Review of Methods and Algorithms for Dynamic Management of CBRNE Collection Assets.

摘要

The Defense Threat Reduction Agency (DTRA) provides reachback support to warfighters combating threats from chemical, biological, radiological, and nuclear weapons. DTRA is actively developing technologies to enable and support real-time integration of data from heterogeneous networks of CBRN (and non-CBRN) sensors positioned in and around an area of operations with physics-based modeling capabilities. The integrated toolset being pursued by DTRA must be applicable to a wide range of operational scenarios including direct force protection, targeted search, long-term threat behavior monitoring and wide area search. These integrated systems of sensors and models will require the development of automated methods (algorithms) for combining sensor information with physics models to perform critical functions such as threat detection, classification, identification, or localization. Development of algorithms that perform supporting functions such as data triage, null hypothesis generation, and reallocation of sensing resources will also be required. Algorithm development is required in order to ensure that the analytic reachback capabilities keep pace with developments in network, sensor, and computation capabilities. This report reviews the current state-of-the-art in algorithms and supporting methodologies and suggests possible directions for future development.