低密度校验码是一种基于稀疏校验矩阵的线性码,LDPC的性能接近Shannon限,并具有简单有效的迭代译码算法[1,2]。针对LDPC原始的置信传播迭代译码算法,文中分别从概率似然以及对数似然的角度对该算法进行改进,改进后的这两种译码算法译码更简单,且其运算量大大降低。 Low density parity check codes are linear codes based on sparse parity check matrix. The performance of LDPC is close to Shannon limit, and has a simple and effective iterative decoding algorithm [1,2]. Aiming at the original LDPC iterative decoding algorithm of belief propagation, the algorithm is improved from the perspective of probability likelihood and log-likelihood respectively. The improved decoding algorithms are simpler to decode and their computational complexity is greatly reduced .