Research On Key Technologies Of Data Dissemination In Software Defined Vehicular Network

Recently,with the economic and social development and the continuous expansion of the city scale,the number of cars increases significantly,resulting in many urban traffic problems,such as traffic congestion,frequent traffic accidents and environmental pollution,which have become a persistent problem that has plagued urban development and people's lives.With the rapid development and integration of modern information and communication technologies,automobile manufacturing industry intelligent transportation systems(ITSs)are expected to solve these sever urban traffic problems.Hence,as the basic and key technology to support ITSs,vehicular ad hoc networks(VANETs)have been concerned by both academia and industry.Efficient data dissemination in VANETs is the fundamental and critical issue to enable all kinds of ITSs.However,due to the inherent characteristics of VANETs and the diverse requirements of ITS applications,it is challenging to enable efficient data dissemination in VANETs.First,due to the high mobility of vehicles and the limited bandwidth of infrastructure in VANETs,the data dissemination normally cannot be completed in a single infrastructure's coverage.Hence,it is urgent to coordinate multiple infrastructures to make better use of bandwidth resources of infrastructures.Furthermore,due to requirements of the emerging ITS applications,such as mass data,low latency and high reliability,it is urgent to coordinate the communication resources of network nodes,especially edge nodes,to enable efficient data dissemination.In addition,the real-time services need to complete the data transmission within given time constraints.In highly heterogeneous vehicular communication environments,the heterogeneities,such as different data size and different data transmission rate,may greatly affect the performance of real-time service.Consequently,it is urgent to dynamically allocate communication resources to enable efficient data dissemination.In view of the above problems,this thesis investigates data dissemination in VANETs in terms of service architecture,model construction,algorithm design and performance evaluation.The main contributions are presented as follows:(1)Study on data dissemination based on network coding and vehicular caching in VANETs.First,by integrating software defined network(SDN)into vehicular network,a software defined vehicular network(SDVN)based service architecture is proposed,where the controller predicts vehicle trajectories based on the vehicle status and coordinates multiple roadside units for data services.Furthermore,based on network coding and vehicular caching,a cooperative coding and caching scheduling problem is formulated.In addition,a binary particle swarm optimization-based coding scheduling algorithm is proposed.According to the features of data dissemination problem,we design a series of rules for searching optimal solutions,such as particle representation,fitness function.Finally,experimental results demonstrate the architecture and algorithm can significantly improve the performance of data dissemination and breakthrough the bandwidth bottleneck of conventional infrastructure-to-vehicle(I2V)based data services.(2)Research on data dissemination based on fog computing in VANETs.First,fog computing is integrated into SDVN architecture,in which the edge nodes of the network are regarded as fog nodes take the advantages of fog computing in terms of low delay and high reliability.Then,a fog computing assisted service architecture in SDVN is proposed to improve the bandwidth efficiency and system performance of data dissemination.Furthermore,a fog-assisted cooperative service problem is formulated,and proves that it is NP-hard.In addition,a clique searching based scheduling algorithm is proposed to make broadcast decisions for cloud nodes and coordinate data scheduling strategies for fog nodes and vehicles.Time complexity of the algorithm is analyzed,which shows feasibility of the algorithm.Finally,the simulation model is built and a comprehensive simulation results demonstrates the superiority of the proposed solution.(3)Research on data dissemination in heterogeneous VANETs.First,based on service architecture integrated SDVN and fog computing,the heterogeneous vehicular environments are further considered,and a service architecture is proposed to improve the performance of real-time data service.Furthermore,for different data sizes and different data transmission rates,a heterogeneous data services problem is formulated,and the NP-hardness of this problem is proved.In addition,a coding-based heterogeneous data scheduling algorithm is proposed,which can dynamically allocate communication resources for data dissemination.Finally,the simulation model is built and the simulation resultst demonstrate the effectiveness of the system architecture and the superiority of the proposed algorithm.