Research And Implementation Of RFID Indoor Positioning System Based On Improved LANDMARC Algorithm

Abstract:With the rapid development of Internet of things, humans demand for location information becomes larger. Location-based services and systems have arose more and more people’s attention. Global Positioning System (GPS) has been widely applied to our lives. Accuracy of GPS positioning may have reached10meters assisted by the cellular network, which solves a lot of practical problems in the fields of our public life. However, this positioning method is limited to outdoor conditions. In recent years, studies of indoor positioning technology have been proposed based on infrared, wireless sensor, ultra-wideband, radio frequency technology RFID, especially RFID technology. RFID technology has the advantages of non-line, non-contact, high-precision, short latency, large transmission range and relatively low cost, which makes it become the main choice of indoor positioning technology. Enclosing RFID indoor positioning, the main contents are as follows:1. Analysis of the current situation of positioning technology and Improved positioning algorithmFirstly, this paper analyzed the advantages and disadvantages of the current RFID positioning algorithms and their scope of application. According to the practical situation of equipments and experimental area, LANDMARC algorithm was chosen as the core algorithm in the experiment. Aiming at the problems of LANDMARC algorithm such as layout, weights, K value and large calculation, this paper proposed an improved algorithm, which could divide experimental area into sub-regions with SVM and determine the value of K dynamically. Then, on the platform of Matlab, this paper evaluated the experiment results by means of estimation error, mean square error and error cumulative distribution function and validated the improvment of the improved algorithm on the aspects of accuracy, computational efficiency and equipment costing. At last, the most suitable reference tag density layout mode in experimental area was given by analyzing the positioning accuracy of different reference tag density.2. Positioning System Design and ImplementationThis paper determined the positioning system needs on the aspects of function, data and technology and completed the design of the system. Then RFID equipments were placed in the Innovation Center office of Jiashan County, Zhejiang Province, which were used to collect walking paths data through improved positioning algorithm. At last, this paper implemented an indoor roaming three-dimensional system with analysis and visualization of RFID spatio-temporal data stored in SQLite database by Unity3D.