Dynamic Adaption of Data Fusion Algorithms

The ORINCON team's proposed research and development addresses the application of adaptive data fusion to the surveillance problem of high-density targets in a cluttered environment; our effort focuses on robust data fusion for surveillance through algorithmic adaptation. Our overall goals are to (1) extend, develop, and control the fuzzy controller for the adaptive data fusion problem; and (2) develop and evaluate alternative architectures for adaptive fusion systems. We propose to extend the development of a fuzzy controller to include time constraints for the solution based on the rate of incoming reports. By monitoring surveillance problem characteristics, such as number of targets, and assignment problem matrices (number of targets and contention), the controller selects either an MHT or an ND approach and specifies algorithm parameters. Thus, the controller adapts to the problem at hand and selects the best fusion approach given the characteristics of the problem. In addition, we intend to focus our research and development toward the automated selection of alternative fusion algorithms.