Research On Collaborative Computing Task Offloading Strategy Based On Edge Computing

Edge computing,as a supplement to cloud computing,can minimize the delay and energy consumption in the process of computing offloading when processing computing tasks generated by devices in the Internet of things and 5G scenarios.However,because of the uneven distribution of computing tasks and the empty window period of task generation,the existing edge computing architecture has a series of problems.On the one hand,when facing massive and intensive tasks,the processing pressure of some edge clouds is too heavy.On the other hand,there are a large number of idle edge servers around the local edge offloading system,which are in an idle state so that their computing resources cannot be reused.Therefore,how to effectively coordinate between multiple edge clouds to achieve joint computing offload to minimize the delay and energy consumption of the offload system has become a hot research area today.In this thesis,we first study the complete computing offload scenario due to the lack of computing power of the terminal node.Achieving the relevance between multiple tasks and the synergy between computing nodes,we propose an optimized computing offload scheme.The CCOSGG,a collaborative offloading scheme targeted at energy consumption,uses task reconstruction and graph segmentation algorithms to map all tasks into unit computing clusters based on relevance,and then simplifies the three-tier collaborative computing offloading system into a two-tier offload structure.The Stackblerg game theory is used to conduct bilateral games on the user side and the server side.The results of simulation experiments show that the scheme can verify the validity of the results,and at the same time minimize the delay and energy consumption costs under different network scales.For some computing offload scenarios where the terminal node has computing capabilities,and also for the correlation and privacy between tasks,the problem of collaborative offloading in the three-tier offload structure,a multi-edge and cloud collaborative computing offloading based on genetic algorithm is proposed.Model(Genetic Algorithm-based Multi-edge Collaborative Computing Offloading Model,GAMCCOM),this computing offloading scheme combines local edge and remote edge to perform task offloading,and uses genetic algorithm to minimize the system cost considering time delay and energy consumption.Compared with the existing simulation experiments of unloading and unloading,under the objective of minimizing time delay and energy loss in some calculation unloading scenarios,the overall cost of this scheme is reduced by about 23%compared with the basic three-layer unloading scheme.Considering only the delay consumption and only considering the energy consumption,the system cost can still be reduced by about 17%and 15%,respectively.Therefore,the GAMCCOM offload method improves system performance for different offload targets of edge computing.Through theoretical analysis and experimental simulation,we verify that the proposed collaborative computing offloading scheme for the two scenarios can effectively reduce the time delay and energy cost of the system.