造方法所构造出的规则QC-LDPC（5334,4955）码的纠错性能明显要优于ITU-T G.709光通信系统标准的RS（255,239）码以及ITU-T G.975.1中的LDPC（32640,30592）码。在BER=10-6时，QC-LDPC（5334,4955）码距离香农极限约1.64dB，其净编码增益（Net Coding Gain, NCG）比RS（255,239）码提高了约1.61dB，同时比LDPC（32640,30592）码提高了约0.9dB，更适用于光通信系统。 （3）在进一步研究有限域的基本特性之后，结合有限域中元素的逆元特性，设计出一种满足行列约束条件的LDPC码构造方法，得到一种新颖的校验矩阵结构。结合有限域构造QC-LDPC码的方法，造出一种适用于光通信系统高码率要求的规则QC-LDPC（5334,4962）码。仿真分析表明：在码率为0.93时，利用此构造方法所构造出的规则QC-LDPC（5334,4962）码的纠错性能明显要优于RS（255,239）码。在BER=10-6时，QC-LDPC（5334,4962）码距离香农极限约1.48dB，其净编码增益比RS（255,239）码提高了约1.77dB，，同时比LDPC（32640,30592）码提高了约1.06dB。该方法构造的规则QC-LDPC（5334,4962）码可考虑作为应用于光通信系统中的超强前向纠错码型中的一种候选码型。 （4）在完成基于有限域的规则QC-LDPC码构造之后，本文又进一步探究基于修饰技术的QC-LDPC码改进方案。提出一种基于计算机搜索基本矩阵中对六环数目影响权重最大的节点，利用修饰技术对权重大的某个点或者某些点进行“置零”操作，达到减少校验矩阵中六环的数目，进而提高构造出LDPC码型纠错性能的修饰矩阵选择方案。利用本文所提出的修饰方案，对本文所提出的基于有限域构造的校验矩阵进行多次修饰改进后可知：第一次修饰和第二次修饰均有效的减少了校验矩阵中六环的数目，并改善了新的校验矩阵所构造的QC-LDPC码型的纠错性能，其中经过第一次修饰之后，基础矩阵中六环的数目减少了128个，在误码率为10-6时，采用修饰技术后得到新的校验矩阵构造出的非规则QC-LDPC（5334,4962）-masking-1码型比修饰前的规则QC-LDPC（5334,4962）码型提高约了0.23dB的NCG，并且修饰后校验矩阵构造的码型保持了与修饰前构造码型相等的码率，仿真验证了本文所提出的修饰方案的可行性及其对修饰技术的研究具有重要的参考意义。
[Abstract]:With the progress of the social information process, the existing video services and the newly commercialized 4G have increased the bandwidth demand, and the demand for the transmission capacity of the optical communication system is also increasing. How to achieve high reliability, long distance, large capacity and low cost transmission is the biggest challenge for high-speed optical transmission technology. Channel coding is the biggest challenge. Code technology is an effective method to improve the reliability of transmission systems, and Low-densityParity-check (LDPC) code is an extremely important reference code type in channel coding technology. With its good error correction performance, it has become a hot spot in the research of channel coding techniques in recent years. Quasi -cyclic Low-density Parity-check, QC-LDPC) code is a kind of LDPC code with long code length, more flexible selection of bit rate, lower coding complexity and better application foreground in optical communication system. Therefore, this paper deeply studies the algebraic construction method of LDPC code in optical communication system.
The topic of this paper comes from the Chongqing Natural Science Foundation Project "Research on the novel structure mechanism of SFEC code in high speed optical communication system" (CSTC, 2010BB2409). The following work is completed mainly:
(1) the basic theory principle of LDPC code is expounded and the coding and decoding theory of LDPC code is discussed. The main factors affecting the error correction performance of LDPC code are verified by simulation. The noise analysis of the main modules in the optical communication system is analyzed, and the simulation model of the optical communication system is built on the basis of MATLAB, and the algebraic construction method of LDPC code is studied for the following. And performance simulation analysis to make a solid paving work.
