Research On Object Position Sensing And Multi-tag Mutual Coupling Suppression In Battery-free RFID

Radio Frequency Identification(RFID)is originally designed for automatic object identification,which may substitute for commonly-used 1D and 2D barcodes in the future.In recent years,an RFID tag is starting to be regarded as a battery-free sensor by many researchers for object localization,orientation tracking,material recognition,human respiratory/heart rate detection,etc.The novel idea makes RFID's application fields significantly extended.This thesis performs a deep research on four issues in RFID: moving object tracking,stationary object localization,device-free object motion sensing and tag mutual coupling suppression.The main contributations of this thesis are as follows:(1)To improve noise tolerance in phase-difference based solutions for moving tag tracking,this thesis proposes a system Track T,which uses a novel localization feature,i.e.,phase periodicity difference between successive phase samples.The system can achieve fine-grained tracking accuracy for a tag moving along known or unknown trajectory.The phase periodicity difference is derived from phase difference,which is usually required to be eliminated as much as possible in previous work.In the proposed system,this ‘bad' variable is instead used to determine the tag position in a surveillance region.For the case where an RFID tag move along a known track with constant speed,Track T only needs to locate the tag's initial position.It relies on the proposed matching rule based on phase periodicity difference to signle out several possible position candidates from the region of interest.Since these candidates are relatively far from each other(more than half a wavelength),the phase difference containing phase periodicity difference can be used in theis step to find an optimal one from them.And for the case where the tag's moving track is unknown to the system,both its initial position and subsequent trajectory need to be estimated.Gven all possible positions in the region of interest as an initial guess,the system adopts a first-order Taylor series expansion to calculate the relative displacement of the tracked tag between successive sampling points.Once collecting enough samples,the system locates the initial position using the same method for the known trajectory tracking.In a scenario with a relatively low interference,the proposed system achieves millimeter-level tracking accuracy when deploying four antennas around the region.(2)To deal with the issues of acquiring a reader antenna's position at each sampling time and improving real-time performance in 3D localization for stationary RFID tags,this thesis present RFMVO that fuses RFID and computer vision for stationary RFID localization in 3D space by attaching a light-weight 2D monocular camera to reader antennas.Firstly,the existing monocular visual odometry only recovers a camera/antenna trajectory in the camera view from 2D images.By combining it with RF phase,RF-MVO uses a model to estimate a scale factor for real-world trajectory transformation,along with spatial directions of an RFID tag relative to a virtual antenna array due to the mobility of each antenna.Then RF-MVO can achieve a novel RFID localization algorithm that does not require exhaustively searching all possible positions within the pre-specified region.To speed up the searching process,this thesis proposes a coarse-to-fine optimization algorithm.Our experiments demonstrate the effectiveness of proposed algorithms and show RF-MVO can achieve 6.23 cm localization error when deploying two antennas.(3)To address the flash effect caused by an obstacle in non-line-of-sight scenarios and weaklytrusted signal features for device-free sensing,this thesis designs RF-HMS that tracks device-free human motion through walls using a tag planar array is grouped by many battery-free RFID tags.The system performs by using a differencing model to remove the flash effect,and then it extracts phase shifts from the model to detect the absence or presence of any moving persons and further identify his/her forward or backward motion direction.The results show that RF-HMS can effectively achieve the accuracy of 100% for moving person detection in a sensing region and more than 90% accuracy for human motion direction recognition.(4)To cope with the distortion in RFID signal fingerprint(i.e.,RSSI and phase)due to tag mutual coupling when multiple tags are simultaneously attached on an object for sensing,this thesis presents RF-Mirror that enables compensating the distortion in two-tag RFID systems based on prior knowledge.Firstly,the thesis models the backscatter signal of a responding tag in multi-tag scenarios,and then formulates novel RSSI-and RF phase-distance models with coupling terms.Secondly,this work designs an algorithm to characterize the coupling effect on tag gain by fusing RSSI and phase in the absence and presence of mutual coupling,without acquiring mutual impedance between tags.Thirdly,the proposed system RF-Mirror performs by using a decoupling algorithm based on an observation that tag mutual coupling is independent of the position of a tag array relative to a reader antenna.Our experiments show the effectiveness of our models and achieves the decoupling errors of 0.422 d B and 0.076 radians in calculating RSSI-and phase-difference.