Research And Implementation Of Mobile Node Positioning Technology Based On Zigbee Network And RFID

Recently, as the Internet of Things and mobile communication technology grow rapidly, people’s demands for services based on location increase quickly. The GPS satelllite positioning technology is currently the most popular way to achieve such goal, however, this approach has obvious disadvantages when it is used indoor, which leads to a lot of studies considering close distance wireless positioning technology.Zig Bee and RFID positioning technologies, two main ways of close distance wireless positioning technology, become the most promising techniques of indoor positioning. But Zig Bee positioning technology can not achieve the goal of positioning the important information of objects, while RFID positioning technology relies on the Internet, and especially constrained by the positioning distance in the process of positioning the tags.Based on the analysis of the possibility of combining the Zig Bee and RFID technologies together, this paper proposes a new composite approach, mobile node positioning technology. The main work is as following:Firstly, combining with the characteristics of close indoor location circumstance, and basing on on the Zig Bee and RFID technology we analyze the mainstream measuring methods: TOA、TDOA 、AOA、RSSI and PDOA; and the basic positioning methods: trilateral positioning and triangulation positioning. Considering the positioning accuracy、cost and system complexity, We propose a composite positioning plan based on Zig Bee network and RFID technology,and then we design a new framework for this positioning system from hardware design and software.Secondly,to solve the problem that classical RSSI ranging model was seriously affected by the circumstance, this paper tries to fix it from three aspects: preprocessing RSSI values、parameter optimization and distance remedy.At the stage of preprocessing RSSI value, we exclude the large RSSI value errors by using the Gaussian function. At the stage of parameter optimization, we get the parameters A and n which are very close to the actual number. At the stage of distance remedy, we build a distance difference correction model tofurther improve the ranging accuracy.Thirdly, to make up the drawbacks of phase difference of arrival(PDOA) that randomly selects frequency in dual-frequency comparison ranging method leading to the low accuracy of measuring, this paper proposes a new method, the phase difference of arrival in frequency domain(FD-PDOA) dual-frequency comparison ranging method. Combining the ideas of least square method and linear regression, this method gives an overall analysis of the phase difference changes on the sample frequency domain, andthen builds a mathematic model to measure the distance.Then, this paper analyzes the principles and shortcomings of centroid locating algorithm based on trilateral measurement, studying the estimationof original coordinates and coordinates correction. At the stage of estimating original coordinates, we introduce an auxiliary beacon nodes to obtain higher accuracy of the initial coordinates. At the stage of coordinates correction, we establish a RMS error function to get the optimal coordinates by making use of the ideas of gradient descent method.Finally, we apply the improved distance measure and positioning algorithm tothe positioning modulein the actual environment.The experiment results show that in the zone of 4*4 m2, the average measurement error of passive tags positioning is about 0.61 m.Positioning accuracy reaches 85%,which is close to the mainstream LANDMARC ocalization algorithm,but this method has lower cost and lower system complexity, which can meet the basic demand in reality.