Application of information-theoretic performance measures to optimization of array spatial configurations for underwater acoustic source localization

Abstract

Information-theoretic measures of acoustic source localization performance provide for performance comparisons that are valid across classes of signal processing algorithms and can be used as performance criteria for the optimization of receiver-array spatial configurations for source localization. This work investigates the use of fundamental information-theoretic quantities, including mutual information of source location and processor output and conditional entropy of source location given processor output, as performance criteria for the optimization of array configurations. Applications of these criteria to the optimization of horizontal and vertical arrays are illustrated in examples of Bayesian localization, conventional beamforming, and matched-field processing of the acoustic field of a time-harmonic source in a range-independent shallow-water waveguide. The optimized array spatial configurations are compared with results obtained using traditional (energy-based) performance measures. [Work supported by ONR.]