Enhanced channel estimation and efficient symbol detection in MIMO underwater acoustic communications

Effective training sequences and reliable channel estimation algorithms are essential for enhancing the performance of multi-input multi-output (MIMO) underwater acoustic communications (UAC). Also, effective interference cancellation schemes are crucial for reliable symbol detection. In this paper, the problem of designing MIMO training sequences is considered. Moreover, we present a sparse learning via iterative minimization (SLIM) algorithm for enhanced channel estimation and reduced computational complexity. Furthermore, RELAX-BLAST, a linear minimum mean-squared error based symbol detection scheme, is implemented efficiently by exploiting the conjugate gradient method and diagonalization properties of circulant matrices. The proposed MIMO UAC techniques are evaluated using both simulated and experimental examples.