江小平,刘 键,丁 昊.基于CSI离散指纹的轨迹跟踪方法[J].中南民族大学学报自然科学版,2018,(4):79-83
基于CSI离散指纹的轨迹跟踪方法
Trajectory Tracking Method Based on CSI Discrete Fingerprint
  
DOI:10.12130/znmdzk.20180416
中文关键词: 室内定位  轨迹跟踪  信道状态信息  离散指纹
英文关键词: indoor location  trajectory tracking  channel state information  discrete fingerprinting
基金项目:国家自然科学基金资助项目(61402544)
作者单位
江小平,刘 键,丁 昊 中南民族大学 电子信息工程学院武汉 430074 
摘要点击次数: 144
全文下载次数: 107
中文摘要:
      针对从Wi-Fi设备中获取的接收信号强度(RSSI)在定位过程中误差较大的问题,设计了一种基于信道状 态信息(CSI) 离散指纹的轨迹跟踪方法.首先经过汉佩尔滤波,巴特沃斯低通滤波去除了CSI异常值和噪声的影响, 然后利用线性回归算法去除相位的频移误差,继而采用协方差矩阵特征值法进行特征提取,最后通过基于CSI子 载波权重的主成分分析法进行了源定位和轨迹跟踪.实验结果表明: 离散指纹点定位的平均正确率在 90%以上, 所设计的算法能有效实现对室内人员的定位和轨迹跟踪.
英文摘要:
      Aim to the problem that the received signal strength (RSSI) obtained from the Wi-Fi device is obvious in the positioning process, a trajectory tracking method based on channel state information (CSI) discrete fingerprint is designed. First of all through the Hampel filter, Butterworth low-pass filter removes the influence of outliers and CSI noise, then the linear regression algorithm removes the phase shift error, next the covariance matrix eigenvalue method is used to extract features,finally the passive location and trajectory tracking are carried out by principal component analysis based on the weight of CSI subcarriers. The experimental results show that the average accuracy of discrete fingerprint location is more than 90%, and the designed algorithm can effectively locate and track the indoor personnel.
查看全文   查看/发表评论  下载PDF阅读器
关闭