Optimal array element localization

Abstract

Measuring ocean acoustic fields at an array of sensors allows the application of advanced signal processing methods such as beamforming and matched-field processing. However, these methods require accurate knowledge of the positions of individual sensors. Typically, an acoustic survey is required to localize the sensors, a procedure referred to as array element localization (AEL). AEL generally consists of measuring the traveltimes of acoustic signals transmitted from a series of known positions to the sensors to be localized. These traveltime data can then be inverted for estimates of the sensor locations using a linearized inversion algorithm. An important issue in AEL is designing the configuration of acoustic source positions: a well-designed source configuration can produce substantially better AEL results than a poor configuration. In this paper, a procedure is developed to determine optimal AEL source locations for arrays of any configuration. This procedure minimizes the condition number of the Jacobian matrix of partial derivatives on which the linearized inversion is based. Minimizing the condition number in effect minimizes the magnification of data (traveltime) errors into model (sensor position) errors in the inversion. Monte Carlo simulations verify that optimal AEL source configurations can lead to significant improvements in sensor localization.