Narrowband underwater vector sensor using michroelectromechanical systems

Abstract

A narrowband underwater vector sensor using microelectromechanical (MEMS) systems is demonstrated. A combination of two pressure gradient MEMS sensors and an omnidirectional hydrophone allows for determination of the direction of arrival (DOA) of income sound over 360 degrees azimuth. The MEMS sensors, inspired on the hearing system of the Ormia-ochracea fly, consist of two wings connected by a bridge and anchored to the substrate by a tortional bean. They are operated with open back to allow a cosine dependence with the angle of incidence in the predominant bending vibrational mode. In the vector sensor, two MEMS sensors are orthogonally arranged to provide a cosine and sine directional pattern. When combined with an omnidirectional hydrophone that shares the same phase center, an arctangent estimator is used for DOA determination. The MEMS sensors are operated near resonance to filter out noise coming from undesired bands and improve the signal-to-noise ratio. A correction algorithm is applied to compensate frequency response differences. Two configurations have been demonstrated with the MEMS sensors enclosed in silicone oil and in air. The standard deviation of the measured DOA error has been computed for both configurations to be less than 5 degrees for various types of sound stimuli.