Research Of Indoor Positioning Technologies Based On BLE

With the accelerated pace of urbanization,shopping malls,subways,parking lots and other indoor space have become increasingly big and complex.The era of internet of things is coming,including smart city and smart home.The above all factors prompt the requirements of services based on indoor location to constantly advance.Currently,Bluetooth Low Energy(BLE)exhibits excellent performance such as low cost,low power consumption,wide coverage,high speed connection and good stability,standing out the numerous indoor positioning technologies.Also,it has been greatly promoted with the popularity of smart phones.Therefore,the research of indoor positioning technology based on BLE has attracted more and more attention.Due to the limitation of power consumption and communication distance,the traditional Bluetooth localization is often based on approximation or propagation model with low precision for the indoor positioning.Yet,the new features of BLE make it possible to use the location fingerprint method which is simple and easy to locate.So,based on BLE technology,a method of using its signal strength fingerprint for indoor positioning was proposed in this paper,to reduce the off-line workload of fingerprint acquisition and improve the on-line matching efficiency and positioning accuracy.This paper is focused on the following researches:Firstly,the development of domestic and international indoor positioning technology,common technologies and algorithms were analyzed.Then,BLE technology and location fingerprinting were introduced.The characteristics of the BLE signal strength were analyzed by building an experimental platform.According to the actual test data,the applicability of the BLE signal strength as the feature for location fingerprint position was studied.Secondly,a method for constructing location fingerprint database based on improved Kriging interpolation was presented,in order to improve the construction efficiency of the off-line location fingerprint database.A small number of samples were used to investigate the experimental variation function of BLE signal in indoor environment.Then,fitting precision of theoretical variation function was improved by the simulated annealing algorithm.Last,the BLE signal strength of unmeasured points was estimated by the ordinary Kriging interpolation.The results show that the proposed method can reduce the workload of fingerprint data acquisition by 50%.Finally,in order to improve the online matching efficiency and positioning accuracy,an on-line locating and tracking algorithm based on fuzzy clustering and particle filter is proposed.The off-line location fingerprint database was clustered by the Fuzzy C-Means(FCM)algorithm.Then the location fingerprint of locating point was classified,and the pattern matching is performed in the class.After obtaining the initial positioning results,the particle filter algorithm fusing the historical state information is used to further improve the positioning accuracy.The experimental results indicate that the algorithm can improve the matching efficiency of fingerprint library and the average positioning error is 1.84 m.