SPEC: Simulating Acoustic Scattering for Wide Aperture Sparse Array Multibeam Receivers

A future deep-ocean bathymetry sensor could be developed from a swarm of disconnected, distributed receivers attached to Autonomous Surface Vehicles (ASVs) forming a very wide aperture multi-beam array. This system would be able to collect high-resolution readings at depths beyond traditional systems. Novel processing techniques must be developed to analyze the data collected by such a system, however to evaluate these algorithms, precisely localized ground-truth data must be available, which is currently not feasible to collect. Additionally, a swarm-based system is capable of reconfiguring its layout to optimize the data collection, however evaluating such a swarm controller requires rapid collection of data. This paper discusses efforts to synthesize the time series data needed for these use cases. Several acoustic simulation tools are evaluated for synthesizing backscatter and bi-static reflection data collected at arbitrary surface positions for multiple detector array and sea floor configurations. In addition, a purpose built solution, Scattered Pressure Echo Capture (SPEC), is described and compared with the established modeling tools. SPEC is able to generate time series data for all examined seafloor configurations, however the other tools are unable to model the deep configuration. In order to provide a reasonable comparison, a scaled version of the problem is developed based on a physical configuration reported in the literature. The tools are used to synthesize receiver data for the scaled version. Near-field beamforming is performed, focusing the receivers at the known bathymetry, and no additional processing is performed. The results are evaluated based on the in-phase signal producing an amplitude spike sufficient for predicting the true bathymetry. Error metrics are presented for each tool including SPEC. In addition, the results for the deep water configurations for which only SPEC was able to handle are given. None of the examined tools generated data from which perfect predications could be made, however this synthesized time series data may be adequate for evaluating future algorithms developed to process the data collected by a swarm-based wide aperture sea floor imaging system.