Bearing estimation using small tetrahedral passive hydrophone array

The NILUS (Networked Intelligent Underwater Sensors) demonstrator system for autonomous underwater sensor networks contains sensor nodes equipped with a tetrahedral passive hydrophone array. The dimensions of the array are small compared to the wavelength of the received acoustic signals. NILUS has been deployed in several sea trials. In the trials, sensor data have been recorded for post-trial analysis in addition to being processed directly at the nodes to provide on-line detection information and bearing estimates, which are sent to an operator through an underwater acoustic communication network. This paper describes the algorithms used for the estimation of bearing (azimuth) and elevation for the tetrahedral array system and the performance based on real-world data. The observed performance is supported by simulation results. Two different methods are used to process the tetrahedral array data. In the first method adaptive beamforming using MVDR (Minimum Variance Distortionless Response) is applied, and the dominant bearings of the beampower output are extracted using a histogram-based algorithm. The second method is based on phase gradients. This algorithm is able to estimate one bearing (azimuth) angle, and also elevation.