非规则LDPC度分布优化设计

Design and Optimization of Degree Distributions of Irregular LDPC

一对好的度分布可以有效降低LDPC的错误平层和编译码复杂度,在AWGN信道下,通过高斯近似分析方法可近似计算给定度分布的LDPC译码门限,利用差分进化算法可优化度分布以获得具有最大门限的度分布,仿真结果表明获得的度分布的译码门限比线性算法优化结果要好0.15dB左右。

Good degree distributions can improve the error-floor and reduce the complication in encoding and decoding of LDPC. Under AWGN channel, Gaussian approximation algorithm can analyze the decoding threshold of LDPC presented by their degree distributions. Using differential evolution, the degree distributions that possess maximal threshold can be gotten. Through simulation results, the optimized degree distributions which possess a 0.15dB better noisy threshold than using linear algorithm.