SBIR/STTR Award attributes
Over the Horizon Radar (OTHR) has been a deployed capability for over 3 decades. OTHR uses the ionosphere to reflect HF radar signals in order to illuminate objects (potential targets) beyond the horizon, giving it a potential effective range of several hundred to a few thousand kilometers. Understanding how the HF radar signals interact and reflect off the ionosphere is crucial to accurate target detection and geolocation, and current state-of-the-art techniques only provide a range accuracy of about ten to forty kilometers. Expedition Technology, Inc. (EXP) and Applied Ocean Science, Inc. (AOS) are pleased to provide Improved Geolocation for OTH Radar (IGOR), which incorporates an innovative. Machine Learning based generative model of the propagation channel in order to refine geolocation estimates and increase the accuracy of range and Doppler measurements. It is also expected that the generative propagation modeling will provide additional utility outside of the OTHR system. Leveraging EXP’s experience in ionospheric modeling, advanced signal processing, and machine learning, At a high level, the problem reduces to a nonlinear, multidimensional estimation problem, with noisy and imperfect data. Since they fundamentally operate in a nonlinear N-dimensional space, Machine Learning (ML) approaches often significantly outperform alternative or traditional approaches.
