改进的LDPC码译码算法

Improved algorithm of LDPC decoding

通过将串行置信度传播机制与归一化BP_Based译码算法相结合,构造出一种改进的LDPC码译码算法。该算法按照校验节点的一定顺序进行置信度传播,改善了置信度传播的收敛特性;同时应用归一化BP_Based算法的置信度更新计算法则,有效降低了译码复杂度,适合硬件实现。在AWGN信道下进行性能仿真。仿真结果表明,本文构造的串行归一化BP_Based算法的译码收敛速度明显快于常用LDPC码译码算法的收敛速度,可以显著提高译码性能。

A novel Low-Density Parity Check （LDPC） decoding algorithm was proposed, which integrates the serial belief propagation mechanism with the normalized BP-Based updating rule. The algorithm propagates the belief according to certain order of the check nodes, thus improves the convergence property of the belief propagation. Meanwhile, the algorithm efficiently lowers the decoding complexity by using the normalized BP-Based updating rule, which makes the algorithm to adapt to the implementation of hardware. Simulation was performed under the AWGN channel and results show that, compared with prior LDPC decoding algorithm, the proposed algorithm is able to accelerate the convergence speed of decoding obviously and enhance the decoding performance significantly