Enhanced Channel Estimation and Symbol Detection for High Speed MIMO Underwater Acoustic Communications

This paper focuses on the channel estimation and symbol detection problems in multi-input multi-output (MIMO) underwater acoustic (UWA) communications. To enhance channel estimation performance, a cyclic approach for designing training sequences and an iterative adaptive approach (IAA) for estimating the channel taps are presented. Sparse channel estimates can be obtained. by combining IAA with the Bayesian information criterion (BIC). Moreover, the parametric RELAX algorithm can be used to improve the IAA with BIG estimates further. Regarding symbol detection, a minimum mean-squared error (MMSE) based detection scheme, called RELAX-BLAST, is presented. Both simulated and experimental results are provided to validate the proposed MIMO scheme.