Deep-sea channel estimation algorithm for orthogonal signal division multiplexing based on fast cluster sparse Bayesian learning

To address the cluster sparsity and long delay characteristics inherent in deep-sea underwater acoustic channels, an orthogonal signal division multiplexing (OSDM) channel estimation algorithm based on fast cluster-sparse Bayesian learning is proposed. Firstly, the sparse Bayesian learning (SBL) algorithm is introduced into OSDM communication systems for channel estimation to achieve enhanced sparse robustness. Subsequently, to effectively exploit the cluster-sparse characteristics of deep-sea channels, the block sparse Bayesian learning (BSBL) framework is adopted. Finally, the iterative mechanism and the sparsity of the estimation results of BSBL are improved via a hyperparameter weighting strategy and dual sparsity constraints. This modification not only substantially accelerates the convergence but also further improves the channel estimation performance. Simulation results based on Bellhop demonstrate that the proposed algorithm reduces the iteration count by approximately two-thirds compared to the traditional BSBL algorithm and achieves a bit error rate on the order of 10^-4 at a signal-to-noise ratio of 18 dB.