A new decoding scheme aided by simulated annealing algorithm is proposed to further improve the decoding performance of successive cancellation (SC) for polar codes at the short block. We use simulated annealing to revise the decoding result of SC which cannot pass the CRC check. To generate the new neighbors, the decoder flips one bit from the set of the least unreliable information bits each time in the estimated source vector of SC decoding. Euclidean distance is used to measure the gap between the new neighbor solution and the received word so that the decoder can obtain a global optimal solution. Simulation shows that the proposed decoder has a performance gain about 0.5 dB in terms of frame error rate (FER) under short blocks in the additive white Gaussian noise (AWGN) channel compared to other basic decoders, while keeping a low time cost through a parameter tuning process.

为进一步提高短块极化码连续抵消(SC)的译码性能，提出了一种辅助模拟退火算法的译码方案。我们使用模拟退火对不能通过CRC校验的SC解码结果进行修正。为了生成新的邻居，解码器每次在 SC 解码的估计源向量中从一组最不可靠的信息位中翻转一位。欧几里德距离用于衡量新邻居解与接收词之间的差距，以便解码器获得全局最优解。仿真表明，与其他基本解码器相比，所提出的解码器在加性高斯白噪声 (AWGN) 信道的短块下的帧错误率 (FER) 方面具有约 0.5 dB 的性能增益，