Signal analysis and data fusion methodologies

The presence of radioactive and nuclear materials (RN) can be detected, indentified or inferred through the detection of emitted neutrons and gammas. For Homeland Security applications a robust detection approach is required, with minimal false alarms and technology suited to the field of deployment. This paper explores methods of combining and analyzing signals from multiple detectors, applying data fusion principles to optimize the performance of detection systems based on passive interrogation. A signal analysis methodology using spectral fits as a means to combine data through fuzzy logic is described and a case study explored.