Solution To Typical Scenes Restoration Based On Smartphone Application Traffic

Nowadays,various types of intelligent terminal devices are springing up as a spinoff of its popularity,such as smart phone,smart tablet and so on.The way of using intelligent terminal devices embraced a sea of changes,the main one being the transfer from traditional phone call and SMS to smart phone-based applications,which lead to some special circumstances such as overloading of base stations becoming more frequent,especially the 3G and LTE network environment.Traditional overload scenes are usually triggered by certain conditions,such as disasters.But nowadays,the over-load has become a more common occurence.Hence,traditional modeling on analysis parameters cannot precisely model the overload scene caused by applications traffic in the 3G and LTE networks.Therefore,how to restore the application using typical scenarios,by analyzing parameters in the base station and application traffic based on smart phone,becomes incrementally critical.This paper focuses on how to restore some typical scenes and analyzing users'behavior on applications,based on application traffic on smart phone and analysis of parameters in base station,in the pursuit of disclosing the risks under some certain circumstances in advance.Here,we demonstrate the design and implementation of typical smart phone-based applications traffic capture tools,which enable the typical application traffic capture on Android platform.Then,the captured traffic shall be cate-gorized by the Naive Bayes algorithm featured with regional adaptability,in search for pure smart phone-based application traffic.Upon above,model on application traffics leveraging traffic modeling algorithm optimized by Markov model.Then the model dedicates in stimulation for application traffic.Using the generated traffics model and analysis parameters in the base station,we restored a typical scene and analyzed users' behavior and applications usage in the scene.We used the generated traffics to predict the analysis parameters and to minimize the difference between real traffic parameters which is obtained from the base station and predicted traffic parameter values.So,we converted the issue of restoration into an issue of optimization to develop the user number of each different application in the scene.At last,we verified the traffic modeling scheme and the scene simulation scheme through stimulation experiments.The experiments results show that the traffic modeling solution give display to the characteristics of the application traffic and restore the burstiness and relationship between the upload traffic and download traffic while using the application.The results of solution to scene restoration exhibit that the scheme is effective to reproduce applications using scenario and the error is within an acceptable range.