多传感器决策融合系统中的决策空间优化划分研究

在假定各局部检测器的决策规则已经给定以及在Ｂｈａｔａｃｈａｒｙｙａ距离最大的意义下，对多传感器融合系统中的决策空间优化划分设计进行了研究。首先基于Ｂｈａｔａｃｈａｒｙｙａ距离准则，把对整个系统决策空间的优化划分解耦为分别对各局部检测器决策空间的优化划分；然后从理论上证明了这种划分设计在最大Ｂｈａｔａｃｈａｒｙｙａ距离意义下的最优性，以及这种基于最大Ｂｈａｔａｃｈａｒｙｙａ距离准则进行优化划分设计的合理性；最后，通过对瑞利起伏环境下信号检测融合问题的数值计算表明，本文方法的性能优于基于Ｊ－散度方法的性能。

OPTIMAL DECISION SPACE PARTITIONING IN MULTISENSOR DECISION FUSION SYSTEM

Under the assumption that the decision rule of each local detector has been given, this paper presents an optimum method for partitioning the decision space in the sense of maximizing the Bhattacharyya distance criterion at the fusion center in multisensor fusion systems. First, based on the Bhattacharyya distance measure, the partitioning of the system decision space is decoupled into the partitioning of local decision space. Then the optimality in the sense of maximum Bhattacharyya distance measure and the reasonableness of selecting Bhattacharyya distance as the objective function in the design of distributed decision fusion systems are proved. Finally, simulation is made and the results show that the performance of the method in this paper outperforms that of using J divergence criterion.