基于量化软信息输入的密钥协商算法  

A Key Reconciliation Algorithm Based on Quantized Soft Information

作　　者：郭福星 辛刚[1] 于大鹏[1] GUO Fu - xing;XIN Gang;YU Da - peng(Institute of Information System Engineering, Information Engineering University, Zhengzhou Henan 450001, China)

机构地区：[1]信息工程大学信息系统工程学院,河南郑州450001

出　　处：《计算机仿真》2018年第6期291-295,共5页Computer Simulation

基　　金：国家863计划资助项目(2015AA01A708)

摘　　要：针对密钥协商中,以量化的硬判决序列作为译码器输入会影响纠错性能的问题,提出了基于交互量化误差(CQA)计算量化软信息的协商算法。该算法利用通信一方的位置索引和另一方的采样值计算量化软信息,并把量化软信息作为低密度奇偶校验(LDPC)码BP译码器的输入来协商密钥。与基于量化硬判决的密钥协商相比,软信息协商算法保留了更多随机变量间的互信息,提升了协商算法的纠错能力。通过仿真表明,基于软信息的协商算法不但在纠错能力上比硬判决的协商算法提高了35%,且当2比特量化相关系数大于0.64小于0.8时生成的密钥长度比硬判决协商算法也至少增加了0.15bits/symbol。In order to solve the problem that the quantized hard decision sequence as the decoder input can affect the error correction performance, we proposed a reconciliation algorithm for computing quantized soft information based on Channel Quantization Alternating in the paper. The location index of the communicating party and the sampied value of the other party were used to compute the quantized soft information which then was used as an input of the low - density parity - check （ LDPC ） code BP decoder to reeoneiliate the key. Compared with the key reconciliation based on quantized hard decision, the soft information reconciliation algorithm can preserve the more mutual information of random variables and improve the error correction capability of the reconciliation algorithm. The simulation shows that the reconciliation algorithm based on quantized soft information can improve the error - correcting ability by 35% than the hard - decision, and the key length is also at least 0. 15 bits / symbol longer than the negotiation algorithm based on hard decision when the number of quantization bits are 2 bits and the correlation coefficient is greater than 0. 64 is less than 0. 8.

关 键 词：密钥协商 量化软信息 交互量化误差 低密度奇偶校验码 置信传播 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]