Research On Communication And Computing Resource Scheduling Mechanism For Security Service In Vehicular Network

With the advent of 5G era,the vehicle has become more and more intelligent and informatization,and the rapid growth of the number of vehicles has led to a large amount of data requiring communicational transmission and computational processing.With the continuous development of the intelligent process of VANET,the safety applications of roads have increasingly attracted the attention of the whole society.And the vehicle security service is constituted of traditional short message alarm information and emerging long message vehicle security applications.In the typical application scenario of the 5G system,Vehicular Ad-hoc Network(VANET)has important research significance.The future intelligent transportation system needs to meet the requirements of low delay and highly reliable broadcasting of warning information,and intelligent processing of computing data at least.The focus of thesis is how to schedule the computing resources and communicational resources of the vehicle to improve the warning performance of the security service in different VANET scenarios.Based on timeliness and reliability requirements of short-message warning information in different vehicle density,a communicational strategy based on vehicle clustering and power adjustment is designed.Firstly,this strategy considers that the clustering vehicle can effectively decrease the redundancy of warning information and improve the utilization of channel resources.At the same time,the algorithm analyzes the degree of channel congestion in different vehicle density scenarios.Under different channel busy degree,power adjustment of cluster vehicles can improve the performance of warning broadcasting.When the channel is relatively busy,the communication power of the vehicle is reduced to avoid channel congestion.When the channel is relatively idle,the communication power of the vehicle is raised to increase the coverage rate of the warning information.And increasing the communication power of the vehicles in the dangerous area to reduce the latency of broadcasting information.Aiming at the problem of communication under non-line of sight(NLOS)intersection scene,the thesis designs a communication strategy based on shadowing effect control.Firstly,this strategy considers the road area has different degrees of shadow fading by roadside buildings in the intersection scene.Based on the shadow detection results,the edge node divides road sections into several types of shadowing attenuation areas according to the shadow effect.The vehicle chooses a reasonable transmission mode in different shadow fading areas,namely vehicle to vehicle(V2V)and vehicle to infrastructure(V2I),this strategy can effectively improve the broadcasting performance.Finally,the analysis of warning performance is carried out for the emerging longmessage security applications,such as 4K/8K ultra clear traffic video,multi-view video splice processing of driverless car,AR-assisted applications related to security and so on.Based on the collaborative analysis of the computing resources of vehicles and edge nodes,the thesis designs a collaborative resource scheduling strategy about vehicle and edge nodes.By studying the relevant literatures of MEC and considering the characteristics of computing resource distribution in VANET,an offloading strategy for vehicular edge computing is designed.The scheduling strategy in VANET is studied by combining the edge computing nodes which deployed on the roadside and vehicular computing resources.When the vehicular equipment generates a computing task request,the computing task will be offloading into the corresponding computing node,through the analysis of markov decision process(MDP)and the training of the reinforcement learning,which guarantee the alarm performance of long message security applications.