多进制LDPC码与大规模MIMO级联相关技术研究

发布时间：2018-04-16 15:32

本文选题：多进制LDPC码 + 多进制变换算法　；参考：《浙江大学》2014年博士论文

【摘要】：现代无线通信系统中,用户对信息传输速率和传输质量的需求日益提高,紧缺的频谱资源与复杂的传输环境逐渐成为发展的主要限制。在这种情况下,大规模MIMO和高性能的信道编码技术开始引起人们的广泛关注。大规模MIMO技术相比传统的MIMO技术,通过增加天线数目,能够在不增加带宽和发送功率的前提下进一步提升系统容量；而多进制低密度校验码(LDPC)由于其优越的抗随机错和突发错能力,可以有效提高系统可靠性。因此,大规模MIMO技术和多进制LDPC码的联合迭代系统,可以兼顾系统的传输速率和可靠性,具有重要研究价值。本文以多进制LDPC码和大规模MIMO检测技术为研究对象,研究了低码率多进制LDPC码基于多进制变换的构造算法,以及多进制LDPC码与大规模MIMO检测联合迭代系统的优化设计方法。首先,论文介绍了MIMO检测技术与多进制LDPC码的基本原理。针对MIMO检测技术,介绍了其系统模型以及线性、非线性两类检测算法,作为大规模MIMO检测算法的理论基础；针对多进制LDPC码,介绍了其基本概念和描述方式,给出了具有快速编码结构的准循环码的编码与构造算法,并介绍了几种常见的译码算法；此外,论文还介绍了最小均方误差(MMSE)检测与多进制LDPC码的联合迭代系统,作为与后续章节中算法复杂度与性能比较的标准。其次,论文研究了多进制变换算法及其应用。重点研究了多进制变换算法在低码率多进制LDPC码构造中的应用,提出了伽罗华域和所有符号的全排列组合中多进制变换系数的选取准则,通过选取最优系数来构造具有最优性能码字,并与传统算法进行比较验证了该算法的有效性。论文还将多进制变换算法应用于混合自动重传请求(HARQ)系统和MIMO发送分集系统中,在不增加复杂度的前提下,获得了优于传统算法的性能。接着,论文研究了大规模MIMO系统中似然上升搜索(LAS)检测与多进制LDPC码的联合迭代系统。在介绍LAS检测算法、符号域M-LAS检测算法的基础上,论文针对大规模MIMO系统的上行链路,提出了高阶调制时M-LAS检测与多进制LDPC码的联合迭代算法。在联合系统中,将译码器输出的硬判决结果反馈回检测器作为初始向量,并将译码校验结果和外信息反馈回检测器,用于决定其搜索集合,可以在缩小搜索范围的同时提升系统性能。仿真中,通过与相应的MMSE系统从性能和复杂度的角度进行比较,验证了LAS检测在大规模MIMO系统中的适用性。最后,论文研究了置信度传播(BP)检测算法与多进制LDPC码的联合迭代系统。在二进制BP检测算法及其高斯近似简化的基础上,论文提出了基于符号域的高斯近似BP检测算法,并将其分别应用于多用户和单用户MIMO系统。在多用户MIMO系统中,研究了符号域高斯近似BP检测算法与多进制LDPC码的联合迭代系统,提出了根据译码校验结果改进反馈外信息的方法。在单用户MIMO系统中,提出了符号域高斯近似BP检测算法与多进制LDPC码的因子图联合迭代算法,将检测器和译码器因子图中的变量节点进行了合并,通过检测节点和校验节点的同时更新加快了系统收敛速度,取得了优于独立检测译码系统与传统联合迭代系统的性能。
[Abstract]:In modern wireless communication system, the user demand for information transmission rate and quality of the increasing shortage of spectrum resources and complex transmission environment has gradually become the main restriction of development. In this case, large scale and high performance MIMO channel encoding technology began to attract attention. Compared with the traditional MIMO technology of large-scale MIMO technology, by increasing the number of antennas, can further enhance the system capacity without increasing the bandwidth and transmit power; and multiple low density parity check (LDPC) due to its superior resistance to random error and burst error ability, can effectively improve the reliability of the system. Therefore, the joint iterative system of large-scale MIMO technology and nonbinary LDPC codes that can take into account the transmission rate and reliability of the system, has the important research value.
Based on the binary LDPC codes and large-scale MIMO detection technology as the research object, the research of low bit rate q-ary LDPC codes constructed based on M-ary transform, and binary LDPC codes and large-scale MIMO detection combined with iterative system optimization design method.
Firstly, this paper introduces the basic principle of MIMO detection technology and nonbinary LDPC codes. The MIMO detection technology, introduces the system model and the linear and nonlinear two kinds of detection algorithms, as the theoretical basis of mass MIMO detection algorithm; for nonbinary LDPC codes, and introduces the basic concept and description are given, with fast the structure of quasi cyclic code encoding and encoding algorithm, and introduces several common decoding algorithm; in addition, the paper also introduces the minimum mean square error (MMSE) detection system combined with iterative and non binary LDPC codes, as the following chapters and algorithm complexity and performance standards.
Secondly, the binary transform algorithm and its application. The application of multi band structure in low bit rate transform algorithm for nonbinary LDPC codes in the proposed selection criterion of Galois field and all symbols of all permutations and combinations of M-ary transform coefficients, by selecting the optimal coefficient to construct with optimal codeword, and compared to verify the effectiveness of the algorithm with the traditional algorithm. The paper will also transform M-ary algorithm is applied to the hybrid automatic repeat request (HARQ) system and MIMO transmit diversity system, without increasing the complexity, have better performance than the traditional algorithm.
Then, the paper studies the likelihood ascent search in large MIMO system (LAS) joint iterative detection system with M-ary LDPC code. The LAS detection algorithm based on M-LAS domain symbol detection algorithm, aiming at the large scale MIMO system uplink, puts forward the iterative algorithm of joint detection of M-LAS and binary LDPC codes of high order modulation. In the combined system, the hard decision decoder output feedback detector as the initial vector, and the decoding results of calibration and external information feedback detector is used to determine the set of search, can improve the system performance in narrowing the search range at the same time. In the simulation, compared with the corresponding MMSE system from the perspective of performance and complexity, verify the applicability of LAS detection in large-scale MIMO systems.
Finally, the paper studies the belief propagation (BP) algorithm and joint iterative detection system for nonbinary LDPC codes. The approximate binary BP detection algorithm and its simplified Gauss on the basis of the proposed approximate detection algorithm in BP domain based on the symbol of Gauss, and was applied in the multi user and the single user MIMO system. Multi user MIMO system, the system of joint iterative symbol domain Gauss approximation BP detection algorithm and binary LDPC codes, decoding is proposed according to the calibration results of the information feedback method. In single user MIMO systems, put forward the factor graph iterative algorithm of joint symbol domain Gauss approximation of the BP detection algorithm and binary LDPC codes the variable node detector and decoder in the factor graph are merged by simultaneously detecting nodes and check nodes update to speed up the convergence rate of the system, obtain a better independent detection decoding system The performance of the system and the traditional joint iterative system.

【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN911.22;TN919.3

【共引文献】