ABSTRACT
This chapter discusses the distributed detection using statistically dependent heterogeneous sensor data and employs copula theory to address the issue of modeling heterogeneous dependence. It also discusses data modeling issues and describes the theory and application of distributed detection using dependent quantized data. A common framework for solving decision problems is to maximize the probability of detection for a predetermined constraint on the probability of false alarm. For the heterogeneous fusion problem, it is also important to model appropriately the statistical dependence and heterogeneity of sensor observations while designing the fusion rule. The first and foremost challenge when designing heterogeneous fusion systems is to adequately model the joint distribution of sensor observations. Using copula theory, one can express a complete joint probability density function for dependent sensor observations. Subramanian and colleagues have addressed the issue of multisensor dependence modeling for detection applications through the use of vines.
