两种降低复杂度的符号翻转多元LDPC译码算法

该文提出两种低复杂度的基于符号翻转的多元低密度奇偶校验码(LDPC)译码算法:改进型多元加权译码算法(Iwtd-AlgB)和基于截断型预测机制的符号翻转(TD-SFDP)算法。Iwtd-AlgB算法利用外信息频率和距离系数的简单求和取代了迭代过程中的乘性运算操作；TD-SFDP算法结合外信息频率和翻转函数特性，对译码节点和有限域符号进行截断与划分，使得只有满足条件的节点和符号参与运算与翻转预测。仿真和数值结果显示，该文提出的两种算法在性能损失可控的前提下，可减少每次迭代的运算操作数，实现性能和复杂度之间的折中。

Two Low-complexity Symbol Flipping Decoding Algorithms for Non-binary LDPC Codes

Two low-complexity symbol flipping decoding algorithms, the Improved weighted-Algorithm B algorithm (Iwtd-AlgB) and the Truncation-based Distance-Symbol-Flipping-Decoding with Prediction (TD-SFDP) algorithm, are presented for non-binary Low Density Parity Check (LDPC) codes. For the Iwtd-AlgB algorithm, the scaling factor of the flipping metric can be replaced by the simple sums of the extrinsic information and the distance-based parameter, which can avoid the multiplication operations in the iterations and thus can reduce the decoding complexity. For the presented TD-SFDP algorithm, the variable nodes and the finite field symbols are truncated and classified based on the extrinsic information frequency and the flipping function. Only those nodes/symbols that satisfy the designed conditions can be involved in the message updating process. Simulations and numeric results show that, the presented two decoding algorithms can reduce the computational complexity at each iteration with a controllable performance degradation, thus can make efficient trade-offs between performance and complexity.