马永涛, 苗新龙, 高政, 裴曙阳
AuthorsHTML:马永涛, 苗新龙, 高政, 裴曙阳
AuthorsListE:Ma Yongtao, Miao Xinlong, Gao Zheng, Pei Shuyang
AuthorsHTMLE:Ma Yongtao, Miao Xinlong, Gao Zheng, Pei Shuyang
Unit:天津大学微电子学院，天津 300072
Unit_EngLish:School of Microelectronics, Tianjin University, Tianjin 300072, China
Abstract_Chinese:在室内无源超高频RFID定位中, 多径传播对定位结果的影响不可忽视．为了提高复杂多径环境下的定位精度, 提出了基于半定规划(SDP)和二阶锥规划(SOCP)的相位差欧式距离拟合定位算法．运用参考标签的位置信息, 对参考标签之间的理想相位差欧氏距离与实际距离进行拟合, 实现标签之间的距离估计．建立待定位标签与阅读器、参考标签距离估计模型, 将多径传播、高斯白噪声、相位差欧氏距离拟合引起的距离估计误差等效为正态分布, 利用参考标签的相位差信息和位置信息计算出正态分布参量．仿真结果表明, 提出的算法在最大多径数目为7时, 定位误差有90% 的概率低于2.3 m, 定位性能优于传统的基于测距的定位算法．
Abstract_English:The effect of multipath propagation can’t be neglected when it comes to localization in indoor passive UHF RFID system. To improve the localization accuracy in dense multipath environment，the phase difference Euclidean distance fitting localization algorithms based on semi-definite programming(SDP)and second-order cone programming(SOCP)were proposed. The location information of reference tags was utilized to establish the relationship between ideal phase difference Euclidean distance and the real distance of reference tags. Therefore，the distance between tags could be estimated. The distance estimation model of tracking tags，readers and reference tags was established. Then，the effect of multipath propagation，Gauss white noise and the phase difference Euclidean distance fitting on distance estimation were modeled as normal distribution. Simulation results show that the localization error of the convex relaxation method is less than 2.3 m with the probability of 90% when the maximum number of multipath is 7，which outperforms the traditional range-based localization algorithms.
Keyword_Chinese:无源标签; 相位差欧氏距离; 半定规划; 二阶锥规划
Keywords_English:passive tags; phase difference Euclidean distance; semi-definite programming(SDP); second-order cone programming(SOCP)

PDF全文下载地址:http://xbzrb.tju.edu.cn/#/digest?ArticleID=5911