面向连续变量量子密钥分发的多边型LDPC码设计与性能分析

为满足连续变量量子密钥分发(continuous-variable quantum key distribution, CV-QKD)应用场景中对高性能低密度奇偶校验(low density parity check, LDPC)码的需求，提出了针对一类具有3种边类型且部分变量节点度为1的多边型LDPC(multi-edge type LDPC, MET-LDPC)码的设计方法。通过掩模到需要的度分布的方式设计左上角矩阵;采用基于多路径外在信息度(extrinsic message degree，EMD)策略的渐进边增长(progressive edge growth, PEG)算法设计左下角矩阵;将各部分矩阵组合在一起完成MET-LDPC码的设计。仿真结果表明，采用掩模加PEG算法设计的MET-LDPC码比单独用PEG算法设计的MET-LDPC码性能更优。

Design and performance analysis of multi-edge type LDPC codes for continuousvariable quantum key distribution

To meet the requirements of low rate, long code length and high performance low density parity check (LDPC) codes in continuous variable quantum key distribution (CV-QKD) application scenarios, this paper proposes a design method for a class of multi-edge type LDPC (MET-LDPC) codes with three edge types and partial variable node degree of 1. Though the check matrix of MET-LDPC codes with this structure can be divided into four parts, only the upper left corner and lower left corner of the check matrix need to be designed based on the optimized degree distribution. The upper left matrix is designed by the method of masking to desired degree distribution, and the lower left matrix is designed by the progressive edge growth (PEG) algorithm based on the multipath extrinsic message degree (EMD) strategy. Then the matrix of each part is combined together to complete the design of MET LDPC codes. Lastly, simulation results show that the masking and PEG algorithm are used to design MET LDPC codes with better performance than the MET LDPC codes designed by PEG algorithm alone, which shows the superiority of the design method presented in this paper.