Research On System Resource Optimization Algorithm Based On Edge Computing

Multi-access Edge Computing(MEC)network provides a new networking mode with faster unloading and richer resources,which brings realizable solutions for fast processing of complex computing tasks.Unmanned Aerial Vehicle(UAV)assisted MEC network makes network deployment and access more flexible and covers a wider range.Under the scenarios of emergency communication and low-cost coverage,it is an important method to improve the effectiveness of MEC system by optimizing user matching and unloading methods.According to the MEC network structure of simultaneous wireless information and power transfer(SWIPT),this paper proposes a user matching algorithm suitable for multi-users and multi-network edge servers.Firstly,considering the difference of users’ needs and diversified energy supply,a system utility function based on computing resources,transmission resources and energy resources is established.Then,aiming at maximizing the utility of the system,the multi-round auction algorithm based on multidimensional knapsack theory is used to match the best users;Finally,simulation verifies the effectiveness and reliability of the proposed user matching algorithm.Aiming at UAV-assisted MEC network,this paper proposes a multi-level MEC network resource optimization algorithm based on energy level and comprehensive performance evaluation,considering the energy level state of UAV,the unloading cost of each network level and the reward function of successful unloading.Firstly,the UAV-assisted multi-level MEC network resource optimization model is established,and the calculation methods of energy consumption and delay corresponding to different unloading levels are given Then,aiming at maximizing the utility function of the system,Markov Decision Processes(MDP)algorithm is used to obtain the optimal unloading strategy.Simulation results show that the proposed algorithm can optimize the system resource allocation and effectively improve the system utility function and user experience of computing and unloading.Fog computing will further miniaturize and discretize edge servers,which can significantly reduce communication delay and improve network capacity.Under the condition that UAV and fog nodes can relay or unload data,considering the choice of communication channel quality,heterogeneous network access and multi-level unloading,this paper proposes a multi-level fog computing(MFC)network resource optimization algorithm based on UAV assistance.The resource optimization model is established,and the MDP algorithm is applied to solve the problem to obtain the optimal unloading strategy.Finally,the performance and applicable scenarios of MEC and MFC are compared and analyzed by simulation.