Research On Task Offloading Strategy Based On Vehicle Edge Computing

With the rapid development of Internet of Things(Io T)and Edge Computing(EC)technologies,Vehicle Edge Computing(VEC)has gradually become a research hotspot.VEC utilizes various communication,computing,and caching resources and migrates computations closer to the vehicle using Edge Computing,through Road Side Units(RSUs),in order to efficiently offload vehicle tasks.However,most vehicle applications have characteristics that are computationally intensive and timesensitive,which presents significant challenges for task offloading scheduling.Additionally,terminal vehicles are also constrained by their own computing capabilities and communication resources.Therefore,researching effective VEC task offloading strategies is crucial.This article studies task offloading strategies under the background of VEC,which mainly includes the following aspects:Firstly,this paper considers the multi-destination joint offloading of tasks in the context of the Internet of Vehicles(Io V),combining the V2I(Vehicle-to-Infrastructure)communication mode between vehicles and roadside units.The tasks are assigned to local devices,edge servers,and base stations to achieve task offloading computation.By satisfying the latency constraints of each vehicle’s task and optimizing the total time and energy consumption required to process all vehicle tasks,this paper formulates a multi-objective optimization problem.A multi-objective task offloading algorithm(MOTOA)is proposed based on a genetic algorithm for solving the problem.The simulation experiment results demonstrate that compared to single-destination offloading(considering only time or energy offloading schemes)and other multi-objective task offloading algorithms,MOTOA achieves a reduction of approximately 32% in latency and saves about 28% in energy consumption.Furthermore,while task offloading can bring benefits to users,it also incurs task offloading costs,which is one of the primary concerns for service purchasers.This paper addresses the issue of task offloading costs by employing Lyapunov optimization methods to ensure the stability of the task queue.In the final solving process,the task offloading costs problem are transformed into a Traveling Salesman Problem(TSP),and a Minimizing Cost Task Offloading Algorithm(MCTOA)is proposed based on the simulated annealing algorithm for solving it.Experimental results show that compared to other offloading algorithms,MCTOA can reduce task offloading costs by 25% while increasing system throughput by nearly 20%.This indicates that the algorithm effectively guarantees the stability of the task queue and minimizes the cost of task offloading in the context of task offloading costs in the Internet of Vehicles(Io V).