Channel Cluster Structure-Based Iterative Channel Estimation and Pilot Design for Deep-Water Acoustic OFDM Communications

In deep-water (DWA) communications, the channel often exhibits two long-separated clusters, with the second cluster exhibiting significant Doppler spread. DWA channels are commonly found in underwater communication networks, where the significant delay spread of the channel causes traditional OFDM channel estimation (CE) methods to severely degrade the communication rate. In this paper, we propose a channel cluster structure (CCS)-based iterative CE method for DWA Orthogonal Frequency Division Multiplexing (OFDM) systems. By leveraging the channel cluster structure, we perform CE in the presence of both inter-block interference (IBI) and inter-carrier interference (ICI) without compromising communication rate. Additionally, to address the problem of IBI deteriorating the measurement matrix, we propose a pilot design scheme that minimizes the coherence of the measurement matrix. We integrate the proposed method with the compressed sensing (CS)-based and the sparse Bayesian learning (SBL)-based algorithms, and apply it to DWA CE. Simulation results demonstrate that the CCS-based iterative CE method outperforms traditional method, and the proposed pilot design further improves CE performance.