An Adaptive Fingerprint Indoor Localization Method Based On Bluetooth Technology

With the improvement of information level of life, our life style has changed a lot. There have been more and more indoor intelligent devices with the development of internet of things (IOT). Bluetooth low energy (BLE) which was invented in 2010 makes Bluetooth smart devices occupy a significant share of market rapidly. As outdoor positioning technology coming mature, people began to concentrate on indoor positioning. Besides being applied in navigation, we can also achieve more intelligent context awareness according to location information. In an environment with Bluetooth devices around, it is necessary to take Bluetooth localization as the future direction of research.This paper has analyzed and summarized the indoor localization methods, as well as the further exploration in the application of low energy Bluetooth technology in the field of indoor positioning. In view of the existing fingerprint indoor localization algorithm, combined with the characteristics of low energy Bluetooth technology, ABIL scheme was proposed, which can make fingerprint database adaptive to realistic environment, also improve the positioning precision and robustness of the system.The main contribution and innovation of this paper are as follows:(1) This paper has analyzed the indoor positioning method and chosen fingerprint method as the main research content to analyze its characteristics and point out its shortcomings.(2) This paper did further analysis and research on the low energy Bluetooth technology, and its application in the field of indoor positioning.(3) By the method of dynamic updating fingerprint database and integrating the concept of group entitativity and perception, let users participate in the maintenance process of fingerprint database and make the fingerprint database change with environment, which can greatly reduce the cost of database maintenance.(4) According to the characteristics of fingerprint database dynamic update, this paper uses bisecting K-Means clustering algorithm with dynamic K value to train on fingerprint database, and gets the dynamic clustering results. Online positioning stages presented a localization algorithm based on the clusters and incorporated multiple clusters, which also achieved the position estimation with nearest neighbor algorithm.(5) This paper achieves ABIL scheme on both iOS platform and cloud server, and establishes a system facility in the laboratory, which verifies the feasibility of solution and achieves higher positioning accuracy.