(2) on the basis of considering the high bit rate requirement of optical communication system, and combining the basic properties of the finite field multiplication group and the cyclic subgroup in the LDPC code algebra construction method, a novel construction method of QC-LDPC code is proposed, and a rule QC-LDPC (53344955) for high bit rate requirement of optical communication system is constructed by using the new construction method. The simulation analysis shows that the error correction performance of the rule QC-LDPC (53344955) code constructed by this method is obviously superior to the RS (255239) code of the ITU-T G.709 optical communication system standard and the LDPC (3264030592) code in ITU-T G.975.1. At BER= 10-6, the QC-LDPC (53344955) code distance to Shannon limit is about 1.64dB, Its net coding gain (Net Coding Gain, NCG) is about 1.61dB higher than RS (255239) code, and increases about 0.9dB over LDPC (3264030592) codes, and is more suitable for optical communication systems.
(3) after further studying the basic characteristics of the finite field and combining the inverse element characteristics of the elements in the finite field, a LDPC code construction method which satisfies the column constraint conditions is designed, and a novel check matrix structure is obtained. A rule QC-L suitable for the high bit rate requirement of optical communication system is created with the method of constructing the QC-LDPC code in the finite field. DPC (53344962) code. The simulation results show that the error correction performance of the rule QC-LDPC (53344962) code constructed by this method is obviously better than that of RS (255239) code when the code rate is 0.93. At BER=10-6, the QC-LDPC (53344962) code is about 1.48dB at the Shannon limit, and its net coding gain is about 1.77dB and is more than LDPC (255239) code. 3264030592) the code QC-LDPC (53344962) code which is constructed by the code about 1.06dB. can be considered as a candidate for the super strong forward error correcting code type applied to the optical communication system.
(4) after the construction of the regular QC-LDPC code based on the finite field, this paper further explores the QC-LDPC code improvement scheme based on the modification technology, and proposes a node with the maximum weight affecting the number of six rings based on the computer search matrix, and uses the modification technique to "zero" operation on a point or point of large weight. In order to reduce the number of six rings in the check matrix, and then improve the modified matrix selection scheme that constructs the error correction performance of the LDPC code, the modified scheme proposed in this paper has been improved by the modification of the check matrix based on the finite domain structure proposed in this paper. The first and second modifications are effectively reduced. The number of six rings in the matrix is checked and the error correction performance of the QC-LDPC code type constructed by the new check matrix is improved. After the first modification, the number of six rings in the basic matrix is reduced by 128. When the error rate is 10-6, the non regular QC-LDPC (53344962) -masking-1 constructed by the new check matrix is obtained after the modification technique is used. The rule QC-LDPC (53344962) code pattern before the code shape ratio improves about 0.23dB's NCG, and the code pattern constructed by the modified check matrix keeps the code rate equal to the pre modified structure code. The simulation proves the feasibility of the modified scheme and it has important reference significance for the study of the modification technology.
【学位授予单位】：重庆邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN929.1;TN911.22
本文编号：2120012
[Abstract]:With the progress of the social information process, the existing video services and the newly commercialized 4G have increased the bandwidth demand, and the demand for the transmission capacity of the optical communication system is also increasing. How to achieve high reliability, long distance, large capacity and low cost transmission is the biggest challenge for high-speed optical transmission technology. Channel coding is the biggest challenge. Code technology is an effective method to improve the reliability of transmission systems, and Low-densityParity-check (LDPC) code is an extremely important reference code type in channel coding technology. With its good error correction performance, it has become a hot spot in the research of channel coding techniques in recent years. Quasi -cyclic Low-density Parity-check, QC-LDPC) code is a kind of LDPC code with long code length, more flexible selection of bit rate, lower coding complexity and better application foreground in optical communication system. Therefore, this paper deeply studies the algebraic construction method of LDPC code in optical communication system.
