A Dynamic Adaptive Hierarchical Game Theoretic Approach for Modeling and Prediction of Asymmetric

Intelligent Automation Inc. proposes a comprehensive and innovative approach for building dynamic intelligent modeling and prediction systems for asymmetric threats that accounts for enemy learning based on distributed hierarchical game theory. The primary goal is to develop an automated tool for enemy course of action prediction. We leverage novel game theoretic, deceptive-sensitive algorithms, emotion and cognition theory and ontology-based information structure to provide real-time enemy estimates to a tactical command. To achieve this goal, first, Exponentially Weighted Moving Average (EWMA) technique and multi-agent distributed learning are fused into spatial-temporal prediction model so that the trends/variations in enemy COA patterns can be identified and even predicted. Second, emotion modeling and deception reasoning are proposed so that the overall model can fit the non-rational reasoning and attacking patterns of terrorists such as suicide bombers. Third, hybrid data fusion is applied to process spatial-temporal information extracted from lower-level sensor data, which will allow the fusion of both structured and unstructured data. Fourth, ontology-based information representation is applied to facilitate the building of virtual battle space with less computation. Fifth, applying Markov game approaches can naturally model the coupling between conflicting parties and provide more accurate reasoning and predictions.BENEFITS: The proposed dynamic threat prediction tool for situation awareness has tremendous applications potential in many military applications. It can also be directly used for developing of advanced mission planning and emergency preparedness decision support systems. IAI is already a contractor on the Future Combat System (FCS) program. We are teamed with Honeywell on the FCS-Platform Soldier Mission Readiness System and with Northrop Grumman on FCS-logistic decision support system. We will leverage these relationships to identify the end FCS customer, and work with these teams to transition our Phase 2 technology into the FCS program. In addition, our industry partner, Lockheed Martin, MITRE can also transition these technologies. The market for military applications is quite large. Other potential commercial applications include law enforcement, border and coast patrol, intrusion detection, business intelligent, global awareness, decision support and the automated inferenceing tools used for diverse purposes from system failure prediction to marketing. The size of this market is not small and may grow rapidly with the commercial demand in homeland security. We expect the aggregated market size will be similar to that of military applications.