Research On Multi Virtual Vehicle Cooperative Service In The Internet Of Vehicles

With the new communication technology and infrastructure,the Internet of Vehicles realizes information exchange and behavior coordination between people and vehicles,which is an important part of intelligent driving technology.In intelligent driving services,machine intelligence is not enough to complete all complex decisions independently.In urban transportation,there will be a long-term mixture of human behavior intelligence and machine control intelli-gence.As a user and provider of intelligent driving services,each vehicle and car synergy requires a combination of centralized and distributed hybridization in a wide range of time and space.In the past,there are three aspects of the lack of human-vehicle synergy behavior,lack of hybrid interaction mode,and lack of behavioral metrics of human-car synergy,which cannot provide support for efficient intelligent driving services.To this end,this paper proposes a hybrid multi-virtual car collaboration framework.The framework first uses the virtual car in the information space to characterize the driver's behavioral intelligence in the human-car synergy,and then abstracts the distributed or centralized phys-ical communication between the car-networking nodes into logical communi-cation in the information space.On this basis,this paper makes an in-depth analysis of intelligent driving services,focusing on the local collaborative ser-vices,global collaborative services and behavioral trust metrics,and has achieved the following results:1)Aiming at the local collaborative service of multiple virtual vehicles in the car network,a service method based on information dissemination and par-ticipation incentive is proposed to complete the entire service process of obj ect discovery and collaborative decision making between virtual vehicles at the edge of the network.The method firstly discovers the collaborative obj ects ac-cording to the customized services,and the service information arrives through the self-organized communication to the individual virtual vehicles that do not have direct communication with them,thereby expanding the scope of the co-operative group.After forming a range of potential service providers,all indi-viduals make their own decisions and complete service delivery.Experiments based on traffic simulation software and communication process simulation show that the proposed method can make efficient information dissemination and group adaptive decision based on the customized content,so as to support the local collaborative awareness service of multiple virtual vehicles.2)In the context of multi-virtual vehicles global collaborative services,a service method based on value iteration and combinatorial optimization is pro-posed to rationally allocate road resources through the combination of autono-mous decision-making on the edge of virtual vehicles and the coordination of cloud-based groups.This method uses a value iterative network to make the edge of the virtual car learn human driving patterns and anthropomorphic path planning,and then through the cloud coordinating node for multiple virtual car path candidate set further combined optimization.Experiments based on real city maps and taxi trajectory data show that the proposed method can effec-tively perform coordinated navigation and diversion in the overall city,thereby improving the traffic efficiency of urban traffic.3)In the aspect of trustworthy measurement of multi-virtual vehicles be-havior,a trusted measurement method based on proximity propagation and cen-tral arbitration was proposed.Through the combination of edge dynamic as-sessment and historical assessment of the cloud,virtual vehicles with bad be-haviors were found to meet the needs of collaborative services for car network-ing.Trusted requirements.This method enables the virtual vehicle to dynami-cally recognize the local environment and individuals by means of communi-cation or observation,and generate vehicle network environment data.After a plurality of virtual vehicles share intelligence,they are jointly calculated to ob-tain dynamic trust,and then combined with the historical trust information of the central arbitration.Carry out accurate identification of individual virtual vehicles.Based on vehicle trajectory and behavior simulation,the proposed method can fully and accurately measure the behavior of virtual vehicles and provide effective support for virtual vehicle coordination.