The Study Of Data Dissemination Enabled By Edge Computing In Vehicular Networks

Internet of everything is the trend of developing communication technologies.As one important part of internet of things,internet of vehicles represents a promising approach to deploy smart city and intelligent transportation.Internet of vehicles is changing people's life style,such as enhancing road safety,improving transportation management as well as providing entertainment services.A variety of vehicular applications are emerging,including path navigation,video downloading,gaming and content fetching.In order to meet these applications,data communication and processing are two key technologies.Different from mobile ad hoc networks,vehicular networks have some inherent characteristics.For example,the mobility of vehicles leads to intermittent connectivity,topology changing causes unstable end-to-end link,time-varying channel impacts transmission performance,and obstacles result in non-line-of-sight transmission.Due to fast mobility of vehicles,frequent topology changing,varying channel with time and complicated urban environments,it is a quite challenging task to design an efficient data dissemination scheme.On the other hand,increasing vehicular applications drives the exponential growth of network data,which imposes huge pressure on current vehicular network architectures in terms of delay and bandwidth.Effective data computation is helpful for improving significantly communication quality.Meanwhile,massive data possess a wealth of information regarding transportation,network and users.Data fusion and mining is better for optimizing transmission decisions.All of these questions need strong computation resources.Traditional network architectures fail to deal with these challenges.Therefore,it is crucial to design a new network architecture along with some novel key technologies to realize efficient data communication and processing.In order to achieve this goal,we aim to integrate data dissemination with mobile edge computing to forward and process data efficiently,which is beneficial in improving service of quality.On a basis of vehicular scenarios,vehicular characteristics and application requirements,we focus on the creation of novel network architecture and the design of data dissemination scheme and computing offloading scheme,with the aim to guarantee the efficiency of data communication and computing.The detailed works are introduced as follows:(1)We propose a novel vehicular communication architecture enabled by mobile edge computing to be the guidance of designing data computation and data communication schemes,where mobile edge computing is deployed to guarantee the quality of data transmission.First,software defined network is introduced into this architecture for dealing with the limitations of existing data dissemination schemes,the global view of which is beneficial for optimizing routing policy and resource management.Then,mobile edge computing is deployed in the network edge to provide proximity service for users.Special,the deployment of mobile edge computing combines inter-layer cooperation and cross-layer cooperation to maximize resource utilization,and designs computing offloading scheme with service caching as well as leverages artificial intelligence to achieve the goal of resource allocation.(2)We propose a task offloading scheme by exploiting mobility in vehicular edge computing(VEC)to guarantee data communication by providing necessary computation resources,which makes the best use of available computation resources in vehicles.The novelty of this scheme lies in not only dealing with the difficulty caused by the mobility of vehicles but also capturing the potential opportunity when implementing tasks.In addition to all onehop vehicles of task vehicle which has tasks to implement,certain two-hop vehicles which meet the given requirements in terms of mobility pattern and computation capacity are also be leveraged to carry out the assigned tasks by the task vehicle.The offloading policy of the task vehicle is made to optimize the utility system of task vehicle.A semidefinite relaxation approach with an adaptive adjust procedure is proposed to solve the formulated optimization problem and obtain the corresponding offloading decisions by taking into the effect of mobility of one-hop and two-hop service vehicles.(3)We propose a connectivity-based and delay-guaranteed geographic routing protocol enabled by edge nodes in vehicular networks to ensure effective data communication.In view of the large size of a modern city,the road map is divided into a series of Grid Zones(GZs).The forwarding of packets is then determined among GZs instead of the whole map to reduce the computation complexity.The best path within each GZ is chosen according to the road weight evaluation algorithm,by which each road segment is assigned with a weight via comprehensively exploiting the real-time link property when connected or historic traffic information when disconnected,where the connectivity model and delay model are built as the routing metric,respectively.In this way,the risk of local maximum and data congestion can be greatly alleviated by taking the transmission delay and traffic conditions into account.In addition,the built backbone link can further avoid the MAC contentions and guarantee the quality of data transmission when forwarding packets,which is composed of a series of edge nodes at intersections and within road segments.