相关条件下基于N-P准则的分布式检测理论研究

Study of Distributed Detection Based on N-P Rule with Correlated Sensors

本文针对并行网络、二元局部判决的情况,研究了在相关条件下基于N-P准则的分布式检测融合算法,给出了在联合概率密度已知和未知两种情况下的最优融合规则的理论推导及相应的解决方案,并在实验仿真的基础上,得到了部分有益的结论。

We study the fusion algorithms of distributed detection based on Neyman-Pearson criteria with correlated condition. We consider a parallel network structure that consists of a number of local sensors and a fusion center. Each sensor makes a binary decision based on its own observation. The detections with known and unknown joint pdf are considered. In both cases, we give the theoretical derivation of optimal fusion rules and corresponding solution. In the end, some useful conclusions are given based on the experiment.