Research On Multi-access Edge Computing Offloading Strategys For Tactical Smart Terminal Tasks

Most of the current researches on offloading strategies for Multi-Access Edge Computing（MEC）are aimed at civilian smart terminal tasks in a city-friendly environment,and lacks attention to tactical smart terminal（TST）tasks.The battlefield environment has challenges such as shortage of communication resources and continuous enemy interference,and military users have more stringent requirements for mission execution than ordinary users,such as lower time delay and higher security.The focus of this thesis is to supplement the lack of research on MEC offloading decisions for TST tasks by studying the salient characteristics of TST and its tasks,and to maximize the satisfaction of military users by formulating appropriate offloading strategies.In this thesis,terminal tasks are divided into small and medium-sized tasks and largescale tasks according to their data scale.Small and medium-sized tasks have two types of task sources: strong mobility terminals and weak mobility terminals.Among them,strong mobility terminals are prone to cell handover and service migration overhead,the weak mobility terminal is easy to be eavesdropped by the enemy,resulting in the leakage of mission information;At the same time,the current offloading strategy lacks adaptation and transformation for the characteristics of large-scale missions,and cannot be directly applied to this research.Based on the above problems,the main research work of this thesis is summarized as follows:First,a strong mobility terminal offloading strategy considering terminal mobility is studied.In this thesis,a probability-based mobility prediction model is designed.Combined with the service migration cost and other models,the problem of maximizing system offloading revenue is established,and an improved genetic algorithm is desiged to effectively solve the mixed integer nonlinear programming problem.The simulation results show that the offloading strategy proposed in this thesis can effectively reduce the unnecessary service migration overhead and improve the system offloading benefits from many aspects,compared with the uninstallation strategy that does not consider terminal mobility,the system uninstallation revenue increases by 123.24% on average.After that,the security offloading strategy of weak mobility terminal combined with Physical Layer Security（PLS）technology is studied.In this thesis,an eavesdropping model based on multi-association eavesdroppers is summarized,combined with the PLS technology,the problem of minimizing system security benefits is combined,and the constrained random successive approximation method is used to solve the non-convex problem effectively iteratively.The simulation results show that the eavesdropping model proposed in this thesis is more realistic than other models,compared with the single factor optimization strategy,the safe offloading strategy of jointly optimizing the delay and energy consumption reduces the system offloading overhead by 8.6% on average.Finally,an edge-cloud collaborative offloading strategy based on the characteristics of large-scale tasks is studied.In this thesis,a terminal energy-saving waiting model is proposed according to the energy consumption characteristics of different modes of the terminal,and the problem of minimizing system overhead is combined combined with other edge-cloud collaborative models,and then an improved particle swarm algorithm is desiged to solve the problem effectively.The simulation results show that the edgecloud collaborative offloading strategy proposed in this thesis is more suitable for largescale tasks than other comparative strategies,without considering the offloading strategy of the cloud computing center,the strategy proposed in this paper can reduce the system offloading overhead by 5.4% on average,and adding the terminal energy-saving waiting model can further reduce unnecessary offloading and cloud migration overhead.