Modeling the Decision Quality in Sensor-to-shooter (sts) Networks for Unattended Ground Sensor Clusters

One of the most promising concepts being considered for use by the Objective Force of the U.S. Army is an automated Sensor-to-Shooter (STS) network. An STS network is a closed-loop, internal feedback system that links various suites of sensors deployed throughout a 3D battle space to a network of weapons platforms (shooters) using optimized communications pathways. The system decision to fire or not is based exclusively on the information generated by this network. Hence, the quality of this decision process is directly dependent upon the quality of the information used to support it. In this paper, we introduce a novel sensitivity analysis framework capable of assessing the marginal contributions to uncertainty made by the various processes and devices of an STS network. This approach extends earlier work in modeling data and process quality for multiinput, multi-output information systems that principally focused on reducing error rates. While this study represents a work-in-progress, we are optimistic that the results can be directly used to identify an information quality critical path defined as an end-to-end pathway through the STS network composed of those devices and processes whose marginal rate perturbations most affect the quality of the final information product at the decision point. Moreover, a simple ranking of these marginal rates can identify and prioritize locations in the network where effective information quality enhancements should be performed to maintain a high quality final information product. This approach will also provide valuable insights as to whether or not continued efforts to improve sensor device precision beyond current levels is warranted.