【摘要】 本文针对水声通信系统中有色噪声干扰和噪声非一致问题,提出了一种适用于有色噪声背景的多通道跳频软解调算法。该方法在最大比例合并的基础上,考虑多通道间噪声强度与频谱特征的非一致性,将其拓展为对各阵元各频点分别加权的二维合并方法。同时,针对有色噪声背景,提出两种基于最大后验概率的软解调算法。通过估计并利用有色噪声频谱结构,逐比特计算软度量值,在提升软解调准确性的同时,降低强噪声频点的软度量值,从而减小译码错误概率。相比于传统软解调算法,所提算法在单通道有色噪声仿真背景下获得2.7 dB性能增益;在多通道非一致有色噪声背景下获得4.7 dB性能增益。在外场航行器舷侧阵通信试验中,所提算法将平均误码率由1.59×10^-2降至0。更多还原

【Abstract】 To address the interference of colored noise, and its inconsistency in the underwater acoustic communication channel,this paper proposes a soft demodulation algorithm for the multichannel frequency hopping-multiple frequency shift keying（FHMFSK） modulation scheme. Based on the criterion of maximum ratio combining（MRC）, the proposed algorithm takes into account the inconsistency of noise intensity and spectral characteristics among channels, and expands MRC to a two-dimensional combining method that individually weights each frequency point of each array element. Meanwhile, two soft demodulation methods based on maximum a posteriori（MAP） probability are proposed in the channel of colored noise. The spectral structure of colored noise is estimated and exploited to calculate soft metrics for each bit. The accuracy of soft demodulation is improved, and the soft metrics of frequency points affected by intense noise are reduced, thereby reducing the bit error probability during decoding. Compared with the traditional soft demodulation algorithm, the proposed algorithm achieves a 2.7 dB performance gain in the simulation channel of single-channel colored noise and a 4.7 dB gain in the channel of multichannel inconsistent colored noise. Additionally, in the field test equipped with flank arrays, the proposed algorithm reduces the average bit error rate（BER） from 1.59 × 10^-2 to 0.更多还原