Rapid Reprogramming Technologies for Electronic Warfare Training

ABSTRACT: Current simulations are limited in their ability to simulate complex systems, not only from a capabilities perspective, but from a rational agents perspective as well. In order to"rapidly and accurately represent agile, reactive, and adaptable threats"significant improvements are required in the development and application of rational agents. BSI proposes to develop a rapidly-reprogrammable cognitive architecture for use in distributed simulations, in which a wide array of tactically relevant variables (including threat data, behavioral tendencies and environmental factors) are dynamically evaluated and a logical response posture is suggested. BSI will design and prototype the Common Cognitive Decision Architecture (C2DA), which will serve as the foundation for this new capability. At a high level, the C2DA should be capable of evaluating Decision Factors (DFs) and formulating a Decision Response (DR) at the entity level. The power to define exactly how these DFs are weighed by the C2DA must be put in the user's hands; thus the C2DA must be rapidly reprogrammable by the non-programmer user. In Phase I, BSI will research and develop an initial set of Decision Factors and Decision Responses, design a prototype graphical user interface (GUI) for the C2DA and a Run-Time Interface (RTI) for interoperability. BENEFIT: The successful completion of this research will result in the innovative application of rational agents to distributed mission simulation. The C2DA has the potential to bring a revolutionary improvement in the behaviors and responses of computer generated forces (CGFs), thus improving the overall efficacy of distributed mission training as used throughout the DoD and by our allies. BSI intends to take an architectural approach that is problem-domain agnostic, meaning that the C2DA could be adapted to non-military distributed (or stand-alone) simulations in the future. The potential applications of the C2DA are literally countless, as it could be used to externalize and expand any simulation that uses (or should use) rational agents to process decision logic.