Research On File Dissemination And Resource Scheduling Scheme In Vehicular Cloud Networks

With the accelerating urbanization process,increasingly prominent traffic problems such as congestion and security risks have caused a drastic decline in public travel experience.Along with the rapid rise of cloud computing information service mode,vehicular cloud computing has become the new transport solution meeting public expectation.Any application in vehicular cloud network is inseparable from the interconnection and efficient transmission of files.Owing to a series of technical limitations,network features and strict requirements of users in the system,efficient files transmission scheme has been challenged and attracted much attention.In view of the above issue,this thesis mainly researches from the aspects of vehicle-side access control and cloud-side resource scheduling.On the one hand,the high mobility of vehicles shortens the lifetime of channels and deteriorates network stability,and the resulting file fragmentation wastes the bandwidth and reduces transmission efficiency.The vehicular cloud system should perform task scheduling properly to optimize the download performance based on the guarantee of file validity and user satisfaction.On the other hand,the cloud platform provides various vehicular application services of different types and urgent degrees for different users,while the instantaneity,dynamic nature,and mission diversity of VANETs ask for higher requirements of resource allocation in vehicular clouds.The cloud platform should allocate resources efficiently to achieve the optimization of system objectives where the quality of resource allocation scheme directly affects the operation efficiency of the entire vehicular cloud system.In view of the task scheduling of vehicular cloud in traffic scenarios,in consideration of insufficient resources and privacy risks of Vehicle-to-Vehicle(V2V)communication,we employed the basic unit Drive-thru scenario as an example to achieve efficient download of large files(Admission Access with Cloud Compression based on data Integrity,AACCP)utilizing admission control of vehicles in the vehicular cloud system.This scheme employs priority considering file importance,data size,and deadline,trying to complete effective files download as much as possible in the prerequisite of file integrity and user tolerance through vehicle access control and task regulation.The theoretical analysis and simulation verification of this mechanism show its superiority over the classical access control mechanisms in terms of its availability and throughput.In response to the cloud resource scheduling problem in vehicular cloud systems,we designed the vehicular cloud resource scheduling mechanism(Maximum Reward based on Priority Resource Scheduling,MRPRS)focusing on the maximum return of the system.This mechanism integrates the priorities and economic factors including the demand degree,request data volume,and user tolerance,thus building a practical system model that considers both channel fading and vehicle collaborative download.We used an improved priority-based dynamic greedy algorithm to approximate the optimal solution for this optimization goal.Simulation verification shows the effectiveness of the mechanism,and reveals the influencing factors of the maximum return of the system considering the priority,which contributes to the parameter adjustment and resource allocation scheme in the resource scheduling of vehicular cloud platforms.The mechanisms proposed in this thesis improve the transmission performance of vehicular cloud systems in terms of access control and resource scheduling respectively.Meanwhile,the research content of this thesis can be further perfected.The system model of the AACCP file distribution mechanism should consider more practical application scenarios.Considering the load-balancing objective in the MRPRS resource scheduling mechanism to perform multi-objective optimization scheduling is a significant research work.