一种有效的用于雷达弱目标检测的算法  

An Effective Track-Before-Detect Algorithm for Dim Target Detection

作　　者：强勇[1] 焦李成[1] 保铮[1] 

机构地区：[1]西安电子科技大学雷达信号处理重点实验室,陕西西安710071

出　　处：《电子学报》2003年第3期440-443,共4页Acta Electronica Sinica

基　　金：国防十五预研基金 :(No 41 30 70 50 4 ) ;国防科技预研跨行业基金 (No 51 4 0 70 2 0 80 1DZ0 1 1 9) ;国防科技重点实验室基金(No 51 4 31 0 50 2 0 1ZS0 1 0 4 )

摘　　要：低信噪比的环境中对弱小目标的检测和跟踪已经在雷达信号处理领域中引起人们的广泛关注 .检测前跟踪 (TBD)方法对弱小目标的检测非常有效 ,其基本思想是为避免恒虚警率 (CFAR)处理带来的信噪比损失 ,而直接对原始数据进行处理 .但这样做运算量是很大的 ,多数场合难以接受 .本文提出了CFAR和TBD联合检测的思想 ,即加入低门限CFAR检测和求秩过程对数据进行预处理 ,引进平滑度的概念以消除一些为伪航迹 .仿真结果表明此算法的运算量较小而比一般的检测方法SNR性能上约提高 2～ 3dB .An important problem in the field of radar signal processing which arouses people's attention widely is the detection and tracking dim moving targets in very low signal-to-noise ratio (SNR) environment.Track-Before-Detect (TBD) approaches are quite effective on the detection of dim targets,and the basic thought of this method is avoiding of the loss of CFAR by processing the raw data directly.An idea of detecting targets by combining CFAR with TBD methods is presented.In other words,low-threshold CFAR and process of ranking before TBD are added,and the conception of smoothness is introduced in order to get rid of false trails.Simulation results show this approach has small amount of computation,and achieves 2～3dB SNR gains comparable to those from other TBD algorithms.

关 键 词：检测前跟踪 动态规划算法 弱目标检测 雷达 信号处理 

分 类 号：TN957.51[电子电信—信号与信息处理]