Cognitively-Engineered Multisensor Data Fusion Systems for Military Applications

Abstract

This research relates to the bodies of knowledge on multisensor data fusion, cognitive engineering, decision aiding, and user interface design.The following sections present an overview of research on these topics. Multisensor Data FusionThe basic concept of fusing information from multiple sources is not new.For millions of years, animals, including humans, have used multisensor data fusion to combine signals from the five senses with knowledge of their environment to create a dynamic model of their world and to help them interact with their surroundings (Grossman, 1998).This basic mechanism of combining data from different sensors has been extended to such areas as imagery, robotics, signal processing, and surveillance applications.It has been shown that, in many cases, using multiple sensors improves robustness and reliability, extends coverage (both spatial and temporal), increases the dimensionality of the measurement space, reduces uncertainty, improves resolution, reduces ambiguity, and improves detection performance (Grossman, 1998; Li & Wang, 2002).However, this task which comes naturally to humans and animals in the physiological realm is considerably more complicated from an engineering and system design perspective.In fusing multisensor data, researchers must consider how the sensors and noise are modeled, how information from different sensors is related and how it can be integrated, how the features from different sensors can be verified against each other, and how to select an