基于准循环LDPC码译码软信息的码辅助帧同步算法

Code-Aided Frame Synchronization Algorithm Based on Quasi-cyclic LDPC Codes Soft Decoding Information

确保较低信噪比条件下的系统帧同步,是LDPC码在系统应用中的关键问题。基于最大似然的准则,提出了一种适合准循环LDPC编码系统的码辅助盲帧同步算法。该算法通过计算不同帧偏移处的信道输出软信息向量满足LDPC码校验矩阵中所有校验方程的概率与违背所有方程的概率的对数似然比值,再根据最大似然值对应的信息向量确定最终的帧同步边界。新算法可借助译码器的部分资源来实现帧同步搜索,提高了译码器的利用率,降低了实现复杂度;无需一次完整的迭代译码过程,减少了同步捕获的时间。仿真结果表明,相比已有的码辅助盲帧同步算法,新算法具有较好的帧同步性能;借助新同步算法仿真获得的系统误比特率和帧错误率接近已有的码辅助帧同步算法的译码性能。

To ensure system frame synchronization at very low SNR is a key challenge of practical application for LDPC codes.Based on Maximum-Likelihood rule,a code-aided frame synchronization algorithm for Quasi-Cyclic LDPC coded system was proposed.For soft information vectors with different frame offsets,log-likelihood ratios presenting the probability ratios of satisfaction of all check restrictions to violation instances are computed firstly,and then the boundary corresponding to the maximum value is chosen as the final frame synchronization boundary.Using part of LDPC decoder hardware resource,new algorithm doesn＇t need additional hardware resource which reduces realization complexity and increases utilization ratio of decoder.It doesn＇t need a full iterative decoding process which saves synchronization capture time.Compared to other code-aided frame synchronization algorithms,favorable synchronization performance and decoding performance using the new algorithm that approaches the performance of existing code-aided frame synchronization are presented by simulations.