Research On Collaborative Offloading Mechanism Of Computing Tasks In Internet Of Vehicles

With the increasing number of vehicles on the road and the development of Internet of Vehicles(Io V)technology,vehicles have become an important part of mobile devices connected to the Internet.Intelligent connected vehicles in the Internet of Vehicles can support various mobile applications,such as image-assisted navigation,vehicle augmented reality,and autonomous driving.Most of mobile applications have high requirements on the computational resources and processing latency.However,the computing capability of a single vehicle is limited,which is a huge challenge for vehicles with limited resources.To meet these challenges,mobile edge computing(MEC)is considered a promising method.MEC introduces edge servers into the mobile computing environment.Mobile devices can offload computing tasks to neighboring mobile users and resource-rich edge servers through wireless networks.In MEC,offloading resource-intensive mobile applications to edge servers can greatly improve the performance of mobile applications and reduce latency.To study the problem of collaborative computation task offloading in the Internet of Vehicles,in the context of Internet of Vehicles based on mobile edge computing,we propose a collaborative computation task offloading problem that combines the vehicle-to-roadside unit(V2R)offloading method and the vehicle-to-vehicle(V2V)offloading method.This problem is transformed into a constrained optimization problem to minimize the overall processing delay with the given wireless channel constraints and contact duration constraints.We proposed a cluster based V2 V offloading scheme and game-theoretical offloading algorithm applying to this problem.Extensive experiment results based on vehicular mobility trace data set show that our proposed scheme had a better performance in decreasing overall processing delay.This paper studies the computation task offloading in the Internet of Vehicles from the perspective of application-level indicators,considering that intelligent vehicle applications are sensitive to computing delays and is divided into high-priority applications and low-priority applications.At the same time,intelligent electric vehicles have gradually become the future development trend,but a key factor affecting its development is still the problem of battery life,that is,the energy storage energy of electric vehicles is limited.Therefore,this study aims to select the best computation task offloading strategy by minimizing system processing delay and saving energy,and maximize the number of applications that meet high priority,optimize computing efficiency and resource utilization.We use potential game to model the system,and propose a potential game based computation task offloading mechanism.Simulation experiments in different scenarios are provided and the simulation results are discussed to verify that the proposed scheme has better performance.