Adaptive sensor fusion with nets of binary threshold elements

Kam, Naim, Labonski, Guez
1989 International Joint Conference on Neural Networks  
Distributed detection and estimation, and sensor fusion, are at present topics of intensive research due to the variety of systems which employ different sensing devices at geographically separated sites for tasks such as target detection and tracking, diversity in communications and threat classification. In this study, we demonstrate a simple distributed-detection scheme whose probability of error can be calculated analytically, and show that it corresponds to a two-layer network of binary
more » ... eshold elements. We proceed to assume that the sensors and the fusion center are subject to sudden unpredictable changes in the environment that they survey, and show how learning algorithms can be used in order to maintain good performance, in spite of these changes. We conclude with an example, involving five unequal sensors which distinguish between two time-varying Gaussian populations of different means.
doi:10.1109/ijcnn.1989.118678 fatcat:c5puyco6wbbwnjlhjn5dmaoehq