Research On Intelligent Fusion Of Bluetooth And PDR Based Smart Phone Indoor Positioning And Tracking Systems

With the rapid development of the Internet and the popularity of smartphone,the demand for location-based service(LBS)is increasing gradually.The service has been widely used in people's livelihood,business,military and etc.In particular,the LBS application of smartphone has developed very fast.The key to LBS lies in acquiring the real-time position of users.Cellphones can get outdoor location service by GPS,but the GPS doesn't work when the GPS signal cannot be received effectively because of the obstacles of buildings in indoor environment.Thus,the high-precise indoor location of smartphone has always been a tricky problem in the positioning field.Considering the high-precise indoor location problem,this paper focuses on studying two indoor location algorithms,which are based on the received signal strength(RSS)of Bluetooth and pedestrian dead reckoning(PDR).On that basis,the paper puts forward the high-precise indoor localization and tracking algorithm of smartphone,based on the blending of Bluetooth and PDR.The main contents of the paper are as follow:Firstly,a Bluetooth positioning algorithm based on the probabilistic polling is proposed.The algorithm utilizes the RSS of Bluetooth to determine the probability of the smartphone's distance away from the Bluetooth beacon.Several Bluetooth beacons help divide the positioning region into different polling areas.Then according to the probability,the polling is conducted and the region with the largest votes would be taken as the positioning result of the smartphone.Experiments have shown that the maximum probability polling method can reduce the effects of RSS's unsteadiness upon the positioning preciseness.Compared with the least square positioning algorithm,the average positioning error is reduced by 57.3%,when the noises standard deviation is 3;the average positioning error is reduced by 68.3%,when the noises standard deviation is 10.In addition,the paper also suggests utilizing the cooperative localization of multi-users to further improve the positioning preciseness.Secondly,a fusion algorithm for smart phone indoor positioning and tracking is proposed,which combines the PDR and the Bluetooth RSS probability voting algorithm intelligently.1)The initial position of a smart phone is localized with Bluetooth RSS.2)While the smart phone is moving,it is tracked with PDR;then,the region of the largest votes is searched around the location of the PDR.3)Location results of PDR and Bluetooth RSS are combined,and the parameters of PDR are modified.The experiment results in the real scene show that: Compared to the pure PDR and Bluetooth RSS probability voting algorithm,the positioning accuracy of the proposed algorithm improves by 39% and 59.1%,respectively;Compared to the WLAN fingerprint + PDR based fusion indoor localization algorithm,the proposed algorithm has a better positioning accuracy and does not need the offline workload of fingerprint algorithm.Finally,to test the feasibility and practicability of the proposed algorithm,the paper realizes the above-mentioned algorithm in smartphone and develops a set of complete indoor localization demonstration system.The system includes Bluetooth beacon node,Android client and positioning server.It applies the positioning algorithm with the above-mentioned blending of the Bluetooth and PDR to conduct real-time positioning for smartphones and provide positioning,navigation,information push and other functions for mobile users.The positioning demonstration system can meet the demands for indoor localization and location service with a preciseness of 1.81 meter through the field tests.