A cognitive group hierarchy game theoretic framework for bandwidth management

Probabilistic predictions are critical in practice on many decision making applications because optimizing the user experience requires being able to compute the expected utilities of mutually exclusive pieces of content. Hierarchy games for decision making are valuable where two or more agents seek their own goals, possibilities of conflicts, competition and cooperation. In most cases, decision making needs to happen real-time, and the latency allowance for predictions is restrictive. The quality of the knowledge extracted from the information available is restricted by complexity of the model. Hierarchy game theory framework enables complex modeling of data in probabilistic modeling. However, applicability to big data is complicated by the difficulties of inference in complex probabilistic models, and by computational constraints. We focus on applying probabilistic models to bandwidth management for Landsat sensors. It is important to recognize the inconsistencies between individual rationality and group rationality for this social as well as engineering problem. Hierarchical game theory models interactions where a situation affects players at multiple levels. Our paper discusses the effect of optimizing the selection of specific bands for faster throughput on mission capabilities. We discuss four levels of hierarchies involving Landsat bands, number of sensors, mission priorities, and responsible organizations. In each case each entity in the hierarchy is treated as a player or agent.