Research On Edge Server Deployment Method In Mobile Edge Computing

With the rapid development of networks,concepts such as smart cities and metaverse,and the increasing variety and number of mobile terminals,huge amounts of data are expected to be generated in the future,making it difficult for traditional cloud computing to meet the low latency requirements of mobile terminals,while the limited computing power of mobile terminals makes it difficult to process huge amounts of data.Mobile edge computing is seen as an important technology to solve these problems.Offloading computing tasks from the mobile device to an edge server closer to the mobile device can effectively reduce network latency and task processing time.However,the proper deployment of edge servers is a challenging issue.There may be a large number of base stations in an area through which mobile users communicate with edge servers.An inefficient deployment strategy will result in unbalanced load between edge servers,high deployment costs and long access latencies.For this reason,this paper investigates edge server deployment strategies.First,we study a robust deployment model for capped edge servers that minimises the weighted sum of network transmission latency and deployment cost.The model enables edge servers to perform well when faced with computing requests that exceed some of the predicted computing power,while allowing edge server operators to choose deployment strategies based on cost budgets or minimum system uptime.The problem is solved using improved simulated annealing and simulated using Shanghai Telecom Base Station Dataset.The experiments show that the improved simulated annealing algorithm improves performance by 9.7% over the control algorithm in terms of cost and latency weighted sum metrics.This paper then investigates the multi-objective optimisation of an edge server deployment model with no upper arithmetic limit,with the optimisation objective of balancing the workload between edge servers,as well as minimising the sum of network transmission delay and deployment cost weighting.This paper proposes a Harris Hawk discrete algorithm based on the Harris Hawk algorithm to solve this problem.Simulation results show that the Harris Hawk discrete algorithm improves the multi-objective optimisation metric by 3.9% and the load balancing metric by 54.15%.