应用于空间通信的LDPC码编译码研究与实现

发布时间：2018-03-31 08:05

本文选题：低密度奇偶校验码　切入点：准双对角线　出处：《国防科学技术大学》2015年硕士论文

【摘要】：低密度奇偶校验码采用稀疏奇偶校验矩阵进行构造,是一类特殊的线性分组码,它具有逼近Shannon限的优异纠错编码性能。该码适用于具有传输距离远、信号衰减大、信号发送功率受限等特点的空间通信信道。然而,LDPC码编码过程通常具有较高的复杂性,在一定程度上制约了它在实际中的普遍应用。对此,本文在空间通信应用背景下,对如何设计性能优、复杂度低的译码器进行了研究,主要工作如下:首先,在对LDPC码的构造方法和编码算法研究的基础上,提出了一种采用准双对角线结构构造奇偶校验矩阵的方法。此种方法继承了IEEE802.16e标准LDPC码的准双对角线结构,并且在构造时消除了对基础矩阵扩展因子的约束条件。通过改进,不但构造出的LDPC码编码简单,而且可以根据设计的不同需求,灵活构造出多种不同码长和码率组合的LDPC码,适用范围更广,更加贴近空间通信的工程应用需求。本文选用(16384,8192)LDPC码,采用这种方法对H矩阵进行构造,给出了快速编码算法和编码电路设计。然后,针对空间数据咨询委员会(CCSDS)推荐的几种深空和近地应用的LDPC码,采用传统译码算法,分别设定不同译码算法、迭代次数、码长和码率分析其对误码率的影响,寻找出有利于LDPC码译码器硬件实现的归一化最小和译码算法。通过对传统并行算法消息传递方式的改进,将串行消息传递机制应用于归一化最小和译码算法中。串行算法相对于传统并行算法,减少了消息更新的迭代次数和运算量,能够改善译码的收敛性、降低数据存储量,更适合低复杂度的译码器设计。对本设计中构造的(16384,8192)LDPC码采用串行归一化最小和算法进行译码仿真,能够取得优异的误码率性能。最后,完成(16384,8192)LDPC码译码器的FPGA硬件实现。通过对最大迭代次数、归一化因子、数据量化方式等影响译码性能的重要因素进行仿真分析,基于FPGA对译码器顶层电路和功能模块进行设计和优化,完成了(16384,8192)LDPC码译码器的硬件实现。通过测试验证,本文的设计方法可以实现高译码性能、低实现复杂度的LDPC码译码器,能够较好的应用于空间通信的相关领域。
[Abstract]:Low-density parity-check codes are constructed by sparse parity check matrices. They are a special class of linear block codes with excellent error-correcting performance which approximates the Shannon limit.The code is suitable for space communication channels with long transmission distance, large signal attenuation and limited signal transmission power.However, the coding process of LDPC codes usually has high complexity, which restricts its general application in practice to some extent.In this paper, under the background of space communication application, we study how to design a decoder with good performance and low complexity. The main work is as follows: firstly, based on the research of the construction method and coding algorithm of LDPC code,A method of constructing parity check matrix with quasi-bidiagonal structure is presented.This method inherits the quasi-bidiagonal structure of the IEEE802.16e standard LDPC code and eliminates the constraint on the expansion factor of the basic matrix.Through the improvement, not only the LDPC codes constructed are simple, but also the LDPC codes with different code length and rate combination can be flexibly constructed according to the different requirements of the design, and the application range is wider and closer to the engineering application requirements of space communication.In this paper, the H matrix is constructed by using this method, and the fast coding algorithm and the design of coding circuit are given.Then, for several kinds of deep space and near-ground LDPC codes recommended by the Space data Advisory Committee, the traditional decoding algorithms are used to set different decoding algorithms, the number of iterations, the code length and the bit rate to analyze their influence on the bit error rate (BER).A normalized minimum and decoding algorithm is found for the hardware implementation of LDPC decoder.Through the improvement of the traditional parallel algorithm, the serial message passing mechanism is applied to the normalized minimum and decoding algorithm.Compared with the traditional parallel algorithm, the serial algorithm reduces the number of iterations and operations of message update, improves the convergence of decoding, reduces the data storage, and is more suitable for the design of low complexity decoder.In this paper, the serial normalized minimum sum algorithm is used to decode and simulate the LDPC code constructed in this design, which can achieve excellent bit error rate (BER) performance.Finally, the FPGA hardware implementation of the LDPC decoder is completed.Through the simulation and analysis of the most important factors such as the number of iterations, normalization factor, data quantization and other important factors affecting the decoding performance, the design and optimization of the top-level circuit and function module of the decoder based on FPGA are carried out.The hardware implementation of the LDPC decoder is completed.The test results show that the design method of this paper can achieve high decoding performance and low implementation complexity of LDPC decoder, and can be used in space communication field.
【学位授予单位】：国防科学技术大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN911.22

【参考文献】