Innovative Approaches to Fusion 2+

Tactical aircraft pilots are provided with ever increasing amounts of information about the battlefield and the threats they face, providing the opportunity for achieving overwhelming information superiority over our adversaries. This potential faces two difficult challenges: 1) each aircraft must be individually capable of synthesizing/fusing all available information into a coherent tactical situational awareness and threat assessment capability and 2) the consistency of the tactical picture for all aircraft involved in a mission must be assured. Overcoming these challenges will allow commanders to incorporate doctrinal changes that allow multiple aircraft to execute highly coordinated missions, greatly improving overall mission effectiveness. A fundamental aspect of any fusion process is the representation of the uncertainty of the data. In Phase I, we demonstrated a hybrid Fuzzy Logic/Bayesian fusion architecture that demonstrated the strengths of these two inference techniques for a single aircraft capability. In Phase II, we extend the Phase I architecture to allow multiple, distributed hybrid Fuzzy/Bayesian networks to be combined into effectively one virtual network through Multiply-Sectioned Bayesian Network (MSBN) techniques. This fusion 2+ architecture is dependent on the output of a level 1 fusion model and will be provided by the Canadian test bed called CASE_ATTI .