Sparse Channel Estimation With Global Optimum Solution for Orthogonal Signal Division Multiplexing in Underwater Acoustic Communication

This paper proposes sparse channel estimation using basis pursuit denoising (BPDN) with a global optimum solution (GOS) for underwater acoustic (UWA) communication using orthogonal signal division multiplexing (OSDM). To realize reliable UWA communication, we need to address delay and Doppler spreads by utilizing the sparse nature of the UWA channel. BPDN recovers a sparse signal from the observed signal, which consists of the sparse original signal and noise components. Conventional OSDM using iterative-based BPDN suffers from large computational complexity and cannot fully utilize the sparse nature of the UWA channel. Focusing on the nature of the pilot vector, we introduced the use of BPDN-GOS for channel estimation. We conducted a simulation of UWA communication based on ray tracing and a sea trial to evaluate the communication quality of OSDM using BPDN-GOS. Our simulation and experimental results demonstrated that OSDM using BPDN-GOS reduced computational time and achieved better communication quality than OSDM with iterative-based BPDN.