Spatial rotation-based direction-of-arrival estimation for uniform circular hydrophone array

Abstract

When using a uniform circular array (UCA) to detect far-field underwater targets, rotational motion of the UCA caused by water currents during snapshot acquisition is inevitable. As a result, the array coordinates of a direction-of-arrival (DOA) estimate is time varying, and subspace-based methods that take a large number of snapshots for super resolution are inaccurate, especially for weak sources. This study makes use of the UCA heading information and the properties of the UCA steering vector to address this problem. A two-step spatial rotation procedure is devised to rotate the source DOAs from array coordinates to geodetic coordinates in real time, and consequently, the UCA remains virtually static. The results of simulation and real data experiments validate that our proposed technique allows high resolution subspace-based methods to be used when there is UCA rotation, and thus improves the performance of DOA estimation for weak sources.