Robust adaptive channel estimation in MIMO underwater acoustic communications

For underwater acoustic (UWA) communications, a robust iterative channel estimation based minimum mean-square-error (MMSE) adaptive turbo equalizer is proposed and studied for multiple-input multiple-output (MIMO) detection. Rather than the classical MMSE or normalized least mean squares (NLMS) estimation algorithms, the improved proportionate NLMS (IPNLMS) is adopted for the iterative MIMO channel estimator. The MIMO channel estimation is performed jointly with the MMSE equalizer and the maximum a posteriori probability (MAP) decoder. With inter-block interference removed, the MIMO MMSE equalization is performed with overlapped information subblocks without guard intervals, thus a high transmission efficiency is guaranteed and performance degradation is prevented. The proposed MIMO detection scheme has been tested by experimental data and proved to be robust against tough MIMO channels. The experimental results for the undersea 2008 Surface Processes and Acoustic Communications Experiment (SPACE08) are reported.