线性分组码的盲识别技术研究

发布时间：2018-02-23 00:09

本文关键词： 信道编码线性分组码循环码 BCH码盲识别　出处：《河北大学》2015年硕士论文　论文类型：学位论文

【摘要】：为保证实际数字通信系统中信息传输的可靠性和安全性,通常使用信道编码技术,该技术在许多领域得到飞速发展,随之信道编码识别技术也开始备受关注。信道编码识别的目的是仅根据截获数据序列,在仅有少量甚至无任何先验信息的条件下识别出截获数据流的编码体制和参数,从而提取原始信息序列。而线性分组码编码简单、纠错能力较强,应用于数字通信的许多领域中。因此本文主要针对二进制线性分组码参数的盲识别问题展开研究。首先简要介绍了分组码知识及编码识别的数学模型和识别参数,为下面章节参数盲识别算法的讨论奠定理论基础。接着总结现有线性分组码参数识别算法的优缺点,进而提出一种利用码重相似度的算法识别码长、同步点,这是根据实际序列与随机序列的码重分布之间的差异最大特性来识别的,并利用仅在数学理论证明中出现的特征——深度分布,识别生成矩阵,从而完成对分组码的盲识别。码重相似度的算法理论过程简单易懂,仿真结果表明该算法在误码率0.01下能有效识别。最后针对分组码的特殊码字——循环码进行研究:对于BCH码的半盲识别问题,根据其循环性,利用辗转相除法求得移位前后码字的最高公因式,其阶数相对于随机序列来说分布极不平衡,根据此差异性,提出一种运用方差差值、平均欧式距离的量化指标分别识别码长的算法,比较这两种量化指标识别效果,进而提出一种新的融合指标识别码长的算法,并构建公因式系数概率分布,识别生成多项式,实现了BCH码的半盲识别。经实验仿真,融合指标识别码长的算法容错性较好。为有效解决高误码率下高码率循环码的全盲识别问题,在BCH码识别算法的启发下,仍利用最高公因式阶数分布之间的差异性特征,通过理论改进和实验仿真,提出基于标准差率差值的最高公因式阶数的循环码全盲识别算法,该算法计算量较少,容错性强,且在0.023的误码率下,对中短码识别效果明显。
[Abstract]:In order to ensure the reliability and security of information transmission in practical digital communication systems, channel coding technology is usually used, which has been developed rapidly in many fields. The purpose of channel coding recognition is to identify the encoding system and parameters of intercepted data streams only according to intercepted data sequences and without any prior information. In order to extract the original information sequence, the linear block code is easy to encode and has strong error correction ability. In this paper, the blind recognition of binary linear block code parameters is studied. Firstly, the mathematical model and identification parameters of block code knowledge and coding recognition are briefly introduced. It lays a theoretical foundation for the discussion of parameter blind identification algorithm in the following chapters. Then it summarizes the advantages and disadvantages of the existing linear block code parameter identification algorithms, and then proposes an algorithm based on the similarity of code weight to identify the length of code and the synchronization point. This is based on the maximum characteristic of the difference between the code weight distribution of the real sequence and the random sequence, and the generation matrix is identified by using the feature-depth distribution, which appears only in the mathematical theory proof. Thus the blind recognition of block codes is completed. The algorithm of code weight similarity is simple and easy to understand. Simulation results show that the algorithm can be recognized effectively under BER 0.01. Finally, the special codeword cyclic code of block code is studied. For semi-blind identification of BCH code, according to its circularity, The highest common factor of the code word before and after the shift is obtained by the method of tossing and rotating phase division, the order of which is extremely unbalanced relative to the random sequence. According to the difference, a difference of variance is proposed. The average Euclidean distance quantization index respectively identifies the code length algorithm, compares these two quantization index recognition effect, then proposes a new fusion index identification code length algorithm, and constructs the common factor coefficient probability distribution, recognizes the generation polynomial, The semi-blind recognition of BCH code is realized. The experiment results show that the algorithm of fusion index identification code length is fault-tolerant. In order to effectively solve the problem of full blind recognition of cyclic code with high bit error rate, the algorithm is inspired by BCH code recognition algorithm. Based on the difference between the order distribution of the highest common factor, and through theoretical improvement and experimental simulation, a full blind identification algorithm for cyclic codes based on the highest common factor order of standard deviation rate difference is proposed. The algorithm has less computation and strong fault tolerance. At the error rate of 0.023, the recognition effect of medium and short code is obvious.
【学位授予单位】：河北大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN911.22

【参考文献】