基于信号检测法的NLOS识别  

NLOS Identification via Signal Detection Algorithm

作　　者：颜俊[1] 陈亮[1] 吴乐南[1] 

机构地区：[1]东南大学信息科学与工程学院,南京210096

出　　处：《信号处理》2009年第12期1848-1852,共5页Journal of Signal Processing

基　　金：高等学校科技创新工程重大项目培育资金项目"基于数字广播电视信号的无线电组合定位"(706028)资助

摘　　要：针对无线定位系统中的非视距传播(NLOS)问题,本文提出采用传统的信号检测方法对接收信号中的NLOS信号进行检测。结合Nokia公司的现场测量数据分析得到NLOS误差信号的统计特性,根据对先验知识的掌握程度,提出分别采用广义似然比检验(GLRT)和一致最大功效检验(UMPT)检测接收信号中是否存在NLOS信号,并分别进行了两种情况下的仿真验证。仿真结果表明:基于信号检测方法的NLOS识别技术具有时间开销小,识别能力强的特点。同时由于算法理论是建立在现场测量数据的基础上,所以算法具有工程应用价值。Non-line-of-sight （NLOS） identification carried significant importance in wireless localization systems. It has been discussed extensively in the former literature. In this paper, we proposed a novel non-line-of-sight （NLOS） identification technique based on conventional signal detection method. It is modeled with the statistical property of Non-line-of-sight （NLOS） signals which was discussed with the field measurements by Nokia. Generalized likelihood ratio（GLR） test and uniformly most-powerful（UMP） test which are the common tools for signal detection were respectively proposed to test if there was some non-line-of-sight （NLOS） signal in the received signal, according to a priori information. The algorithm was simulated in the two different situations at the same time. Simulation results indicate that the proposed algorithm, compared with those present algorithms, was more efficient and accurate. The results also suggest that it may have great value in project applications, because the theory analysis was based on field measurement.

关 键 词：定位 NLOS 信号检测 广义似然比检验 一致最大功效检验 

分 类 号：TN911.7[电子电信—通信与信息系统]