基于NEYMAN-PEARSON准则的最优分布式量化检测融合算法

研究分布式NEYMAN－PEARSON量化检测融合系统的性能优化问题，融合系统由融合中心及多部传感器构成。各部传感器对同一目标或现象进行观测，并将量化后的观测信息传送至融合中心，融合中心对各部传感器的量化信息进行融合，并作出系统的最终判决。文中推导了融合系统检测性能最优化的必要条件。并在此基础上给出了各部传感器的最优量化规则，采用这一最优量化检测方法对分布式水声信号检测系统的性能进行优化，所得系统性能明显优于单部声呐的检测性能。

Optimum Detection Fusion Algorithm for Distributed and Quantized Neyman-Pearson Detection Systems

This paper considers the optimization of system performance for distributed and quantized Neyman Pearson detection systems. The distributed detection system consist of multiple sensors and a fusion center. Each sensor observes the same phenomenon and transmits its quantized observation to the fusion center. The fusion center integrates the quantized sensor observations and makes a final global decision. The necessary conditions for optimum detection for the distributed detection systems are derived, and the optimum sensor quantizer mappings are obtained. Using the optimum detection scheme to optimize the performance of a distributed underwater signal detection system, the system performance obtained is much better than that of individual sonars.