Helios Remote Sensing Systems proposes to develop a method for passive 3D localization (range, elevation and azimuthal bearing) of an emitter located over a variable sea surface. Our effort focuses on developing machine learning methods for passive 3D localization of an overwater emitter under a variety of weather, sea-states, and ducting conditions from a single receiving surface platform. Our approach features an RF solution incorporating appropriately located sparse 2D antenna arrays utilizing spatio-temporal processing. The spatial measurements from multiple, appropriately placed, simultaneously collecting receivers are combined to decompose the RF multipath channel. Our solution for extracting out emitter characteristics includes learning algorithms, which are applied for spatio-temporal pattern recognition. These learning methods will be trained under a variety of antenna architectures and sea conditions to incorporate the complexities of the RF surface channel into the learning method during Phase I.
