一种改进的线性分组码的全盲识别算法

针对线性分组码编码参数的盲识别问题,根据实际与随机序列码重概率分布间较大的差异性,提出了利用两种特征参数(码重标准差率差值、码重信息熵)分别同时识别码长和起始点的算法.根据这两种算法的不足又进一步改进,提出一种对这两种特征参数进行融合来同时识别码长和起始点的算法.在此基础上,通过建立矩阵进行化简获得生成矩阵,从而实现线性分组码的全盲识别.理论分析及实验仿真表明该方法简单易行,容错性较强,在误码率为0.025条件下对中短码识别率达到90％,误码率为0.005条件下对中长码识别率高于80％.

A Improved Blind Recognition Method of Linear Block Code Parameters

In view of the problem of the blind recognition of linear block code parameters, the code length and synchronization point are simultaneous identified by the recognition method based on the two characteristic parameter(Standard Error Rate Difference of code weight and Information Entropy of code weight)that is proposed by the difference of code weight distribution probability between the actual sequence and random sequence .According to the deficiency of these two algorithms, then a new method of fusion of two characteristic parameters is put forward that code length and synchronization point can be simultaneous identified.And through establishing matrix and simplifying Matrix ,and identify the generator matrix to achieve a blind identification linear block codes.Theoretical analysis and simulation experience show that the recognition method is simple and has better error-tolerance, and the recognition method 90% with middle and short code length can be recognized when BER is 0.025 and has better performance more than 80%.about Slightly long code in 0.005BER.