Research On Visualization Techniques Of Spatio-Temporal Data In Wireless Roaming

With the increase of wireless network scale,users use more wireless networks than traditional networks.When users roaming in the wireless network,data exchange between mobile devices and AP.These data have temporal and spatial characteristics.Researchers have paid more attention to the advantages of visualization in dimensionality reduction and anomaly detection when dealing with spatio-temporal data.Therefore,using visualization to mine user behavior has great value and significance for monitoring data changes,discovering data anomalies and mining user relationships.Through the in-depth study of visualization theory and technology the better application of visual analysis method,this paper proposes a visual analysis model of space-time data in wireless roaming scenarios.The model includes data acquisition and storage module,data processing and analysis module,data structure and visualization module.According to different data analysis methods,the visualization module is divided into direct visualization module,feature visualization module and similar user portrait module.Through the research of user social relationship mining,users with similar trajectories are more likely to have intimate social relationships.In traditional methods,trajectory similarity algorithms,such as LCSS,only consider the relative position in time sequence,ignoring the continuity of trajectory.In the data processing and analysis module,a trajectory similarity calculation method based on spatio-temporal data is proposed.The trajectory distances under the spatio-time characteristics are calculated by segments.Then sum the similarity of the segments,and get the whole trajectory distance as the similarity.In addition,Users with high similarity are clustered by clustering method to find the user groups that may have social relations.This paper validates the scheme by collecting,processing,analyzing and visualizing the wireless roaming data of Tianjin University of Technology campus network.The experimental results show that the method is effective in finding users with social relations.The visual analysis method can monitor the trajectory and feature changes of users,portray similar users and depict users' behavior habits.