基于对数似然率的软判决维特比译码算法

Soft-Viterbi Decoding Method Based on Log-likelihood Ratio

提出了一种基于对数似然率的适合于M-QAM解调的软判决度量计算方法.该方法依据最大后验序列估计准则,按照QAM符号中每个bit值将QAM星座图进行分割,计算出在接收到某个符号后,对应符号中每个bit的为0或1的似然程度作为度量值.仿真结果表明,该计算方法比传统方法有更好的误码性能以及较低的复杂度.

In this paper, a method based on LLR （log-likelihood ratio） for soft bit information computation is proposed. It partitions QAM-constellation by the bit value, compares the proba- bility of 0 or 1 of every bit and computes soft bit value for soft-input Viterbi decoder after receiv- ing the whole QAM symbol according to MAPSE（maximum a posteriori sequence estimation）. Compared with the orthodox method, a lower bit-error probability and lower computation com- plexity can be achieved.