Mobile Cloud Service Access And Offloading Strategies Based On Game Theory

With the development of mobile Internet technology,resource management and optimization of mobile terminals such as computing,storage and network has become a new challenge.Mobile Cloud Computing(MCC)has received extensive attention as the current mainstream technology to address this challenge.However the mobile characteristic of mobile terminal makes the network access problem become the main factor restricting its performance improvement Therefore,the integration of the network handover prediction strategy into the computing task offloading technology is an effective means to realize the optimization of energy consumption and delay of the equipment in the mobile process.This thesis mainly studies the theory and optimization algorithm of energy consumption and delay cost when mobile devices in mobile cloud perform computing tasks.Major achievements include:1.The network handoff selection for the mobile terminal while moving causes the user to compete and thereby reduce the data transmission rate problem,this thesis introduces the principle of predictive handover,takes the remaining connection time and the received signal strength as metrics,and introduces non-cooperative game theory on the basis of the multi-criteria decision algorithm,and proposes the candidate network prediction access model.This model not only considers the attribute requirements of both wireless networks and users through multi-criteria decision algorithms,but also considers the user's personalized requirements and benefit optimization for wireless network selection through non-cooperative game models.The experiments show that this model not only effectively improves the accuracy of user's network selection,but also reduces the ping-pong effect of network connection caused by mobility and effectively improves the network selection effect of mobile users.2.To optimize the energy consumption and delay in the offloading process of mobile users,this thesis proposes a computational task offload strategy based on potential game theory.This strategy selects the potential game theory to model the offloading decision of the mobile devices and obtains the final offloading decision of each user through the Nash equilibrium point.The experimental results show that the strategy proposed in this thesis can meet the cost optimization of the whole system when the mobile user is offloaded,which is optimized by 10.3% compered with the traditional typical offloading scheme.