针对IRA-LDPC码类的半随机半代数结构设计

Semi-random and semi-algebraic structural design for IRA-LDPC codes

提出用半随机半代数结构的设计方法来构造IRA-LDPC码的信息位所对应的奇偶校验矩阵H d。与现有结构化LDPC码相比,所给出的H d矩阵的结构化紧凑表示阵列的独特优势在于:可使H d矩阵中每个1元素的位置坐标均能用数学表达式计算得到,不仅极大地降低了随机奇偶校验矩阵对存储资源的消耗,而且还为LDPC编解码器的低复杂度硬件实现提供了可能性。与现有工业标准中的LDPC码相比,所提出的IRA-LDPC码在误码率与信噪比的仿真性能方面也占有优势。

A method of semi-random and semi-algebraic structure was presented for constructing the low-density pari-ty-check matrix that corresponds to the information bits of the IRA codes. Compared with the existing structural LDPC codes, the distinct advantage of the presented compact structural array for information-bit-corresponding matrix is that the position coordinate of each 1 element in this matrix can be calculated by a determinate algebraic expression, which not only reduces the consumption of memory resource for the random parity-check matrix, but also provides the potential probability for designing low complexity hardware circuit of the LDPC encoder/decoder. In addition, compared with the existing practical LDPC codes in industrial standard, the presented IRA-LDPC code is also slight preponderance in the performance of simulation in bit error rate and signal noise ratio （BER-SNB）.