Research On Spectrum Sharing Between Multiple Operators And Computation Offloading In Vehicular Networks

In recent years,Cellular Vehicle-to-Everything(C-V2X)technology has attracted much attention in academia,for it can be compatible with existing celluar networks.However,the rapid increase of vehicles and the data traffic of vehicular business makes the spectrum resources of the licensed band become scarce.In order to copy with the challenge,C-V2 X technology can share the Intelligence Transportation System(ITS)band with the Dedicated Short Range Communications(DSRC)technology to increase the available resources of the C-V2 X network.Moreover,emerging applications such as autonomous driving,Augmented Reality(AR)and Virtual Reality(VR)generally involve complex computation and have strict delay requirement,whereas vehicles usually have limited computing ability.Mobile Edge Computing(MEC)technology deploys computation resources to the edge of vehicular networks to meet the demand of vehicular computation-intensive applications.This thesis aims at the spectrum sharing between the C-V2 X and the DSRC networks,the computation offloading based on Vehicle-to-Vehicle(V2V)and Vehicle-to-Infrastructure(V2I)communication in the vehicular edge computing(VEC)system.Firstly,we study the spectrum sharing between multiple network operators in vehicular network.In order to estimate the time-varying load of C-V2 X and DSRC networks,we propose a Quality of Service(Qo S)assessment mechanism.In order to avoid the networks misreporting their bids to obtain revenue,we use the Vickrey-Clarke-Groves(VCG)reverse auction mechanism to force the networks bid truthfully and ensure vehicular users to make appropriate network access decisions.Moreover,to achieve better spectrum efficiency and Qo S provisioning for vehicular users,we propose a spectrum transaction algorithm based on Carrier Sensing and Adaptive Transmission(CSAT)mechanism,with which C-V2 X and DSRC networks can negotiate about spectrum sharing.Simulation results verify the effectiveness of the spectrum transaction algorithm and reveal the impact of vehicle density and transmission rate on the Qo S of vehicular users.Secondly,we study the reliability issue of the computation offloading based on V2 V communication in the VEC system.Compared with the traditional task replication method where each task has multiple replicas and each replica is offloaded to one vehicle.Here,a task vehicle allocates the computation tasks and communication resources to its neighboring service vehicles through the V2 V link,and avoids the degradation of computation efficiency.Specifically,we formulate a joint computation task and bandwidth allocation problem with the goal of minimizing the task completion delay under computation offloading reliability constraint.In order to efficiently solve this non-convex optimization problem,we propose a two-layer iterative algorithm based on the Penalty Method and the Concave-Convex Procedure(CCCP)method.Simulation results verify that the algorithm converges quickly and the task completion delay is only about 30% compared to that of the task replication method.Finally,we study the joint allocation of communication and computation resources in a distributed manner based on V2 I communication in the VEC system.The resource allocation problem is modeled as an auction process,and the utility functions of vehicles and the VEC server are formulated.Each vehicle independently determines the proportion of offloaded task,the amount of communication and computation resources based on its utility function.By analysing utility function of vehicles,we derive that the amount of communication and computation resources requested by each vehicle are monotonously non-decreasing with the resource price decreasing.The price selection algorithm of the VEC server based on Dutch auction is designed.Simulation results illustrate the effectiveness of the proposed algorithm.