A Strategy To Build Network Of RSU Based On Real Traffic Data

As the number of vehicles increases continuously,urban traffic congestion is getting worse.Development in information acquisition and processing technology,sensor technology,communication technology,wireless sensor network technology makes it possible for every vehicle to equip a sensor unit and a communication unit.Sensor unit can acquire data like track of a vehicle etc.These data can provide important data basis for traffic control,vehicle navigation,emergency rescue,traffic accidence avoidance and etc.The network of vehicles can transfer these collected data to data centers or other vehicles,it can also transfer orders from data centers to vehicles.The internet of vehicles provides a new way of thinking to improve urban traffic conditions.Due to vehicles’ nature,however,it is impossible to form a stable and usable wireless network using only vehicle-to-vehicle communication.Previous works pointed out that the deployment of road side units(RSUs)can significantly improve the connectivity and performance of the entire vehicle networking.RSUs are connected by wire,and they have far more powerful communication capabilities and storage capacity compared with vehicles.The network of RSU actually becomes the backbone of the internet of vehicles,which plays an important role in internet of vehicles.Currently,there are lots of studies on RSU,such as,vehicle coverage,road segment coverage,area coverage,buffer usage and etc.Most of these works are from the perspective of data collection and acquisition.These studies is suitable for applications to transfer data to data centers,but rarely consider the impact the deployment of RSUs have for the communication between vehicles.Furthermore,these studies focus on the question of placement of RSUs,but ignore the subsequent problems such as wiring,route establishment and cache usage reduction.This thesis’ s contributions to network of RSU in internet of vehicles are stated as follows:1)We study the impact of the deployment of RSU for communication between vehicles,modeled the communication from vehicle to RSU,and then from RSU to vehicles.2)Based on the communication model,we propose a new criterion to evaluate the RSU deployment,and propose several heuristic algorithms to optimal RSU deployment based on this new criterion.3)We point out that the assumption in previous works that RSU’s buffer is infinite is not reasonable.And we are able to reduce RSU cache usage with limited affecting the performance of the network through our partial-RSU-hold algorithm and destination RSU prediction algorithm.4)We provide a robust wiring strategy for RSUs,which is robust because the network of RSU can keep its connectivity even if one of the links in the network is damaged.At the same time,the cast of wiring is minimal.5)We propose protocol for route table generation.This protocol makes the RSUs aware of the topology of the network,and generate route table based on this topology.This protocol must meet the requirements to generate the route with the shortest path and require little calculation.RSUs should be aware when there is a link breakdown,and according to the change in network topology,automatically update the route table,making the RSU network connected.When the breakdown recovers,the RSUs should have a mechanism to restore the route table before the breakdown occurs,so that its route path is the shortest.