分布式CFAR信号检测

Distributed CFAR signal detection

带有分布式传感器和融合中心的检测系统已应用于监视系统中．文中提出了多传感器检测空间的划分，以及先分区融合再系统融合的方法，并且在不同的检测区域，多传感器采用不同的门限；已往所有传感器探测同一区域，融合中心采用所有传感器判决的ＡＮＤ、ＯＲ和“Ｋ／Ｎ”规则，和这些规则相比，文中的方法更加合理和实用．利用Ｂｏｏｌｅａｎ代数，文中还推导了多传感器检测系统的检测概率和虚警概率表达式．对于由几个单元平均恒虚警率（ＣＡ－ＣＦＡＲ）检测器组成的系统，通过分区融合和系统融合，可以在融合中心得到总的判决．对于Ｒａｙｌｅｉｇｈ目标和干扰，固定系统的虚警概率为某一要求值，通过使系统的检测概率最大来获得每个ＣＦＡＲ检测器的最佳门限．最后，举例说明了这种方法的应用并分析了其性能．

Detection systems with distributed sensors and data fusion are increasingly being used bysurvelliance systems.In this paper,multiple sensor detection space partitioning is considered.An approach to the fusion in each detection area and then the fusion at system level is present-ed.Compared with methods available by which all sensors conduct detection in the same areaand AND,OR and“K out of N”fusion rules are used at the fusion center,the approach ismore reasonable and practicable.Boolean algebra is used to derive closed-form expressions forsensor-system detection probability and false alarm probability.For a system consisting of sev-eral cell-averaging constant false alarm rate(CA-CFAR)detectors,the decision fusions in ar-eas are used to obtain the overall decision at the data fusion center.For the Rayleigh target andinterference,the optimal threshold of the individual CA-CFAR detector is obtained for a fixedlevel of false alarm in the data fusion center in order to maximize the total probability of detec-tion.Finally,the application of the approach is illustrated with an example and an analysis ofits performance is also made.