Research On Occupational Residence Acquisition Based On Mobile Phone Signaling Data

This paper studies the mobile phone signaling data.A density-based time series spatial data clustering method and LCSS are proposed after the research of the positioning principle of the cellular base station,the clustering algorithm of the resident point,and the extraction algorithm of the commute route.The algorithm of the commute route studies the problem of obtaining the occupational residence from the mobile phone signaling data,and extracts the commute route from it.The research shows that the method of occupational residence extraction proposed in this paper uses the time series information contained in the mobile phone signaling data to more accurately identify the place of residence and the commuting route between them.The proposed clustering algorithm and parameter calibration algorithm can identify the occupational residence distribution of different motion characteristics,and have better recognition effect.This paper provides a valuable exploration for the research of mobile phone signaling data in the geographical background and the analysis of individual residences in the city.Firstly,this paper analyzes the advantages and disadvantages of different positioning methods and the sources of error in mobile phone signaling data through the research on the positioning method of cellular base stations.On this basis,the preprocessing of mobile phone signaling data is deeply analyzed,and the processing methods and processing requirements of different error data are given.After analyzing the mobile phone signaling data of individual users,the resident points in the mobile phone signaling data are analyzed from the perspective of clustering.DBSCAN is a density-based data clustering algorithm that automatically determines the number of clusters and can be applied to the resident point analysis of mobile signaling data.Based on DBSCAN,this paper introduces time dimension parameters and proposes a DBSCAN-T algorithm for time series data.The adaptive method is used to automatically calibrate the algorithm parameters and support the cluster analysis of time series data.Finally,based on the analysis results of the resident clusters,according to the attribute law of the occupational residence,from the perspective of time,the KMeans clustering method is used to identify the position of the residence.After analyzing the distribution of the occupational residence,the users who meet the commuting rules are calculated by the LCSS algorithm to obtain the similarity of the route and obtain the commute route of the user.The comparison between the experimental results and the real data shows that the DBSCAN-T algorithm can well identify the place of residence,and the error is within 100 meters.The LCSS algorithm accurately extracts the common routes of users based on the analysis of the on-the-job residence.