一种基于量化预处理的低复杂度LDPC译码算法

为了降低低密度奇偶校验(Low Density Parity Check，LDPC)码译码算法的复杂度，提出了一种基于量化预处理的LDPC迭代大数逻辑译码算法。该算法在迭代译码过程中，校验节点采用基于伴随式的信息处理方式，避免了外信息的计算；同时，变量节点基于回传的伴随式信息进行可靠度偏移大小的计算，并结合与当前码位相对应的调制映射信息进行可靠度偏移方向的设计。迭代更新时，变量节点采用基于信息匹配的可靠度更新规则。迭代前的量化预处理能避免实数乘法运算进入迭代过程，使其只涉及整数加法操作和逻辑操作。仿真结果表明，在保持译码性能的前提上，所提算法具有更低的译码复杂度。

A low-complexity LDPC decoding algorithm based on pre-processing quantization

To reduce the computational complexity,an iterative majority-logic low density parity check(LDPC) decoding algorithm is proposed based on pre-processing quantization.In the proposed algorithm,syndrome-based message processing is employed at check nodes,which can avoid the extrinsic information computation.At variable nodes,the shifting-step of the reliability-message is computed according to the received syndrome message.Meanwhile,the shifting-direction is designed by combining the modulation information with respect to the current bit.Furthermore,the message-matching strategy is adopted in the iterative decoding.With the pre-processing quantization,the proposed algorithm can avoid the complicated real multiplication and only involves the integer addition and logic operations,resulting in low complexity.Simulation results show that the proposed algorithm has low decoding complexity while maintaining excellent decoding performance.