摘要： LT码是喷泉码的一种，采用LT码的BP译码算法时，当处理完度数为1的数据后，译码将无法继续，造成停止集为空集，影响译码效率。为此，提出一种新的二次译码算法，在译码停止后，针对剩下的编码分组仍然具有可译性的特点，进行进一步译码。仿真结果表明，该算法在译码开销为0.2~0.5的情况下，译码效率普遍提高20%左右。

关键词: 喷泉码, LT码, Raptor码, BP译码算法, 度分布函数, 二次译码算法

Abstract: LT code is a kind of fountain code. When data with degree 1 is proceeded while using BP decoding algorithm, the decoding has to be stopped, which causes stop-set empty, consequently affects the decoding efficiency. This paper proposes a new decoding algorithm for this problem which called the secondary decoding. It carries on a further decoding for the remaining encoding packets with translatability after decoding is stopped. Simulation results show that the secondary decoding increases the efficiency of decoding. When the cost of decoding is between 0.2 and 0.5, the decoding efficiency is generally improved by 20%.

Key words: fountain code, LT code, Raptor code, BP decoding algorithm, degree distribution function, secondary decoding algorithm

中图分类号: 

• TP301.6