Research On Vehicle-road Coordination Task Offloading Strategy For Vehicle Platoon

The vehicle platoon uses the Internet of Vehicles technology to form a relatively stable communication link among a group of vehicles,and to coordinate the control of the vehicles,maintaining a relatively close fixed distance to follow the movement,which can reduce resistance and avoid invalid acceleration,thereby maximizing road throughput.It has a positive impact on energy consumption,reducing energy consumption and improving driving safety and comfort.With the help of the relatively stable network topology among vehicles,the vehicle platoon centrally controls the distributed computing resources,which provides a promising paradigm for solving the limited computing power of a single vehicle and using on-board computing resources to perform computingintensive and delay-sensitive tasks.In order to effectively utilize the computing resources of vehicle platoon and meet the goal of low-latency vehicle tasks,this paper focuses on the task offloading problem in the vehicle platoon scenario,and proposes a task offloading algorithm based on the low-latency vehicle-road coordination in the vehicle platoon scenario and dynamic vehicle platoon scenario.Task offloading strategies for pricing incentives.This paper mainly includes the following two research contents:(1)The problem of low-latency task offloading strategy in the vehicle platoon scenario is studied.According to the characteristics that vehicle platoon can provide relatively stable network topology,a task offloading strategy is proposed with vehicle platoon member nodes,vehicle-connected platoon communication nodes and roadside units as service nodes.First,according to the characteristics of vehicle platoons that can communicate with each other and form a relatively stable network topology,vehicle nodes with idle computing resources within the communication range of vehicle platoons are included in the scope of task offloading service nodes,and the roadside units equipped with edge servers are also considered.As a node that can provide offloading services.Secondly,due to the relative motion between the vehicle platoon and other nodes,the task offloading delay is constrained by calculating the link expiration time,and the delay and energy consumption during the task offloading process are modeled and converted into Markov order.In the decision-making process,a task offloading algorithm for vehicleroad coordination is proposed.The simulation results show that this algorithm reduces the time delay by 27.9% compared with the task offloading algorithm only assisted by platoon.(2)The problem of task offloading incentive mechanism in vehicle platoon scenario is studied.Aiming at the differences in information such as remaining energy,link status,and available computing resources of service nodes that provide offloading services,an incentive mechanism for dynamic pricing is proposed,which can dynamically price the computing resources and energy provided according to the comprehensive situation of each service node.First,according to the principle of diminishing marginal utility,the amount of computing resources provided by the service nodes is adjusted on the basis of the basic unit price,and the initial pricing plan is determined according to the energy,computing resources,communication links and other conditions of the nodes.After the contract is concluded according to the contract theory,the vehicle platoon leader node transmits the comprehensive information of the node to the roadside unit for unloading decision.After the task offloading is completed,the final price is determined according to the completion of each node,and the compensation is paid.The simulation results show that the task offloading strategy reduces the delay by 10.7% compared with the greedy strategy.