Abstract

In this paper, three time-varying channel equalization schemes are studied in the underwater acoustic (UWA) Orthogonal Frequency Division Multiplexing (OFDM) communication system. The equalization algorithms are the zero-forcing (ZF) equalization algorithm, and the minimum mean square error equalization (MMSE) algorithm and the serial interference cancellation (SIC) equalization algorithm. Among the schemes, there is a problem of needing a large amount of operation when obtaining the inversion of the channel matrix. Then, to reduce the computation complexity of channel matrix inversion, the band approximation of the channel matrix, the serial equalization and the LDL^H decomposition are also studied. To evaluate the efficacy of the algorithms studied in this paper, numerical simulation and the field experiment are both conducted. The simulation results proof that each equalization algorithm can work appropriately under different time-varying conditions, and valid the reliability of each simplified algorithm under the same Doppler factor. The results of two sets of field experiment also prove that the simplified algorithm eliminates the influence of the residual narrow band Doppler to a certain extent, and a better effect is obtained while a channel estimation algorithm with higher accuracy is combined.

中文翻译：

---

水下声OFDM通信系统中的时变信道均衡

摘要

本文在水下声学（UWA）正交频分复用（OFDM）通信系统中研究了三种时变信道均衡方案。均衡算法是强制零（ZF）均衡算法，最小均方误差均衡（MMSE）算法和串行干扰消除（SIC）均衡算法。在这些方案中，存在一个问题，当获得信道矩阵的求逆时，需要大量的运算。然后，为了降低信道矩阵求逆的计算复杂度，信道矩阵的频带近似，串行均衡和LDL ^H还研究了分解。为了评估本文研究算法的有效性，进行了数值模拟和现场实验。仿真结果证明了每种均衡算法在不同的时变条件下都能正常工作，并且在相同的多普勒因子下验证了每种简化算法的可靠性。两组现场实验的结果也证明，该简化算法在一定程度上消除了残留窄带多普勒的影响，结合了高精度的信道估计算法，取得了较好的效果。