多传感器相关数据的分布式判决方法

Distributed Detection for Multisensor Correlated Data

提出多个传感器数据在相关情况下实现的一种分布式判决方法.通过使用正交变换方法实现对数据解相关,从而采用较为简单的似然比判决形式,从理论上证明了这种变换的可能性及提出构造相应变换矩阵的方法.通过对两种模型的仿真结果表明,该变换方法能有效地解除传感器数据之间的相关性,更好地进行分布式判决.

A new method of distributed detection is presented where data from multisensor are correlated. Orthogonal transformation is adopted for the decorrelation of data, so a simple likelihood ratio method can be used. The paper also proves the probability of transformation in theory and presents the transform matrix. Two kinds of models in simulation show the effect of the method in distributed detection systems.