Collaborative Task Allocation Mechanism Of Multi-Level Edge Network Of Internet Of Vehicles For Smart Highways

As the future direction of road development,smart highways provide a solid foundation for the rapid development of a new generation of smart vehicles.With the introduction of road alert,road monitoring,assisted driving and other services as well as the rapid.growth of the number of vehicles,it is difficult to meet the growing demand for smart highway services by relying on centralized computing in the cloud alone.Mobile Edge Computing(MEC)technology can greatly reduce the network latency caused by transmission by handling most of the computing tasks at the edge side,while relieving the pressure of computing in the cloud.However,compared with the huge number of vehicles on the smart highway,the load of a single base station is still high,and it is difficult to meet the low latency requirements of vehicle-generated services within the coverage area.It is of great value to study the collaborative task allocation technology of edge network for smart highway.The current research on collaborative task allocation in edge networks focuses on the vehicular network scenario,mainly in terms of computation and caching of vehicle tasks.The optimization mainly focuses on task latency and energy consumption,but there is less relevant research on the impact of vehicle speed on task offloading.At the same time,the problem of load balancing between servers by frequent server switching of vehicles has yet to be solved.In order to meet the requirement of low latency of the task by vehicles under the massive task in the smart highway scenario,a large number of Road Side Units(RSUs)are deployed on both sides of the highway to join the edge side network to share part of the computation task,and a multilevel network is used to efficiently process the tasks of different sizes.A multi-level intelligent collaborative computing mechanism(MICCM)for the Internet of Vehicles for smart highways is proposed.Firstly,a multilevel intelligent collaborative computing model based on cloud,MEC server and RSU is established to distribute the tasks offloaded to the multilevel network.Secondly,with the objectives of task latency minimization and network load balancing,the multi-level intelligent collaborative computing algorithm based on DDQN is proposed to solve the task splitting and allocation problem.Simulation results show that MICCM can reduce the service delay and keep the network load balance while completing the service demand.For the problem of increasing delay caused by frequent server switching due to the high-speed movement of vehicles in the smart highway scenario,we propose collaborative edge tasking based on location prediction(CETLP)mechanism by combining the characteristics of vehicle movement and using location prediction to collaborate with vehicle tasks in the same direction from the horizontal level to realize the low delay demand of vehicle tasks and provide efficient and sustainable services for vehicles.Firstly,the delay and load balancing-oriented edge task collaboration model is established by combining the characteristics of vehicles moving at high speed in the smart highway scenario.Secondly,the DDQN-based edge task collaboration algorithm is proposed to collaborate the current task toward vehicle movement based on vehicle location prediction and task delay requirements to further reduce task transmission delay and make full use of edge-side network resources.The simulation results show that the CETLP mechanism can further reduce the task completion delay and is very suitable for the current scenario.