In this paper, we address information exchange problem in heterogeneous fusion networks (decision networks). A fusion network is a set of connected nodes in which fusion nodes (decision-agents: DAs) consume information produced by other sources nodes (e.g., sensors, other fusion nodes), and information is exchanged across a web of connected nodes. Information value is assessed based on partial utility function. This value, representing the DA's utility, is modeled as a time depending function. Routing in a fusion network is not just about getting data from one point to another. Routing needs to optimize a set of end-to-end goals driven by the application requirements, while considering network resources. We model this problem as a bi-objective optimization problem that maximizes the overall utility of the network and reliability of the generated paths. A multi-objective genetic algorithm (MOGA) is proposed to solve such an NP-hard problem. The empirical results are also presented.
展开▼