Task Scheduling And Incentive Mechanism In Vehicular Fog Computing

With the development of Internet of Vehicles technology,the types of vehicle applications become more and more diverse,resulting in an explosive growth in the number of service requests.However,the limited computing capacity and storage resources of the vehicle make it difficult to meet the resource requests for many computation-intensive tasks.For this reason,Vehicular Fog Computing(VFC)came into being,which was to dispatch the task to the fog node near the vehicle for processing.As the fog node was close to the task requesting vehicle,the distance of the task transmission was shortened and the task delay was effectively reduced.Task scheduling,as the primary link in VFC,is very important in reducing energy consumption and time.Therefore,it is very necessary to design a reasonable task scheduling scheme.Meanwhile,because existing studies assume that all vehicles are willing to play the role of fog nodes in the task scheduling process of VFC.Actually,vehicles are selfish,so it is also very necessary to design a reasonable incentive mechanism to encourage fog nodes to actively complete tasks.In particular,this thesis has done the following work for the above two problems:1.In order to reduce the time and energy consumption during the task execution,it is need to schedule some tasks to the fog node,and the fog node help to complete the tasks.Firstly,the communication and calculation models of the process are proposed,Meanwhile,the objective function is obtained as the fitness function.Then,this thesis proposes a task scheduling scheme using genetic algorithm,which solves two main problems: task execution position(local execution or fog node execution)and task execution order.The scheme constantly optimizes the individual through the process of coding,initialization,evaluation of fitness value,selection,crossover and mutation,so as to achieve the purpose of reducing energy consumption.Finally,the simulation experiment of the task scheduling scheme is carried out,and the experimental results show that the proposed task scheduling scheme can effectively reduce the time delay and energy consumption.2.In order to encourage the client vehicle to actively participate in the task scheduling and the fog node to actively complete the task,an incentive mechanism is designed in this thesis.Firstly,considering the devaluation of the task value and the failure of some fog servers to complete the task on time,the models of buyer,seller and auctioneer are established respectively,and the actual social welfare of the whole auction process is obtained.Then,an incentive mechanism using double auction is designed,which consists of three processes: winners selection,winners matching and expenses calculation,the winners matching process is divided into three steps: cutting,matching,temporary costs calculation,using the greedy algorithm and threshold payment way match winners and temporary costs calculation,meanwhile,the honesty of the incentive mechanism is analyzed.Finally,through simulation experiments,it is verified that the matching scheme using greedy algorithm has less running time.The incentive mechanism meets the requirements of individual rationality,budget balance and computational efficiency.Certainly,it can capture the potential task depreciation.