一种全局最优的多感测器检测融合算法  

An Algorithm for Global Optimization of Multisensor Detection Fusion

作　　者：刘英坤[1] 冯新喜[1] 党宏刚[1] 潘平俊[1] 

机构地区：[1]空军工程大学电讯工程学院,陕西西安710077

出　　处：《现代防御技术》2008年第1期50-53,共4页Modern Defence Technology

摘　　要：在实际检测中,由于带宽、信道衰落和信道噪声等因素的影响,从本地检测器至融合中心的信道通常无法保证为理想传输信道,另外由于本地检测器和融合中心的距离比较远,会加剧信道的非可靠性和增加信道的传输能量,于是就在本地检测器和融合中心之间增加中继节点。然后利用贝叶斯规则对所有节点的判决规则进行联合最优化。最后通过仿真对系统的检测性能进行了比较。In the actual detection, due to bandwidth, channel fading and channel noise, the ideal channel isn＇t assured from the local detector to fusion center. In addition, owing to the distance between the local detectors and fusion center being further, this can make the reliability of channel worse and the energy of channel transmission increasing, so the intermediate nodes are added from local detectors to fusion center. Then the detection rule of every node is optimized jointly according to Bayes rule. At last the detection performance is compared by simulation.

关 键 词：贝叶斯风险 二进制对称信道 全局最优化 

分 类 号：TN919[电子电信—通信与信息系统] TB114.1[电子电信—信息与通信工程]