Quality of information for data fusion in net centric publish and subscribe architectures

This paper examines data fusion and target tracking issues involved within net centric publish and subscribe architectures with respect to the quality of information (QOI) provided to the end user. These architectures are commonly selected in shared knowledge environments to enable access to all information by all users. DoD, DOJ, FAA, and other government organizations are pushing net centric concepts to move from stovepiped, closed architectures with limited data accessibility toward open architectures where dissemination of new data sources is intrinsically supported. Although these net centric initiatives (such as DoD's GIG) promise tremendous aggregate bandwidth, sensor-to-user bandwidth, end-user processing power and storage capacity, and other real world issues are expected to impact operations. For example, the soldier of the future may require moving target updates on a bandwidth-limited wireless PDA. This paper examines how the Air Force Research Laboratory's Joint Battlespace Infosphere (JBI), a publish-and-subscribe system in conjunction with the OGC's sensor Web enablement (SWE) initiative, can be used to update clients in a quality-of-information (QOI) paradigm rather than a quality-of-service (QOS) paradigm. We conclude that net centric architectures can support operational flows that improve the quality of information while reducing resource usage over traditional stovepipe systems. We also conclude that net centric tool implementations should specifically address QOI.