User Mobility Research And Interest Region Mining Based On Cellular Networks Traffic

With the rapid development of wireless communications technology,more and more people use mobile phone to internet.And with the rapid development of urbanization,the mobility range has become larger and larger,which take a greater challenge for city and road planning,base station building.In order to make city planning and building more reasonable,we should understand user mobility.However,there are too many people in city and their behavior is different from each other,which make researching independent user mobility meaningless.In reality,there are some certain mobile features during user group.So we should analyse user group and model their behavior,which can improve city building and guide the future.In this paper,based on DPI data of cellular networks,we research the mobility of user group.The main research contents and innovations are as fellows.First of all,the four typical data types in cellular networks and intrinsic value are given.Then big data storage technology HDFS,big data computation technology Spark and Web technology SSM architecture are introduced.And the big data platform for data analysis and algorithm implementation is built.Secondly,user group mobility behavior is analysed in detail,which include user mobile distance,user mobile range and mobile pattern.The difference between cities of different economic level is compared.The result show that,user mobile distance approximate conform power law distribution,area of trajectory convex hull approximate conform truncated power law distribution and the more advanced the economy is,the stronger the mobility of the users.The result of user frequent pattern show that user group have different mobile regular in different time during a day.Then,GMM model is applied to mine the function zone of city based on user mobile feature of base station,which represent the relation between user and function.The probability of base station area belong to each function is mined.Then the user mobile feature of each function zone is researched.The result show that there are four function area which have typical user mobile feature based this DPI data.Finally,the probability HITS model is proposed to mine the interest regions in city and use Spark to implement the model with distributed computing.The interest region of weekday,holiday and different time of the day is mined.The result show that the distribution of interest region is different from weekday,holiday and different time of a day.But the main region has little change and different interest region have different user mobile feature.This paper research user group mobile behavior from multiple dimensions based on DPI data.The result can provide guidance for urban planning,base station building and can improve the utilization efficiency of the cellular network resources.