Sensor Fusion Dynamic Scenario Descriptor

Radiance proposes to design, implement and test multi-sensor (RF/IR/Vis) multi-platform data fusion algorithms, advancing the state of the art in Rao-Blackwellised Particle Filter (RBPF) algorithms in a dynamic Bayesian network framework. Our innovations provide a multi-sensor (RF/IR/Vis), multi-geometry picture of threat scenarios for C2BMC"s requirement for a single integrated picture of the battlespace. Our algorithms adapt with knowledge from one sensor applied to networked sensors, adapting with changes in operational environments and threats. Our algorithms offer potential enhanced performance and risk reduction for C2BMC and Ballistic Missile Defense RF and EOIR sensor elements. The data fusion algorithms we propose provide robust adaptation to changing physical environments, with seamless fusion of disparate and diverse sensor data, including diverse engagement geometries. RBPFs capture non-linear behavior with Particle Filter techniques integrated with linear algorithmic approaches where appropriate, resolving issues associated with traditional Particle Filter techniques, most significantly those due to particle depletion. The network will fuse data from multiple RADAR, airborne Electro Optical Infrared (ABIR-like), and satellite-based EOIR (PTSS-like) systems. We provide a Proof-of-Principle, demonstrating functionality and the ability to generate a multi-geometry, multi-sensor realization of the scenario that can adapt with network sensor data.