Research Of Rsu Deployment Strategy For VANET

In recent years,with the rapid development of Chinese automobile manufacturing industry and the popularity of private cars,the transportation field has emerged a series of practical problems to be solved,such as road congestion,frequent traffic accidents,low transport efficiency and so on.The study of Vehicular Ad Hoc Network(VANET)provides a feasible path to solve the above problems,which has aroused the concern of more and more researchers.Data dissemination is a basic and essential issue in VANET.This thesis focuses on the RSU deployment problem for information dissemination.This thesis firstly collects the vehicle trajectory captured from real traffic environment.Then we use mean filtering and interpolation to solve the problem of large time granularity and poor data accuracy.According to the specific contents of data distribution,we study the RSU deployment scheme for short message data transmission and for long message data transmission.The short message is small but the real-time requirement is high.In view of the real-time requirement,this thesis models the RSU deployment problem as a maximum vehicle coverage problem of how to find the optimal deployment positions for maximizing the number of vehicles which is covered under the limited number of RSUs.In addition,we introduce the transmission probability model taking into account the rapid vehicle movement,the signal interference,and other factors may lead to communication interruption,and redefine the concept of vehicle coverage.The long message is large but its latency is lenient.Due to the large amounts of data,the distribution of long message data needs to be transmitted for a long time.Therefore,the vehicle can only acquire part of the data by remaining in an RSU communication range.If the vehicle wants to obtain complete message data,it must meet multiple RSUs.Moreover,the RSU type is not fixed,there are many types to choose from,different types of RSU have different communication ranges and deployment costs.Based on the above scenario,this thesis proposes a RSU deployment strategy for long message transmission,which minimizes the costs of RSU deployment while ensuring the successful distribution of long message data.These RSU deployment problems in this thesis belong to the NP-hard problem.Due to the problems complexity and the heavy computation required by the exhaustive method,an evolutionary algorithm is designed to solve them.The algorithm is tested in real-world datasets and compared to a greedy algorithm and the results show that our algorithm is effective and can find better results than the greedy algorithm.