On the performance of geo-acoustic estimation for a distributed sensor array

The vision underlying the Widely scalable Mobile Underwater Sonar Technology (WiMUST) project is that of developing advanced cooperative and networked control / navigation systems to enable a large number (tenths) of marine robots towing small acoustic arrays to act as a coordinated team for seismic sub-bottom imaging. The space-time coherent processing of bottom returns requires the ensemble of short acoustic arrays to be seen as a single spatially distributed sensor array. Since the vehicles are free to move along range, cross-range and depth the resulting distributed sensor array may take, at least conceptually, any spatial shape. With array shape freedom comes the question of which is the most suitable (or optimal) array geometry for sub-bottom imaging and inversion. The answer to this question hinges, among others, on the definition of performance of a seismic sub-bottom profiling system. Determining the optimal sensor array geometry is clearly a ill-posed problem, since the optimal geometry is itself bottom dependent, and there is no such environment as “one size fits all”. This work addresses several criteria for sub-bottom profiling system performance including gain, resolution and probability of detection. Two physical models will be tested: one based on acoustic wave reflection used in traditional seismic imaging, and another normally used in matched-field bottom properties estimation, that includes propagation and refraction. Simulations to support the theoretical developments and algorithms were obtained on a scenario inspired in a real environment off the coast of Peljesac (Croatia).