SBIR/STTR Award attributes
With the delivery of the operational multi-static active coherent (MSAC) source to the NAVAIR anti-submarine warfare system, the opportunity for complex waveforms and real-time updating of the ping-sequencing for an active sonobuoy field will be available. In order to maximize the use of this revolutionary source, there is a need to understand how multiple ping types and sequences lead to the combined probability of detection (CPd) of the full field and the full ping cycle without the extraordinary cost of testing at sea. The current system planning algorithm does not include combine propagation angle and doppler for target echo or angle of incidence for surface or seafloor reverberation. There is a need to improve the fidelity of the modeling to determine the differences between a large assortment of ping signal types. In this SBIR, Applied Ocean Sciences (AOS) proposes to develop a system performance model for system CPd which incorporates the physics of doppler and travel time and coherent acoustic phase to accurately determine performance in a complex ocean environment. Reactive behavior of an intelligent adversary to pings received on the threat submarine will be developed and included in the system. The action and policy spaces and the ASW system (with ping selection) and the reactive adversary will be formulated into a reinforced learning (RL) algorithm to provide machine learning solutions for each. This ping-optimizer, in the presence of a reactive adversary, will be developed in Phase II demonstrating a 25% improvement in performance and transitioned to the FLEET via the ASPECT system in Phase III.
