摘要： 在二进制输入加性高斯白噪声信道中传输LT码时，采用高斯近似方法预测置信传播译码算法的误比特率性能不够准确。为此，提出一种改进的高斯近似方法，其中，输入节点度分布采用泊松分布，相应的软信息为高斯混合物，在此基础上给出一种LT码度分布优化方法。仿真结果证明，该方法相比同类方法性能更优越。

关键词: 喷泉码, 二进制输入加性高斯白噪声信道, 置信传播译码, 渐进分析, 高斯近似

Abstract: When LT codes are transmitted in the Binary Input Additive White Gaussian Noise(BIAWGN) channel, traditional Gaussian approximation method, which is used to predict the bit error rate performance of Belief Propagation(BP) decoding algorithm, is not accurate. So this paper proposes an improved Gaussian approximation method to predict bit error rate performance. The degree distribution of input node is Poisson distribution and the corresponding soft information is a Gaussian mixture. Based on this method, a new method used in the design of good degree distribution for LT codes is given. Simulation results show that the improved method is more accurate than the previous one.

Key words: fountain code, Binary Input Additive White Gaussian Noise(BIAWGN) channel, Belief Propagation(BP) decoding, asymptotic analysis, Gaussian approximation

中图分类号: 

• TN911.22