Research On Edge Server Placement Strategy In Multi-access ME

The combination of multi-access edge computing（MEC）and 5^thgeneration mobile communication technology（5G）has attracted more and more attention from scholars,and MEC improves service quality by offloading user terminal tasks to nearby edge servers with computing and storage resources.The placement of edge servers is related to the delay of task offloading and system energy consumption,and the placement of edge servers is an important issue in the field of edge computing.Therefore,this paper studies the placement of edge servers in multi-access edge computing,and discusses how to reasonably deploy the number and location of edge servers from the perspective of multiple scenarios,including:First:This thesis conducts research on the placement of edge servers in 5G metropolitan networks,weighting the user average delay,edge server load balancing,and edge server average resource utilization as indicators to measure the overall system performance.A method for edge server placement based on firefly algorithm（FA-ESP）is proposed,which improves the traditional firefly algorithm by using K-Means algorithm to determine the initial position,and iterates with average user delay to investigate the rationality of the number and location of edge server deployment in the Shanghai base station dataset.Simulation results show that compared with mainstream placement methods,the proposed improved firefly algorithm significantly reduces the delay effect when placing 20 edge servers,with a reduction of 16.2%compared to the traditional K-Means algorithm,and the overall performance is superior to the compared algorithms.Secondly:This thesis focuses on the placement of edge servers in 5G and vehicular edge computing scenarios.Current deployment strategies for edge servers have not considered the situation where edge servers are fully loaded and users need to forward service requests to other edge servers.To fully utilize the idle resources of other servers and improve the resource utilization of edge servers,this paper proposes an edge server placement method based on ant colony optimization and simulated annealing（ACOSA-ESP）.The model allows for user tasks to be forwarded to multiple edge servers.The Canopy algorithm is used for coarse clustering to determine the number of edge servers to be placed.Then,the ant colony optimization algorithm is combined with the simulated annealing algorithm to explore the placement of edge servers in the Nanjing roadside unit dataset,with priority given to optimizing system energy consumption while also optimizing edge server load balancing and reducing vehicle terminal delay.The weighted combination of these three factors is used to construct a system cost model.Finally,the optimal number and placement of edge servers for the vehicular edge computing system is obtained.