Research On Task Offloading Strategy In Mobile Edge Computing

Today,the rapid development of the Internet of Things technology has resulted in a massive number of interconnected devices that generate vast amounts of data.However,in the traditional cloud computing framework,data and applications need to be transmitted over the internet to cloud data centers for processing and storage,which can lead to problems such as high latency,high energy consumption,and privacy breaches.In contrast,mobile edge computing can leverage edge nodes located close to the end devices,such as mobile base stations,smart routers,and edge servers,to extend computing resources and services to the network edge,thus enabling more efficient data processing and application execution.Therefore,based on these advantages,computation offloading technology,as one of the core technologies of MEC(Mobile Edge Computing),has attracted widespread attention from both academia and industry.However,as mobile applications become increasingly complex,the computational tasks that end devices need to perform continue to increase,and the size of the computational tasks also becomes increasingly large.In this case,end devices with limited computing power cannot handle large-scale,high-load computational tasks.Therefore,optimizing and improving the computation offloading strategies for different application environments has become a key factor in enhancing user experience.Based on the above reasons,this paper aims to study the computation offloading strategies in the MEC under the scenario of multiple users and a single server,and the main innovative work is as follows:(1)A computation offloading strategy based on the improved whale optimization algorithm for MEC is proposed.This strategy further considers factors such as mobile device transmit power,edge server core allocation,and device energy consumption to build a system-wide profit maximization computation offloading model based on the traditional computation offloading model in the cloud computing framework.Furthermore,an improved whale optimization algorithm is designed to optimize the computation offloading strategy for the new MEC computation offloading model by integrating multiple strategies such as Circle chaotic mapping,adaptive threshold,adaptive weight,and Levy flight.Simulation results show that under the system profit model designed in this paper,the improved whale optimization algorithm can significantly improve system profit and achieve lower time cost.(2)Research is conducted on user collaboration in the MEC scenario,and a single-user collaboration computation offloading strategy based on the improved seagull algorithm is proposed.This strategy considers problems such as poor channel quality of mobile devices far away from the base station and insufficient use of computation offloading advantages,and designs an auxiliary computation offloading model for a single computing busy device,a single computing idle device,and an MEC server.The strategy uses idle devices close to the base station with better channel quality for auxiliary computation and offloading while minimizing energy consumption under the time delay constraints.Furthermore,an improved seagull algorithm based on a random step length is proposed to quickly solve the offloading direction for each computational task.Simulation results show that the improved algorithm has better performance than the original algorithm and can obtain better offloading strategies to effectively reduce system energy consumption.(3)The above collaboration computation model is further extended to the scenario of multiple users,and a multi-user joint collaboration computation offloading strategy based on orthogonal frequency division multiple access is proposed.This strategy integrates all mobile devices within the base station range and considers each mobile device as a potential edge server to perform auxiliary computation for other devices.At the same time,to address the problem of the impact of a large number of task transmissions on device energy consumption in the above scenario,this paper considers time delay constraints for each task and solves the slow convergence problem of the traditional artificial gorilla troop algorithm by proposing an improved algorithm.The simulation results show that the proposed algorithm can effectively reduce the energy consumption of the system while achieving better computation offloading strategies.