论文部分内容阅读
目前,LDPC码的译码主要采用MacKay和Neal提出的BP(Belief-Propagation)迭代译码算法。但其水平步骤和垂直步骤的孤立使部分外信息在下一次迭代中才能对译码产生帮助。为了使这部分外信息得到更有效的利用,本文提出两种以外信息的传递路径为导向的改进BP算法,用于LDPC码的译码。模拟显示,在加性高斯白噪声信道下,本文提出的两种方法在译码复杂度的不变或基本不变的情况下,有效加速了迭代的收敛,得到了更好的译码性能。
At present, the LDPC decoding mainly uses the BP (Belief-Propagation) iterative decoding algorithm proposed by MacKay and Neal. However, the isolation of horizontal steps and vertical steps can make part of the external information available for decoding in the next iteration. In order to make more effective use of this part of the external information, this paper proposes two improved BP algorithms oriented to the transmission of other information, which are used to decode LDPC codes. The simulation shows that under the additive white Gaussian noise channel, the two proposed methods effectively accelerate the iteration convergence under the condition that the decoding complexity is invariant or almost invariant, and obtain better decoding performance.