马永涛, 高鑫
AuthorsHTML:马永涛, 高鑫
AuthorsListE:Ma Yongtao, Gao Xin
AuthorsHTMLE:Ma Yongtao, Gao Xin
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:被动式定位(DFL)中的射频层析成像(RTI)算法主要运用在无线传感器网络(WSN)中．而物联网的快速普及使得无线射频识别(RFID)网络随处可见, 所以本文提出了在无源RFID网络下实现RTI的被动式定位算法．针对多目标情况下的伪目标问题, 提出了一种基于Adaboost的机器学习算法来去除伪目标, 最终在目标数目未知的前提下实现至少3个目标的准确识别．实验结果表明提出的方法可行性很高, 在定位误差为0.7 m时目标识别准确度达到86% ．
Abstract_English:The radio tomographic imaging(RTI)algorithm in device-free localization(DFL)was mainly used in wireless sensor network(WSN). While with the rapid development of the internet of things，the wireless radio frequency identification devices(RFID)network can be seen everywhere. So the idea of realizing the RTI algorithm in passive RFID network is proposed. The Adaboost-based machine learning algorithm was presented to remove the false goal for multiple target situations. Finally，the accurate identification of at least three targets was realized on the premise that the target number was unknown. Experimental results show the feasibility of the proposed method is very high，and the target recognition accuracy reaches 86% when the position error is 0.7 m.
Keyword_Chinese:被动式定位; 射频层析成像; 无线射频识别; 机器学习
Keywords_English:device-free localization; radio tomographic imaging; wireless radio frequency identification devices; machine learning

PDF全文下载地址:http://xbzrb.tju.edu.cn/#/digest?ArticleID=5991