基于存储机制的LT码编译码方法  

Memory-based encoding and decoding of LT codes

作　　者：姚渭箐[1] 易本顺[1,2] 

机构地区：[1]武汉大学电子信息学院,湖北武汉430072 [2]武汉大学深圳研究院,广东深圳518057

出　　处：《系统工程与电子技术》2018年第1期165-170,共6页Systems Engineering and Electronics

基　　金：国家自然科学基金(61371125);深圳市基础研究项目(JCYJ20150630153917254)资助课题

摘　　要：提出一种基于存储(memory-based,MB)机制的Luby变换码的编译码方法,来实现信息在二进制删除信道(binary erasure channel,BEC)中的可靠传输。首先,发送端的编码器采用泊松鲁棒孤子分布(Poisson-robust soliton distribution,PRSD)产生普通编码包,同时产生携带存储信息的"存储包"。然后,源源不断在BEC中发送编码包和"存储包"给接收端。接收端的译码器根据接收到的"存储包"的数量不同,采取不同方式对输入包进行译码。如果"存储包"全部被接收,则所有输入包都能通过"存储包"中的存储信息直接获得;如果部分"存储包"丢失,则结合"存储包"和置信传播(belief propagation,BP)算法进行译码;如果所有"存储包"丢失,则仅采用BP算法进行译码。仿真结果表明,相比LT码的传统编译码方法,采用PRSD-MB方法可以大大降低误比特率,提高编译码效率。The memory-based(MB)encoding and decoding scheme of Luby transform(LT)codes is proposed to improve the reliability and efficiency of information transmission over the binary erasure channel(BEC).The encoder at the sender generates the common encoded packets by using the Poisson-robust soliton distribution(PRSD),and generates several memory packets carrying the memory information.Then these encoded packets and memory packets are sent over BEC.According to the number of memory packets received,the decoder recovers the input symbols by using different methods.If all memory packets are received,the whole input symbols can be obtained from the memory information in the memory packets directly.If some memory packets are lost,the memory packets and the belief propagation(BP)algorithm are combined to recover the input symbols.If all memory packets are lost,only the BP algorithm is employed to decode the input symbols.Simulation results show that compared with traditional methods,the proposed scheme achieves a great reduction in the bit error rate,and improves the encoding and decoding efficiency.

关 键 词：Luby变换码 二进制删除信道 泊松鲁棒孤子分布 存储包 

分 类 号：TN911.22[电子电信—通信与信息系统]