基于似然比检验的超宽带信道检测与定位算法

发布时间：2019-03-19 18:18

【摘要】：超宽带(Ultra Wide Band,UWB)室内定位系统的定位性能主要受信号非视距(None Line Of Sight,NLOS)传播影响。为此该文提出一种基于信道统计量(Channel Statistics Information,CSI)的信道NLOS状态检测法。该方法首先在IEEE802.15.4a信道模型下对均方根时延扩展和平均超量延迟的概率分布函数进行建模,作为信道标准分布。再以信道瞬时分布与标准分布间的KL散度为检验统计量做似然比检验(Likelihood Ratio Test,LRT)来鉴别信道状态。同时提出一种基于LRT的定位算法LRT-Chan算法。该算法能有效利用受NLOS污染的测距数据提高定位精度。仿真结果表明:LRT信道状态检测法在全部UWB信道中都能获得较高检测准确率;在定位锚点(Anchor Node,AN)分布不理想的NLOS环境中LRT-Chan算法也能取得较高定位精度。
[Abstract]:The positioning performance of ultra-wideband (Ultra Wide Band,UWB (UWB) indoor positioning system is mainly affected by the propagation of signal non-line-of-sight (None Line Of Sight,NLOS). In this paper, a channel NLOS state detection method based on channel statistic (Channel Statistics Information,CSI is proposed. Based on the IEEE802.15.4a channel model, the probability distribution functions of root-mean-square delay spread and average excess delay are modeled as the standard distribution of the channel. Then the KL divergence between the instantaneous distribution and the standard distribution is used as the test statistic to perform the likelihood ratio test (Likelihood Ratio Test,LRT) to identify the channel state. At the same time, a LRT-Chan algorithm based on LRT is proposed. The algorithm can effectively utilize the NLOS contaminated ranging data to improve the positioning accuracy. The simulation results show that the LRT channel state detection method can achieve high detection accuracy in all UWB channels, and the LRT-Chan algorithm can also obtain high location accuracy in the NLOS environment where the location anchor point (Anchor Node,AN) distribution is not ideal.
【作者单位】：桂林电子科技大学信息与通信学院;广西精密导航技术与应用重点实验室;卫星导航系统与装备技术国家重点实验室;
【基金】：国家自然科学基金(61362005,61561016) 广西自然科学基金(2013GXNSFA019004,2014GXNSFAA118352)~~
【分类号】：TN925

【相似文献】