论文部分内容阅读
鉴于大衍数列构造的校验矩阵不存在短环和原模图具有高速译码、低译码平台、低译码门限的优点,将大衍数列与原模图相结合,提出了一种基于大衍数列的原模图QC-LDPC(quasi-cyclic low-density parity-check)码新颖扩展方法。该扩展方法是利用数学规则和大衍数列固定项差对应的值单调递增的特点替换原模图中的元素。该方法构造出的校验矩阵不存在4环,并且用该方法所构造的YD-QC-LDPC码具有优越的纠错性能。仿真结果表明,利用该方法构造的码率为0.5的YD-QC-LDPC(3000,1500)码,在误码率为10-6时,比同码率的基于等差数列构造的APS-QC-LDPC(3132,1566)码提高了1.1d B,并与基于大衍数列构造的DY-QC-LDPC(3000,1500)码相比,其净编码增益(NCG)有显著的提高。
In view of the fact that the parity check matrix constructed by Dayu sequence does not exist the advantages of high speed decoding, low decoding platform and low decoding threshold for the short loop and the original mode, The new method of quasi-cyclic low-density parity-check (QC-LDPC) expansion of the archetype of large-derivative sequences. The extension method is to replace the elements in the original model with the feature that the value corresponding to the fixed item difference of the math sequence and the number of monotone increases monotonically. The proposed method does not exist a 4-check matrix for the parity check matrix and the YD-QC-LDPC code constructed by this method has superior error correction performance. The simulation results show that the YD-QC-LDPC (3000, 1500) code with the code rate of 0.5 constructed by this method shows better performance than the APS-QC -LDPC (3132,1566) code increased by 1.1d B, and its net coding gain (NCG) was significantly improved compared to the DY-QC-LDPC (3000,1500) code based on the Da Yanyan sequence.