The Research On Low-cost And Large-scale Urban Vehicular Network For Data Gathering In Smart Cities

5G technology enables the mobile telecommunication network to access massive Io T devices,laying the underlying network foundation for the smart city of connecting all things.However,the non-profit public welfare nature of smart city applications makes the cost of 5g network data transmission generated by urban data collection limited by the limited government budget.If the idle information resources of mobile intelligent devices widely existing in the city can be used to transmit urban non-real-time IOT data for free,it will greatly improve the ability of urban data collection and promote the rapid development of smart city related applications.The next-generation electric intelligent vehicles have the characteristics that portable intelligent devices(smartphones,smart wearable devices,etc.)lack: high power,plenty of electricity,less equipment weight and volume limitation,and rich resources of computing,communication,storage,and sensor.The large-scale urban vehicle network is more suitable for smart city.The traditional VANET(Vehicular Ad-hoc Network)has the characteristics of dramatic changes in wireless communication environment and highly dynamic network topology.VANET nodes,which are lack system resources,are difficult to communicate and cooperate with and cannot be deployed on a large scale.In order to solve this contradiction,VANET architecture introduces the network infrastructure roadside unit RSU.However,the deployment of additional RSU makes VANET expensive and lacks competitiveness compared with the mature Telecom mobile network.Therefore,building a new low-cost and large-scale urban vehicular ad hoc network independent of roadside units has become the main challenge of collecting urban vehicle information resources for non-real-time data transmission in a smart city.In this dissertation,the essential characteristics of urban vehicles,the urban environment,and the existing urban traffic network are introduced into the top-level design of urban vehicular ad-hoc network architecture to reduce the system complexity.On this basis,a low-cost urban vehicula network architecture with low system overhead,high wireless communication efficiency,and good scalability is constructed.The main work and innovative achievements of this dissertation include:(1)Static Characteristics of Urban VehicleThe static characteristics of urban vehicles are revelated by studying the monitoring data of the private cars in Changsha.It breaks away from the shackles of traditional vehicular network assumption that vehicle nodes are in the moving states.It implies that the parking state of vehicles and the spatial-temporal distribution of vehicle parking areas can be introduced into the vehicular network architecture design.(2)Simplified Model of Urban Vehicular Daily ActivitiesConsidering that the static / moving states of vehicles are unstable and unpredictable in reality,this dissertation proposes a simplified model of urban vehicular daily activities which abstracted from the parking states with longer duration.The model uses the static characteristics of urban vehicles to simplify the number of vehicle states and improve the predictability of vehicle states.The system complexity and the corresponding network system overhead of urban vehicular network is significantly reduced from the source.(3)Low-cost and Large-scale Urban Vehicular Network FrameworkIn this dissertation,a Low-cost and Large-scale Urban Vehicular Network Framework is proposed.It utilizes the change of vehicle node's own motion state and location to switch the role of vehicle node in the network to implement the cooperative interaction of each node.The long-term parking vehicles establish a long-term stable wireless communication connection with Io T equipment and other vehicles,which ensures the efficiency of vehicle data transmission in urban vehicle network and avoids the defects of the short lifetime and unstable link in the traditional vehicular network.Compared with the traditional vehicular network relying on roadside units or the telecom network relying on network infrastructure such as base station,the decentralized urban vehicular network has the advantage of low cost in construction,operation and maintenance.(4)Multi-scale Division Method of Urban Vehicular Network AreaIn order to solve the impact of uneven distribution of urban vehicle parking area,this dissertation proposes a multi-scale regional division method.It makes the urban vehicular network area self-expansion adaptive to the distribution of various vehicles in the real city,and can effectively improve the performance of urban vehicular network.This dissertation studies the scalability of the urban vehicular network.Using real vehicle monitoring data to analyze the performance of the urban vehicular network proposed in this dissertation,and verify that as a self-organizing network without external infrastructure,the network performance improves with the increase of system scale,which can meet the task requirements of collecting urban vehicle resources for transmitting Io T data within the city.