基于Polar码的水声通信信源信道联合译码方法

A joint source-channel decoding algorithm based on polar codes for underwater acoustic communication

水声信道时延扩展较长,频域选择性衰落严重,导致水声通信提升可靠性困难。同时,水声通信实际发送的信源中通常存在残留冗余,传统方法难以利用这部分冗余,从而导致一定的带宽浪费。针对该问题,提出了一种基于Polar码的水声通信信源信道联合译码方法。该方法根据Polar码的译码结构,以信源状态转移关系为基础构建信源信道联合译码网格图,综合信源转移概率和信道转移概率计算统一的序列后验概率,实现了信源译码和信道译码的一体化联合优化。该方法能够利用信源符号间相关性导致的残留冗余抵抗信道差错,在不降低通信速率的情况下进一步提高水声通信系统可靠性。湖上实验结果表明,在相同条件下,所提方法可将系统误码率由传统分离译码的1.9×10^(−2)降低为4.0×10^(−4);当传输图像序列时,传统分离译码方法的误码率为3.7×10^(−3),而所提方法则在实验中实现了无误码通信。

It is difficult to improve the reliability of underwater acoustic communication(UAC)due to the long-time delay and severe frequency fading of underwater acoustic channels.On the other hand,there are always some source residual redundancies in practical UAC.Traditional methods cannot use these redundancies.That leads to some waste of bandwidth.To overcome these problems,a joint source-channel decoding algorithm based on polar codes for UAC is proposed.According to the decoding structure of polar codes,the joint source-channel decoding trellis is constructed based on the source transfer relationship.The posterior probabilities are computed by combining the source transfer probability and the channel transfer probability.In this way,the source decoding and channel decoding are optimized jointly.This algorithm can take advantage of the source residual redundancies for error control decoding and can improve the reliability of UAC without reducing the communication rate.The lake experimental results show that the proposed algorithm can reduce the bit error rate(BER)from 1.9×10^(−2)of the traditional separate decoding algorithm to 4.0×10^(−4).When transmitting photographs,the proposed algorithm can reduce the BER from 3.7×10^(−3)of the traditional separate decoding algorithm to 0.