Synthetic Aperture Sonar (SAS) provides high-resolution seabed imagery but is extremely sensitive to platform motion. Among translational perturbations, vertical heave introduces strong phase errors that broaden the main lobe, raise sidelobes, and generate ghost targets, while most existing work relies on navigation sensors, motion estimation, and autofocus to compensate these errors after data acquisition. Here we investigate a complementary, geometry-driven approach: designing receive arrays whose three-dimensional sensor placement intrinsically suppresses heave-induced phase error before any postprocessing is applied. Starting from a cylindrical multi-receiver SAS configuration, we derive a first-order sensitivity measure of the array response to vertical displacement and use it as a design objective to select sparse sensor subsets from an underlying dense grid. Two compact, fabricable heave-robust arrays are obtained: (i) an anti-heave dual helix with z-symmetric sensor pairs, and (ii) a heave-nulling dual ring with azimuthally interleaved elements on two vertically separated rings. Point-target simulations for a 3×3 grid under sinusoidal heave with 0.5 m peak amplitude show that, for the same number of sensors, these layouts reduce phase-error RMSE by roughly 40–50% and yield PSLR/ISLR improvements of about 10–13 dB, substantially relaxing motion-compensation requirements and improving SAS image robustness in realistic sea states

Keywords