论文部分内容阅读
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。
In order to improve the performance of distributed CFAR detection based on binary local decision, a new type of distributed CFAR detection scheme based on local observed signal-to-noise ratio (called R-type scheme) is proposed. It features local processing using the CFAR algorithm to form locally observed SNR estimates, which are then passed to the data fusion center. Compared to S + OS, the R solution not only reduces the amount of traffic between the local processor and the data fusion center in half, but also relaxes the requirements of the local observation more than the S + OS. Under three typical background conditions and two local processors, one of the schemes is analyzed: OS-R-CA, and its closed-form solution to the detection and false alarm performance is deduced and compared with COS , S + OS and other distributed CFAR detection performance comparison. The result shows that the detection performance and false alarm control ability of OS-R-CA are close to S + OS.