极化码及其与MBM级联技术研究

发布时间：2018-01-13 23:31

  本文关键词：极化码及其与MBM级联技术研究　出处：《浙江大学》2017年硕士论文　论文类型：学位论文

  更多相关文章： 极化码 SCL译码算法 MBM技术 级联系统 低复杂度ML检测

【摘要】：信道编码是无线通信系统中对抗干扰和衰落不可或缺的技术手段。极化码是最新提出的一种理论上能达到香农极限的信道编码,编码和译码复杂度低,具有广泛的应用前景。MBM(Media-Based Modulation)是一种新型的调制传输技术,它将部分信息承载到信道状态上,从而提高数据传输效率和物理层安全性。将极化码和MBM级联,能够以较低的复杂度提升系统性能。本文主要研究了极化码的译码算法、极化码与MBM级联方案以及级联系统的检测译码技术。本文首先详细介绍了极化码和MBM的原理特性。本文详细阐述了极化码的极化现象和原理,引进密度进化和高斯近似方法来度量信道可靠性。然后对极化码的编译码算法进行 了说明,并对 SC(SuccessiveCancellation)和 BP(BeliefPropagation)译码算法做了仿真性能对比。针对MBM技术,本文将MBM应用到MIMO(Multiple-Input Multiple-Output)系统中,并对相同数据速率下不同多天线技术方案进行了性能比较。然后本文对极化码译码算法进行了研究,提出了改进的双阈值剪枝策略和动态路径拓展的SCL(SC List)算法。在原来双阈值剪枝策略的基础上,提出了新的阈值划分方法。新策略将选取路径度量值过程的复杂度从O(LogL)降低为O(L)。利用分支策略和剪枝策略,本文提出了一种动态路径拓展的SCL算法,简称DPE-SCL(Dynamic Path Extention SCL)算法。这种译码算法不固定候选路径数,允许其动态变化,由分治策略和剪枝策略来控制路径数的变化。仿真结果表明,DPE-SCL算法在低信噪比下有更好的性能,但是以路径数增多为代价,对应地,高信噪比下性能略有下降,此时具有较低的译码复杂度。最后研究了极化码与MBM的级联方案。本文利用极化码与MBM各自的特点,提出了一种新的级联方式,称为PC-MBM(Polar Coded MBM)。它将极化码的固定位比特映射到信道序号MBM消息上,而将信息位比特映射到SBM(Source-Based Modultaion)消息上。由于固定位比特已知,在接收端就不必对信道序号进行估计,直接可以用于SBM消息的检测,检测复杂度大幅下降。接着利用信道极化特性,将PC-MBM信道转化为并行独立的子信道,根据子信道的信道容量来进行极化码编码构造。针对级联系统的检测译码,本文基于软判决ML检测算法,提出了两种低复杂度的ML(Maximum Likelihood)检测方法。首先借鉴球形译码的思路,将搜索空间限定在一定范围内。其次,提出了区域限制的ML检测方法,将搜索空间进一步减小,使得计算复杂度更低。
[Abstract]:Channel coding is an indispensable technique for anti-jamming and fading in wireless communication systems. Polarization code is a newly proposed channel coding which can reach Shannon limit theoretically. The complexity of coding and decoding is low. MBMN Media-Based Modulation is a new modulation and transmission technology, which carries some information to the channel state. In order to improve the data transmission efficiency and physical layer security, polarization code and MBM concatenation can improve the system performance with low complexity. This paper mainly studies the decoding algorithm of polarization code. The scheme of polarization code concatenation with MBM and the detection and decoding technology of concatenated system. Firstly, the principle and characteristics of polarization code and MBM are introduced in detail. The polarization phenomenon and principle of polarization code are described in detail in this paper. Density evolution and Gao Si approximation are introduced to measure channel reliability. And BPU Beliefpagation). The simulation performance of the decoding algorithm is compared. Aiming at the MBM technology. In this paper, MBM is applied to the MIMO(Multiple-Input Multiple-Output system. The performance of different multi-antenna schemes at the same data rate is compared. Then the decoding algorithm of polarimetric codes is studied in this paper. An improved two-threshold pruning strategy and a dynamic path extension (SCL(SC list) algorithm are proposed, based on the original two-threshold pruning strategy. A new threshold partition method is proposed, which reduces the complexity of the process of selecting path weights from OGlogL to OFL, and makes use of branching strategy and pruning strategy. In this paper, a dynamic path extension SCL algorithm is proposed. Short for DPE-SCL(Dynamic Path Extention SCL, this decoding algorithm does not fix the number of candidate paths, allowing it to change dynamically. The simulation results show that the algorithm has better performance under low SNR, but at the cost of increasing the number of paths. At the same time, the decoding complexity is low. Finally, the concatenation scheme of polarization code and MBM is studied. In this paper, the characteristics of polarization code and MBM are utilized. In this paper, a new concatenation method, called PC-MBM(Polar Coded MMP, is proposed, which maps the fixed bit bits of the polarization code to the MBM message with the serial number of the channel. The information bits are mapped to the SBM(Source-Based module on the message. Because the fixed bit bits are known, it is not necessary to estimate the serial number of the channel at the receiving end. It can be directly used in the detection of SBM messages, and the detection complexity is greatly reduced. Then, the PC-MBM channel is transformed into a parallel independent sub-channel by using the polarization characteristics of the channel. Polarization code coding is constructed according to the channel capacity of subchannels. For the detection and decoding of cascaded systems, this paper is based on soft decision ML detection algorithm. Two low complexity ML(Maximum likelihood-based detection methods are proposed. Firstly, the search space is limited to a certain range by using the idea of spherical decoding. An area limited ML detection method is proposed to further reduce the search space and reduce the computational complexity.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN911.22
，

本文编号：1421066