论文部分内容阅读
在分布式传感器检测融合系统中，各传感器将其对同一假设的判决传给融合中心，该中心将这些判决结合起来形成最终判决，文中提出了使错误概率为最小的次最佳判决融合方案，它包括融合中心的最小错误概率（ＭＥＰ）检验和各传感器的似然比检验。基于这种融合方案，文中分析了不带判决可信度和带有判决可信度两种情况下的接收机工作特性（ＲＯＣ），结果表明：带有可信度信息的判决融合在高信噪比时其性能优于不带可信度信息的判决融合。
In the distributed sensor detection and fusion system, each sensor passes its decision on the same hypothesis to the fusion center, and the center combines these decisions to form a final decision. In this paper, a sub-optimal decision fusion scheme is proposed to minimize the error probability. It includes the Fusion Center’s Minimum Error Probability (MEP) test and the likelihood ratio test for each sensor. Based on this fusion scheme, receiver operating characteristics (ROC) are analyzed in the two cases without decision credibility and decision credibility. The results show that the decision with credibility information converges on the ROC Its performance is better than the decision fusion without credibility information when the noise ratio is higher.