摘要
本文针对(2,1,l)卷积码提出一种维特比矩阵译码算法,通过引入整形、合并和动态选择等辅助模块,实现了所有环节的矩阵处理,构建出具有单一结构的并行译码器。由于只需要更改一部分模块的内部参数便可获得不同卷积码译码器,因此非常有利于分析和设计。仿真实验表明,在运算量更少的情况下,矩阵译码器可以取得接近最优的译码性能。
This paper proposes a Viterbi matrix decoding algorithm to the （2,1,l） convolutional codes. Some auxiliary modules, such as reshaping, merging and variable selection, are brought to realize the matrixing processing in all aspects of decoder, and make a parallel decoder with single structure. All kinds of convolutional decoders can be obtained by changing the internal parameter of some modules, so it is very favorable to build and design. Simulation results demonstrate that the matrix decoder can achieve near-optimum performance under the less computation.
出处
《电路与系统学报》
CSCD
北大核心
2012年第3期115-120,共6页
Journal of Circuits and Systems
基金
重庆市自然科学基金(2010BB2240)
关键词
卷积码
状态转移
维特比译码算法
矩阵化
convolutional codes
state transition
Viterbi decoding algorithm
matrixing
