Improved array weighting for short nested arrays

Abstract

Often linear acoustic arrays are comprised of shorter nested subcomponent arrays cut for higher frequencies. This reduces costs and complexity by trading-off performance, specifically in resolution, SNR, and grating lobes. Typically each sub-aperture is processed individually, and the results incoherently combined over a frequency bandwidth of interest. Here several methods are shown superior to conventional bearing time records on measured at-sea data. Only methods that can be implemented in real-time and on low size, weight, and power systems are considered. Min and product array processing across the sub-apertures results are shown to have advantages, even though the nulls and grating lobes are not optimally coordinated. Finally, convex optimization methods are used to create array weights, over the entire irregularly spaced array. Stability is enforced via a regularization criteria added to the objective function. [Work Supported by ONR and the Royal Swedish Navy.]Often linear acoustic arrays are comprised of shorter nested subcomponent arrays cut for higher frequencies. This reduces costs and complexity by trading-off performance, specifically in resolution, SNR, and grating lobes. Typically each sub-aperture is processed individually, and the results incoherently combined over a frequency bandwidth of interest. Here several methods are shown superior to conventional bearing time records on measured at-sea data. Only methods that can be implemented in real-time and on low size, weight, and power systems are considered. Min and product array processing across the sub-apertures results are shown to have advantages, even though the nulls and grating lobes are not optimally coordinated. Finally, convex optimization methods are used to create array weights, over the entire irregularly spaced array. Stability is enforced via a regularization criteria added to the objective function. [Work Supported by ONR and the Royal Swedish Navy.]