The topic of this paper comes from the Chongqing Natural Science Foundation Project "Research on the novel structure mechanism of SFEC code in high speed optical communication system" (CSTC, 2010BB2409). The following work is completed mainly:
(1) the basic theory principle of LDPC code is expounded and the coding and decoding theory of LDPC code is discussed. The main factors affecting the error correction performance of LDPC code are verified by simulation. The noise analysis of the main modules in the optical communication system is analyzed, and the simulation model of the optical communication system is built on the basis of MATLAB, and the algebraic construction method of LDPC code is studied for the following. And performance simulation analysis to make a solid paving work.
(2) on the basis of considering the high bit rate requirement of optical communication system, and combining the basic properties of the finite field multiplication group and the cyclic subgroup in the LDPC code algebra construction method, a novel construction method of QC-LDPC code is proposed, and a rule QC-LDPC (53344955) for high bit rate requirement of optical communication system is constructed by using the new construction method. The simulation analysis shows that the error correction performance of the rule QC-LDPC (53344955) code constructed by this method is obviously superior to the RS (255239) code of the ITU-T G.709 optical communication system standard and the LDPC (3264030592) code in ITU-T G.975.1. At BER= 10-6, the QC-LDPC (53344955) code distance to Shannon limit is about 1.64dB, Its net coding gain (Net Coding Gain, NCG) is about 1.61dB higher than RS (255239) code, and increases about 0.9dB over LDPC (3264030592) codes, and is more suitable for optical communication systems.
(3) after further studying the basic characteristics of the finite field and combining the inverse element characteristics of the elements in the finite field, a LDPC code construction method which satisfies the column constraint conditions is designed, and a novel check matrix structure is obtained. A rule QC-L suitable for the high bit rate requirement of optical communication system is created with the method of constructing the QC-LDPC code in the finite field. DPC (53344962) code. The simulation results show that the error correction performance of the rule QC-LDPC (53344962) code constructed by this method is obviously better than that of RS (255239) code when the code rate is 0.93. At BER=10-6, the QC-LDPC (53344962) code is about 1.48dB at the Shannon limit, and its net coding gain is about 1.77dB and is more than LDPC (255239) code. 3264030592) the code QC-LDPC (53344962) code which is constructed by the code about 1.06dB. can be considered as a candidate for the super strong forward error correcting code type applied to the optical communication system.
(4) after the construction of the regular QC-LDPC code based on the finite field, this paper further explores the QC-LDPC code improvement scheme based on the modification technology, and proposes a node with the maximum weight affecting the number of six rings based on the computer search matrix, and uses the modification technique to "zero" operation on a point or point of large weight. In order to reduce the number of six rings in the check matrix, and then improve the modified matrix selection scheme that constructs the error correction performance of the LDPC code, the modified scheme proposed in this paper has been improved by the modification of the check matrix based on the finite domain structure proposed in this paper. The first and second modifications are effectively reduced. The number of six rings in the matrix is checked and the error correction performance of the QC-LDPC code type constructed by the new check matrix is improved. After the first modification, the number of six rings in the basic matrix is reduced by 128. When the error rate is 10-6, the non regular QC-LDPC (53344962) -masking-1 constructed by the new check matrix is obtained after the modification technique is used. The rule QC-LDPC (53344962) code pattern before the code shape ratio improves about 0.23dB's NCG, and the code pattern constructed by the modified check matrix keeps the code rate equal to the pre modified structure code. The simulation proves the feasibility of the modified scheme and it has important reference significance for the study of the modification technology.
【学位授予单位】：重庆邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN929.1;TN911.22
【参考文献】
相关期刊论文 前3条
1 袁建国;叶文伟;毛幼菊;;光通信系统中基于LDPC码的SFEC码型研究[J];光电子.激光;2009年11期
2 袁建国;许亮;黄胜;王永;;光通信中一种基于有限域循环子群的QC-LDPC码构造方法[J];半导体光电;2013年06期
3 张晓宏;易小波;刘翔;Peter J.Winzer;;ALU 1006传送技术以及向T比特时代的演进(英文)[J];中国通信;2013年04期
本文编号：2120012
本文链接：https://www.wllwen.com/kejilunwen/wltx/2120012.html
