Mobile User Trejectory Analysis Based On Wireless Detection

With the rapid development of wireless access technology, such as Global Position System (GPS), WiFi, and mobile cellular networks, which provides much convenience for us to collect a large number of user location information. The development of wireless access technology not only brought convenience to people, but also provide the data base for us to research and analyze the mobile user trajectory.The mobile user trajectory analysis has several potential applications such as recommendation service, network optimization, social networking,and so on, for example, getting the current location of an individual or predicting the next location of an individual can provide the recommendations of the corresponding surrounding restaurants,supermarkets or gas stations; in the wireless network, mobile user trajectory analysis can be used to optimize the corresponding network protocol, improve the system performance of wireless network, and improve user experience; mobile user trajectory analysis can reflect the user activity to create a better social network.In this paper, wireless access technology is used to build the wireless data analysis system called WiCloud analysis system, mainly using WiFi technology to collect the mobile user spatio-temporal trajectories. When the WiCloud system collects the user’s trajectory information, the user does not know that his movements are recorded, so these trajectory information can reflect the individual’s actual activity regularity. About WiCloud analysis system and mobile user spatio-temporal trajectory, in this paper the main work is as follow:(1) Make the main research points of mobile user trajectory, and comparison among GPS, mobile cellular networks and WiFi technology to access user information. Make introduce of existing research on mobile user trajectory analysis, and make analysis of common four methods in the mobile user trajectory data dining.(2) Make summary of engineering basic of this research, the WiCloud analysis system, which uses WiFi detector to collect student’s spatio-temporal trajectory data. The WiFi access points are deployed at the network edge and collects user information emitted from all the mobile users in real time. WiCloud analysis system used MVC design pattern and can provide the smart store, smart campus, attendance system and other applications.(3) Based on the spatio-temporal trajectory information that WiCloud analysis system collect, the paper researches the similarity of the mobile user trajectory. Based on the similarity index and decision tree model, the paper presents student relationship measure model to measure the relationship between two users in wireless network. Meanwhile, there is a strong tie with student relationship measure and the student’s sex, so the feature set adds the similarity indicator whether same of the two students’sex. Finally, the paper evaluates the efficiency of the student relationship measure model based on the real spatio-temporal trajectory that the WiCloud analysis collected, and the experiment results approve that student relationship measure model is feasible for measuring user’s relationship.(4) Based on the spatio-temporal trajectory information that WiCloud analysis system collect and Markov chain model, the paper researches the mobile user trajectory prediction. Since the activity of a student in campus is closely related to the time at which the activity occurs, the paper presents an algorithm of dividing the Markov chain according to the notion of the time which coined as the Trajectory Prediction Algorithm (TPA). The WiCloud analysis system is used to evaluate the efficiency of our prediction algorithm, and the experimental results show that the TPA has increased the accuracy of prediction for over 30 % than the original Markov chain.