Design And Implementation Of RSSI-Based Wireless Indoor Localization System

With the rapid development of wireless communication technology and increased demand of location based services in recent years, wireless positioning technology has been applied in more extensive areas. Positioning technology of GPS and base station are widely applied to outdoor positionin. However, these technologies couldn’t satisfy the needs of indoor positioning. Meanwhile, wireless local area networks attract more and more attention due to its extensive coverage, universal access standard and stable signal strength. Nevertheless, how to increase the precision, how to improve real timing and how to optimize energy-efficiency of the positioning system are urgent problems to be solved.In order to enhance the precision and efficiency of indoor positioning algorithms, the thesis make a deep study on indoor signal environments and fingerprint positioning algorithms, then design an indoor positioning system. The major contributions of the dissertation are as follows:(1) Based on wireless indoor positioning theory, the research on RSSI fingerprint of WiFi indoor positioning algorithms is carried out in this thesis from respective of offline phase and online phase. Besides, this thesis analyzes the elements that affect the precision of RSS database, such as distribution of RSS characteristics, involution of AP, noise reduction and cluster algorithms and also compares the online matching algorithms.(2) A RSSI adaptive clustering algorithm of matching the EM-based subordination degree vector is proposed in this thesis, which is a model based iteration algorithm. To combat the drawbacks of traditional indoor positioning sysem, the algorithm preprocesses the RSSI fingerprints and uses an adaptive algorithm to choose the fingerprint cluster core for the offline phase. In addition, EM algorithm is utilized to calculate the subordination degree vector for further RSSI fingerprint feature extraction. Experiments verify the presented RSSI fingerprint clustering algorithm could improve the accuracy and decrease the error of positioning.(3) A wireless indoor positioning system based on C/S structure is designed and implemented. The WiFi fingerprint data is stored and processed in PHP cloud server with offline and online clients for different usage. The improved RSSI databse foundation algorithm and proposed matching algorithm are employed to increase the indoor positioning precision. The experiments show that the mean error of the positioning system is 1.4m and the average response time is 155.2ms.