删除信道下LT码译码符号分布模型  

Decoding symbol distribution model of LT code in deleted channel

作　　者：张永祥 牛芳琳[1] 张思 王力正 ZHANG Yongxiang;NIU Fanglin;ZHANG Si;WANG Lizheng(School of Electronics and Information Engineering,Liaoning University of Technology,Jinzhou,121000,China)

机构地区：[1]辽宁工业大学电子与信息工程学院,辽宁錦州121001

出　　处：《长江信息通信》2022年第7期4-8,共5页Changjiang Information & Communications

摘　　要：在删除信道中,采用LT码作为信道编码时,由于受到噪声随机性和度1符号分布随机性的影响,合法接收者和窃听者在译码成功时会接收到不同数量的编码符号,产生不同的概率密度分布。针对译码成功时不同概率密度分布的问题,提出了一种基于GMM来对不同删除概率下得到的概率密度函数进行统一表示。首先通过Xie-Beni指数确定观测数据的最佳聚类数K,然后利用改进后的FCM算法求观测数据的初始值,并将其作为EM迭代算法的初始值,将最终迭代结果作为GMM的参数,建立删除信道下接收端译码成功时译码符号数量的概率密度分布。In the deleted channel, when using LT code as channel coding, due to the influence of noise randomness and degree 1symbol distribution randomness, legitimate receivers and eavesdroppers will receive different numbers of coded symbols and produce different probability density distributions when decoding is successful. Aiming at the problem of different probability density distributions when decoding is successful, a unified representation of probability density functions under different deletion probabilities based on GMM is proposed. Firstly, the optimal clustering number k of the observation data is determined by Xie Beni index,and then the initial value of the observation data is obtained by using the improved FCM algorithm, which is used as the initial value of the EM iterative algorithm, and the final iterative result is used as the parameter of GMM to establish the probability density distribution of the number of decoding symbols when the receiver decodes successfully in the deleted channel.

关 键 词：LT码 EM算法 Xie-Beni指数 高斯混合模型 FCM算法 

分 类 号：TN911.22[电子电信—通信与信息系统]