Automated Distributed Data Fusion of Correlated Space Superiority Events

Period of Performance: 07/31/2012 - 05/06/2013


Phase 1 SBIR

Recipient Firm

Data Fusion & Neural Networks
1643 Hemlock Wy
Broomfield, CO 80020
Principal Investigator


ABSTRACT: To support the JSpOC Mission System, AFRL needs to develop and test intelligent process management software that integrates diverse sets of automated distributed data fusion tools for multi-source correlated space superiority events. These tools need to be assessed during their use and managed in near-real-time to maintain distributed consistency and improve their performance based upon operator and contextual feedback. To do this affordably AFRL needs to develop an intelligent (i.e., learning) system that performs Data Fusion-Process Assessment & Process Management (DF-PA&PM) of the space superiority system for co-orbital/proximity, NFL, RFI, and directed-energy scenarios. DF&NN proposes to deliver a prototype of such an intelligent distributed fusion management system operating on scenarios driven by space catalog, NASIC-based intelligence, GPS, ALPS, space weather (processed by DF&NN E-SAS/ACU/CAOS-D), and LMC provided WARGODDESS-based SIGINT, IMINT, & MASINT data, plus SBSS & SBIRS OPIR data. DF&NN proposes to deliver: 1. Space superiority scenario real and simulated data and/or events with truth events and relationship tracks and their characterizations 2. Multi-source distributed event and relationship fusion test articles based on DF&NN Bayesian Fusion Node and Charles River Integrated Mining Service 3. DF-PA&PM intelligent software 4. Performance Assessment and Management evaluation of these test articles on selected scenarios BENEFIT: There is significant need for the JSpOC Mission System (JMS) and across DoD for the intelligent assessment and management of distributed Data Fusion & Resource Management (DF&RM) systems that will enable each DF&RM system to adapt its capabilities over time to meet changing mission requirements. This research will provide the technical architecture functional decomposition of this capability and how it integrates with other higher level DF&RM functions. The software products of this work will include distributed multiple source entity and relationship assessment (i.e., Level-1&2 DF), DF-PA&PM Level-4 DF&RM, and Level-4 PAPM tools which will be extendable to other government and commercial applications. A prime example is the application to LMC government sites where such products can be used to cue multi-INT sensors and countermeasures. DF&NN also plans to extend this HLF Bayesian Fusion Node (BFN) and DF-PA&PM software to automated distributed fusion on law enforcement Computer Aided Dispatch systems. Without this interoperability, decision makers across jurisdictional boundaries are blind to the real time overall operational picture. The HLF and DF-PA&PM intelligent software will also provide these capabilities for Homeland Defense, counter-narcotics, power-grid, and numerous other DF&RM systems management.