The analysis and design of distributed detection based on star configurations

In this paper, the distributed detection problem for sensors with star configuration is systematically investigated under the assumption that the fusion rule is based on Neyman-Pearson (N-P) criterion. Then it is proved that local decision and local fusion rules should also be N-P tests in order to obtain global optimization. The computational method of the thresholds is presented for the case where the constant alarm false rate processors are used in each sensor. Finally some calculation examples are given, which demonstrate that the performance of the optimal distributed detection approaches to the one of central detection under unconstrained communication bandwidth. The results given in this paper are of important significance for the point or small area multisensor defense system.