一种新型的卷积码混合译码算法

A Novel Hybrid Decoding Algorithm for Convolutional Codes

该文提出了一种级联的卷积码混合译码算法。该算法由两级译码实现,第1级采用置信传播(Belief-Propagation,BP)算法,而第2级采用修改的维特比译码(Modified Viterbi Decoding,MVD)算法。BP首先对接收序列进行预译码,并利用伴随式将译码输出的对数似然比值分为可靠的和不可靠的两类。不可靠的对数似然比值用接收符号取代,可靠的部分硬判决为编码符号,它们共同组成混合序列。随后,MVD对该混合序列作进一步纠错译码。仿真表明,与传统的维特比算法相比,所提出的混合译码算法的误码性能只有很小的损失,其译码平均复杂度在中高信噪比条件下有明显降低。

A concatenated hybrid decoding algorithm for convolutional codes is presented. The algorithm is complemented by using two-stage decoding, where the first stage uses the Belief-Propagation （BP） algorithm, while the second stage uses the Modified Viterbi Decoding （MVD） algorithm. Firstly, the received sequence is pre-decoded by BP, and its outputs are divided into two groups which are reliable Log-Likelihood Ratios （LLRs） and unreliable LLRs. The hard decision symbols corresponding to reliable LLRs and the parts of received symbols corresponding to unreliable LLRs are formed a hybrid sequence, which is further corrected by MVD. Simulation shows that compared with the conventional Viterbi decoding algorithm, the proposed algorithm has a little performance deterioration with much lower average complexity at moderate-to-high signal to noise ratio.