Stochastic fusion of heterogeneous multisensor information for robust data-to-decision

In this paper, based on the measure-theoretic probability theory and the theory of stochastic differential equation (SDE), a stochastic fusion framework is proposed for the heterogeneous sensor network for the purpose of robust decision making. In this framework, for each sensor, its sample space and the corresponding σ-algebra are defined. Then, random variables, which are designed to meet the requirements of the operation in the battle field, are defined over the pairs of sample space and its σ-algebra. After that, the conditional expectation is taken for those random variables conditional on the union of σ-algebras to finish the information fusion process. Furthermore, to make the decision making process more robust, the undesired uncertainty in the fused information is hedged out based on the theory of SDEs, before the fused information is used for the decision making.