Research On Trajectory Tracking Method Based On Long-distance RFID

As wireless communication technology and mobile terminals in people's daily lives highly popular, the various applications ased on location information have become one of the hot topics in the field of indoor localization. However, the GPS GPS positioning system that has outstanding performance of the outdoor localization was affected by the blocked of the wall, so it does not apply to the indoor environment. Researchers began to apply ultrasonic,infrared, RSSI technologies to the studey of indoor localization, but these methods either high cost or affected by multipath interference and accuracy is not satisfactory, practical applications has been limited. So this paper analyzed long-range RFID technology on the transmission level of wireless signal, extract the channel state information features and apply it to the research of the target intrusion detection, location identification and trajectory tracking problems. The main contributions of this paper are as follows:1. Based on the transmission principle of long-range radio frequency signal, this paper apply the channel state information(CSI) which is in the physical layer of wireless signal to target intrusion detection as a new attempt. Put forward the Time Series-based target intrusion Detection Method(TSDM). TSDM extract the features of channel state information and then projected the CSI of different moments onto time series.By this way, the target intrusion detection problem is converted to a time series anomaly detection problem. In the detection stage, the signals data is analysised and processed by an improved EMA method. Addition, the measures of confirms and alarms are designed in the warning stage.2. Through research and analysis of previous indoor positioning methods, this paper put forward a Clustering-based Fingerprint Localization Method(CFLM). CFLM use the channel state information as the fingerprint date in location identification, it has more robust features. Since the introduction of the idea of clustering algorithm effectively improve the speed and accuracy of fingerprint matching, due to the differentiation K-value strategies in the recognition stage,the CFLM obtain a more accurate positioning results.3. Based on the research of target intrusion detection and location identification, characteristics of the channel state information was further extraction and utilization; this paper put forward a Vector Projection-based trajectory Tracking Method(VPTM). A multi-antenna measurement techniques is used to extract the phase information of CSI in VPTM, and the velocity vector of different direction are calculated by carrier phase ranging principle. The trajectory of target is decomposed to different measurement direction in the coordinate system whose origin is the current location of target, and then the decomposition equation of speed is established according to the principle of vector projection principle. At last, the equation of trajectory moving is deduced,and the location of different time can calculate by use CFLM to calculate the initial pdsition.