An information theory approach to assess acoustic-environmental significance

Autonomous underwater vehicles (AUVs) rely heavily on the acoustic environment for tasks related to sensing, navigation, and communication. In advanced applications of these vehicles, such as those involving hazardous or highly dynamic regions, their internal representation of the oceanic environment can prove essential to the mission success. Our current method is a virtual ocean simulator described by environmental empirical orthogonal functions (EOFs), a pre-determined set of sound speed perturbations; and corresponding recombination coefficients, weights assigned to recreate a specific sound speed profile that can be updated via acoustic communication. Given the limited bandwidth of acoustic communications, it is necessary to transmit the most vital information to update the virtual ocean representation—the minimum necessary to most effectively capture the real-time variability. Here we present the framework for determining the acoustic significance and uncertainty of an environmental coefficient-EOF pair, using an upcoming experiment in the Beaufort Sea as a case study. [Work supported by the Office of Naval Research.]Autonomous underwater vehicles (AUVs) rely heavily on the acoustic environment for tasks related to sensing, navigation, and communication. In advanced applications of these vehicles, such as those involving hazardous or highly dynamic regions, their internal representation of the oceanic environment can prove essential to the mission success. Our current method is a virtual ocean simulator described by environmental empirical orthogonal functions (EOFs), a pre-determined set of sound speed perturbations; and corresponding recombination coefficients, weights assigned to recreate a specific sound speed profile that can be updated via acoustic communication. Given the limited bandwidth of acoustic communications, it is necessary to transmit the most vital information to update the virtual ocean representation—the minimum necessary to most effectively capture the real-time variability. Here we present the framework for determining the acoustic significance and uncertainty of an environmental coefficient-EOF ...