基于分块矩阵变换的线性分组码盲识别

为了解决传统矩阵分析方法存在的误码扩散问题,提出了一种基于分块矩阵变换的线性分组码盲识别方法.首先,将截获序列按照估计码长构造出分析矩阵,将分析矩阵分块后分别进行矩阵下三角变换;然后,以各列列重为度量,根据相关列重量的统计分布特性设置相关列阈值,并统计出符合阈值的相关列的个数,当相关列的个数最大时即为真实码长的情况.该方法还可以识别码字同步点,识别方法简单.理论分析及仿真结果表明,该识别方法的容错性能较好,在误码为5%的条件下,对(15,7)线性分组码的正确识别率依然能达到80%.

Blind recognition of linear block codes based on block matrix transformation

To solove the problem of error diffusion in traditional matrix analysis methods, a linear block code blind recognition method is proposed based on block matrix transformation. Firstly,the analysis matrix is constructed according to the estimated code length,and the matrix is transformed into the low triangular matrix after the analysis matrix is divided. Then,the thresholds of the relevant columns are set according to the statistical distribution characteristics of the relevant column weights,and the number of correlation col-umns corresponding to the thresholds is counted when the number of related columns is the largest. The method is simple and can also identify codeword synchronization points. Theoretical analysis and simulation results show that the fault recognition performance of the method is good,and the correct recognition rate of (15,7) linear block codes can reach 80% under the condition of bit error rate of 5%.