Research On Wireless Traffic Management Based On The Characteristics Of Mobile Internet Services

With the rapid spread of mobile smart terminals,mobile internet services like WhatsApp,WeChat and Line have gained considerable momentum in recent years.These services differ from traditional telecom services,such as voice and short messages,in that they provide an always-on connectivity.In addition,they are developed in accordance with the “client-server” model,which regards the mobile communication network as a single pipeline to carry the service information.Unfortunately,this design paradigm becomes to the uncoordinated initiator between emerging mobile internet services and the network transmission mechanics in current mobile networks,leading great challenges to mobile networks and user equipments(UEs).To deal with these challenges,it is of great theoretical and practical value to explore the transmission characteristics of emerging services,and then to quantify its impacts on mobile networks,and finally to design dynamically adaptable network transmission mechanism.Based on the deep study of “always-on” mobile internet services,this thesis firstly described the interaction process between UE and mobile networks,and then innovatively modelled the traffic of mobile internet services with the parameters of “always-online” needed keep-alive messages(KAMs)and user-triggered messages(UTMs).In order to verify the accuracy and generality of the proposed traffic model,various models of UTM proposed in related works were considered.The simulation results showed that the proposed traffic model could well describe the mobile internet services when UTM follows any kind of distribution.On this basis,this thesis firstly mapped the service traffic with the radio resources in Long Term Evolution network,and then described the process of radio resource control(RRC)state transition triggered by mobile internet services,and finally modelled the network signaling load and UE energy consumption with KAM period,UTM mean interval and RRC release timer.The simulation results showed that only longer KAM period can be used to reduce network signaling load and UE energy consumption simultaneously.Based on the above research results,this thesis finally proposed a wireless traffic management method based on traffic shaping of KAM in mobile internet services.Three modes of buffered KAMs to be centrally transmitted were proposed,that is timeout of shaping process,overload of UE buffer and UTM to be transmitted.The probability of the three modes and the network signaling load,UE energy consumption and shaping delay under different modes were modelled with stochastic processes theory,which finally formulated the average signaling load and energy consumption per unit time of a UE as well as the average shaping delay per KAM.The simulation results showed that no shaping delay of UTM were incurred.Moreover,while the average shaping delay per KAM was constraint to one minute,the average energy consumption per unit time decreased significantly,whatever the RRC release timer is.Besides,the average signaling load per unit time decreased simultaneously when RRC release timer was set within sixty seconds.This showed signaling load and energy consumption can be decreased simultaneously with the proposed method,thus improving user experience and the efficiency of control plane signaling resource at the same time.