论文部分内容阅读
该文利用高斯近似法,提出一种基于最小均方误差(MMSE)准则的Tanh法则优化近似新方法。提出反对称分布与同构广义对称分布新概念,推导出同构广义对称分布条件下若干重要结论,并给出Tanh法则最优近似式的计算实现方法。加性高斯白噪声(AWGN)信道下,对一系列(3,6)规则低密度校验(LDPC)码的实验仿真显示,与传统Hagenauer近似法相比,该最优近似方法在不明显增加译码复杂度前提下,对LDPC码译码性能够带来一定改善。
In this paper, Gaussian approximation method is used to propose a new Tanh’s Law approximation method based on the minimum mean square error (MMSE) criterion. A new concept of antisymmetry distribution and isomorphism general symmetry distribution is proposed, and some important conclusions under the condition of isomorphism generalized symmetry distribution are deduced. The calculation and realization method of Tanh’s law optimal approximation formula is given. The experimental simulation of a series of (3, 6) LDPC codes under additive white Gaussian noise (AWGN) channel shows that compared with the traditional Hagenauer approximation, the optimal approximation method does not increase significantly Under the premise of code complexity, decoding performance of LDPC code can bring some improvement.