基于局部观测信噪比的新分布式CFAR检测

为了提高基于二元局部判决的分布式ＣＦＡＲ检测的性能，提出一类新的基于局部观测信噪比的分布式ＣＦＡＲ检测方案（称为Ｒ类方案）。其特点是以ＣＦＡＲ算法做局部处理以形成局部观测的信噪比估值，然后将其传送给数据融合中心。相对于Ｓ＋ＯＳ，Ｒ类方案不仅使局部处理器和数据融合中心间的通信量减少了一半，而且对局部观测的要求也比Ｓ＋ＯＳ宽松。在三种典型背景环境中和两个局部处理器的条件下，分析了其中一种方案：ＯＳ－Ｒ－ＣＡ，推导出了它的检测和虚警性能的闭形解，并将其与ＣＯＳ，Ｓ＋ＯＳ等分布式ＣＦＡＲ检测进行了性能比较。结果表明，ＯＳ－Ｒ－ＣＡ的检测性能和虚警控制能力保持在了与Ｓ＋ＯＳ接近的水平。

New distributed CFAR detection scheme based on SNR of local observations

A new type of distributed constant false alarm rate (CFAR) scheme based on the ratio of signal to noise (SNR) of local observations (referred as R type scheme) is presented, to improve the performance of distributed CFAR detection based on binary local decision. Its characteristics are that CFAR algorithms are used in local processor to form the estimation of SNR of local observations, and then the estimation is transmitted to the data fusion center (DFC). Compared with the S OS, the R type scheme not only saves a half of the amount of data transmitted between local processor (LP) and DFC, but also overcomes the drawback of S OS, which is not suitable to the practical use. In three classical backgrounds, it analyses one of them, OS R CA for two sensor network, and derives the closed form solutions of detection and false alarm performance, and compares its performance with the COS and S OS distributed CFAR detections. Result shows that the detection and false alarm performance of OS R CA remains at the level similar to that of S OS.