Cognitive mission planning and system orchestration

We present an approach to increase likelihood that an automated reasoning system will make the best decisions for a system and sensor resource manager (SSRM) with limited resources. The introduction of automated game theory to mission planning allows for a modular architecture and scalable approach across a range of platforms and facilitates rapid integration of new modes and technology. Game theory enhances a multi-objective management system, enabling it to intelligently coordinate and optimally achieve disparate mission objectives of a converged system. Our solution may also save money by offering a Pareto efficient process for mission plan validation, which reduces time needed to validate scenarios. An area of significant contribution in our work is our new genetic linear optimization, which is an innovative, developmental extension of our efforts comparing genetic algorithms with linear programming. We show results in receiver-operating characteristic (ROC) curves and compare the performance. We also quantify the performance improvement across all combinations of decisions or players in a linear program. We further enhance decision making by combining the best parts of multiple solutions. Our design provides insight into the most important or optimal weights for degrees of freedom and can be used to efficiently tune those weights. We also discuss tradeoffs for improving processing time with a sensitivity analysis.