Abstract The low identification rate of the Orthogonal frequency division multiplexing (OFDM) based subcarrier modulation in underwater acoustic multipath channel is an important issue. Therefore, we proposed a novel Expectation Maximization-Block-Quasi Hybrid Likelihood Ratio Test (EM-Block-QHLRT) method which effectively improved the identification rate while using the blind channel impulse response (CIR) estimation and likelihood. Initially, CIR is obtained by using clustering, and then the CIR is updated iteratively by EM-Block method. Further, the subcarrier modulation is identified using QHLRT. The influence of iteration times, the length of symbol and influence of the number of blocks in EM are analyzed by using extensive simulation. The identification rate of subcarrier modulation of binary phase shift keying (BPSK), QPSK, 8PSK and 16QAM are presented using different signal noise ratio (SNR). In addition, the identification rate of subcarrier modulation based on Average Likelihood Rate Test (ALRT) as well as QHLRT with EM are compared. Simulation results showed that the identification rate of the proposed EM-Block-QHLRT method can reach to more than 90% when SNR is higher than 5 dB. Finally, the performance of the proposed EM-Block-QHLRT method is verified using sea trial data.

中文翻译：

---

基于块期望最大化和似然的水声OFDM副载波调制识别

摘要 基于正交频分复用(OFDM)的副载波调制在水声多径信道中识别率低是一个重要问题。因此，我们提出了一种新的期望最大化-块-准混合似然比测试（EM-Block-QHLRT）方法，该方法在使用盲信道脉冲响应（CIR）估计和似然性的同时有效地提高了识别率。最初通过聚类获得CIR，然后通过EM-Block方法迭代更新CIR。此外，副载波调制是使用 QHLRT 识别的。通过广泛的仿真分析了EM中迭代次数、符号长度和块数的影响。二进制相移键控（BPSK）、QPSK、8PSK 和 16QAM 使用不同的信噪比 (SNR) 呈现。此外，比较了基于平均似然率测试（ALRT）以及QHLRT与EM的副载波调制识别率。仿真结果表明，当SNR高于5 dB时，所提出的EM-Block-QHLRT方法的识别率可达90%以上。最后，使用试航数据验证了所提出的 EM-Block-QHLRT 方法的性能。