Role of adaptive beamforming with random and irregular arrays

Abstract

When we design an array to accommodate as many as 4–5 octaves of bandwidth, the design must provide a variety of apertures and element spacings. Further, the concept of operations may be such that no structure is available for maintaining a regular array structure (even if we actually wanted one). One example of this is an array of freely drifting sonobuoys. Clearly, with a limited number of elements (far fewer than would be needed to provide half wavelength spacing for the highest frequency over the aperture needed at the lowest frequency), compromises must be made. Solutions include combinations of nested apertures, geometrically increasing spacings, and random spacings to varying degrees—in short irregular and random arrays. Such designs have very poor conventional beam response patterns, with very high sidelobes, compared to the fully populated arrays spaced at half wavelength. However, these objections are a red herring. We will show that adaptive beamforming produces beam responses with very low sidelobes. We will also show that irregular and random arrays indeed outperform regular arrays (of equivalent cost), if a wide bandwidth and range of interference directions are considered. We will also briefly review requirements for array calibration for adaptive beamforming.