Research On Data Transmission And Caching Strategies In Vehicular Networks

As one of the most promising cross-disciplinary applications involving 5G,transportation and automobile,the Internet of Vehicles has become an important development direction of strategic emerging industries in China.Various Internet of Vehicles applications include road safety,traffic efficiency,autonomous driving,infotainment,etc.These applications put forward new requirements for mobile communications such as low latency,high reliability,large bandwidth,and high mobility.In this thesis,we aim to address three key challenges of Internet of Vehicles:(1)How to guarantee data security which is crucial for safety?(2)With the rapid increase of mobile Internet traffic,how to ensure safety,reliability,and stability of data transmission?(3)How to design efficient caching strategies using limited storage spaces?This thesis mainly focuses to address three issues: First of all,intrusion detection technologies based on machine learning has been gradually applied to the intelligent vehicle networks.However,the actual detection performances of the intrusion detection system implemented on vehicles has not been tested.Secondly,vehicle networking for information interaction generates massive amounts of data.Clustering can improve the efficiency of traffic information collection and distribution,but there are few studies on data transmission optimization within clusters.In addition,repeated downloading of popular contents will increase backhaul loads.Caching technologies can effectively reduce transmission delays and costs by caching popular contents on road side units(RSUs)and vehicles close to the requester.Most of the existing caching schemes consider RSU and base station(BS)caching.Without taking cache capacity on vehicles into consideration.To address the aforementioned issues,this thesis conducts in-depth research on data transmission and caching strategy of Internet of Vehicles through modeling and simulation analysis in different scenarios.The main contributions are as follows:(1)We build and train an intrusion detection system based on simulation data.For scenarios where RSUs are less deployed,we propose a fuzzy logic-based intelligent detection system deployment scheme is proposed,which can effectively identify malicious vehicles based on the set threshold values.(2)Based on complex network theory,we propose a clustering algorithm based on generalized distance,a cluster head selection algorithm based on fuzzy logic,and an optimization model of intra-cluster data transmission.Our work aim to reduce the information transmission load of the entire network for urban traffic data service collection and distribution scenarios.The comprehensive communication scheme can effectively improve the network throughput in clusters and reduce transmission delays.(3)We expand an existing single-layer caching scheme for the Internet of Vehicles,and derive the average content delivery delay and average content delivery cost.We then propose a caching scheme so that the caching capacity at both vehicles and RSUs can be fully utilized to minimize the overall transmission delay and cost.We further propose an alternate dynamic programming search(ADPS)-based algorithm and a low complexity cooperative greedy algorithm to solve the optimization problem in this scheme.In addition to numerical simulation,we develop an infotainment content transmission system on the MK5 platform and implement the prototype model.Results show that the proposed caching strategy can effectively reduce content transmission delays and costs.