A general data fusion architecture
Hervaldo Sampaio Carvalho, Wendi Heinzelman, Amy Lynn Murphy, Claudionor Coelho · 2003
Abstract: Data fusion is an important component of applications for systems that use correlated data from multiple sources to determine the state of a system. As the state of the system being monitored and available resources change, the general data fusion framework should change dynamically based on the current environment and available resources in the system. To achieve this goal, we have proposed a general Data Fusion Architecture (DFA) based on the Unified Modeling Language (UML) and using a taxonomy based on the definitions of raw data and variables or tasks. The DFA can be reconfigured according to the measured environment and availability of the sensing units or data sources, providing a graceful degradation in the view of the environment as resources change. We have shown that we can apply the DFA to different domains and applications, including a test bed health monitoring application. Keywords: Data fusion architecture, adaptation, UML, resource management.