分布式贝叶斯检测融合优化算法及其计算机仿真

A Distributed Detection Fusion Algorithm under Bayesian Criterion and Its Computer Simulation

该文研究了利用分布式多传感器获得全局决策的分布式信号检测问题。在这种检测系统中各传感器将其各自关于观测对象的决策传送至融合中心,融合中心根据融合规则给出全局决策。研究重点是基于贝叶斯准则的分布式并联检测融合系统的数据融合理论,给出了使系统全局最优的融合规则和传感器决策规则,提出了对融合规则和传感器决策规则进行优化计算的非线性高斯-赛德尔算法,具体讨论了两相同传感器、两个不同传感器和三个相同传感器在具有独立观测时的数据融合问题。给出了利用本文所提算法对上述几种情况进行计算机仿真的仿真实例。仿真结果表明:融合系统的性能相对传感器有显著改善,采用三个相同传感器的融合系统,其贝叶斯风险下降了26.5%。

In this paper, the signal detection problem is considered when distributed sensors are used and a global decision is desired. Local decisions from the sensors are fed to the data fusion center and the center then yields a global decision based on a fusion rule. The data fusion theories under Bayesian criterion are researched, and the focus is placed on the parallel structure. Fusion rules at the fusion center and the decision rules of sensors are presented. A nonlinear Gauss-Seidel algorithm for the optimal computation of the fusion rules and the decision rules of sensors is proposed. Then, computer simulation for the algorithm is achieved for the fusion cases with two identical sensors, two different sensors and three identical sensors. The results of the computer simulation show that the performance of the fusion system, as compared with the sensor, has been significantly improved. For the case there are three identical sensors in the fusion system, Bayesian risk is reduced by 26.5%, compared with the sensor.