Distributed Battlespace Management Systems

"The military battlespace is a complex, dynamic, and open multiagent system requiring timely and distributed information fusion and decision making. The objective of this project is to develop a distributed battlspace management system using the advancedmultiple target tracking, the recursive modeling method, the dynamic programming, and Bayesian learning techniques. The innovations of this project include: 1) application of a Markov Chain Monte Carlo data association approach for solving complicatedmultitarget-multisensor tracking problems using inexact inputs received from diverse sensors; 2) development of a fuzzy Bayesian network to create a situation modeling structure and computational architecture for uncertain inference; 3) development of arecursive modeling method for enabling an agent to select its action and to coordinate with other agents by modeling their decision making in a multiagent domains; 4) application of a dynamic programming technique for solving the optimization problem forautonomous agents; 5) application of Bayesian learning for updating the agent's belief about the other agents by revising the probability concerned with the other agent's model based on their observed behavior; and 6) testing, demonstration and evaluationof the performance of the multiagent battlespace decision making system using a standard battlespace scenario in an intelligent virtual simulation environment. The research results of this S