Distributed Battle Management Techniques

The Decisive Analytics Corporation (DAC) Team has a unique combination of skills, technology, and experience that allows us to demonstrate an advanced approach for distributed sensor data fusion in Ballistic Missile Defense (BMD) engagements using Bayesian Networks. We seek to extend the success of our Phase I effort supporting the development of a C2BMC that is, as required by the original solicitation, "distributed in execution while being homogeneous in data and algorithms." In particular, we showed that an advanced framework for distributed probability modeling and inference known as Multiply-Sectioned Bayesian Networks (MSBNs) may be used to support distributed object discrimination in the future BMD System. We demonstrated the trade-offs between communication requirements and solution accuracy through various implementations of this framework using a small-scale dynamic Bayesian Network. Through this Phase II SBIR effort, we propose to develop the MSBN framework to support the specific requirements for each of the Hercules discrimination models currently under development, and to demonstrate its effectiveness in an environment suitable for live fire testing.