Active sonar imaging techniques are essential for underwater target detection and tracking, particularly in the complex acoustic environments of shallow water. One of the major obstacles in these scenarios is reverberation from the seafloor and the sea surface, which can significantly degrade image quality. A core objective in active sonar imaging is applying beamforming techniques to achieve a narrow mainlobe and low sidelobe levels. Although conventional beamforming (CBF) is widely adopted for its simplicity and robustness, it faces intrinsic performance limitations. This study reformulates the beamforming process as a single-snapshot complex-valued least absolute shrinkage and selection operator problem. To address this problem, the complex approximate message passing (CAMP) algorithm is proposed. CAMP enhances computational efficiency by avoiding matrix inversion and improves convergence speed by using the Onsager correction term. Experimental validation in an anechoic pool demonstrates that the proposed method achieves significant imaging performance compared to the CBF method. Additionally, the anechoic pool setup is leveraged to optimize the threshold parameter in the CAMP algorithm. Further validation of a monostatic configuration through lake trials confirms that the proposed method significantly suppresses sidelobe levels and improves image resolution, resulting in cleaner sonar images and improved suppression of background interference.
Zeng et al. (Sun,) studied this question.