基于稀疏信道重建技术的相邻多目标序贯检测算法研究  

Research on sequential detection of adjacent multi-targets based on sparse channel estimation

作　　者：吴培荣[1] 王昭辉[2,3] 蔡惠智[2] 刘云涛 张秋生[2] 

机构地区：[1]东南大学无线电工程系,南京210018 [2]中国科学院声学研究所,北京100190 [3]中国科学院研究生院,北京100190

出　　处：《声学学报》2010年第3期359-365,共7页Acta Acustica

摘　　要：为进行强目标相关旁瓣干扰下的相邻弱目标检测,采用稀疏重建理论进行多目标方位估计。高信噪比情况下,由方位估计结果即可完成目标检测;对于低信噪比弱目标回波,为提高系统检测能力,结合方位估计结果,提出了两种检测算法,前者类似于传统CLEAN算法,从能量角度进行目标检测;后者则利用相关函数能量集中于主瓣的特点,通过计算将目标方位估计结果中非零元素置零前后匹配滤波峰值的差值,采用Page-test序贯检测器进行多目标检测。仿真和试验数据处理结果表明,相同检测概率下,第二种方法具有更加优良的弱目标检测性能。High sidelobe peaks of auto-correlation function in the matching filter output incur interference to the detection of other adjacent targets. To achieve adjacent targets detection, sparse reconstruction theory was employed to obtain targets＇ position estimation. In the scenario with high SNR （signal to noise ratio）, estimated results can be used directly for targets＇ detection; otherwise, i.e. in the scenario with low SNR, two detection algorithms are proposed. The first one, similar to CLEAN algorithm, utilizes the energy changes of matching filter output to determine the presence of targets, while the second one, considering the feature of auto-correlation function, combines main lobe change calculation and Page-test sequential detector to achieve multi-targets detection. Simulation and experimental data processing results demonstrate that the second algorithm achieved better weak targets＇ detection performance.

关 键 词：多目标检测 序贯检测器 稀疏信道 测算法 重建技术 目标方位估计 弱目标检测 CLEAN算法 

分 类 号：U666.7[交通运输工程—船舶及航道工程]