Sparse MIMO synthetic aperture sonar processing with distributed optimization

Synthetic aperture sonar (SAS) provides high-resolution seafloor mapping by processing coherently the backscattered signal recorded over consecutive pings. Multiple-input multiple-output (MIMO) configurations improve the spatial sampling along the synthetic aperture at the expense of reduced image contrast due to the cross-correlation between the transmitted waveforms. In this paper, MIMO SAS imaging is formulated as a sparse signal reconstruction problem and is solved in a distributed way with the alternating direction method of multipliers (ADMM). Sparse reconstruction with ADMM outperforms conventional MIMO SAS imaging in terms of resolution and sidelobe suppression, alleviating the impact of waveform correlation.