As an important vertical application field of 5G technology,the Internet of vehicles has great application value and social benefits.With the increasing popularity of 5G network,the Internet of vehicles will usher in large-scale deployment,which ensures traffic safety through the continuous information interaction between nodes and roadside units.However,in the network,there are some selfish nodes to publish false information for their own interests,or some rogue nodes send a lot of malicious data to destroy the communication network usually.As a result,it is difficult to establish a reliable and stable trust relationship between nodes,and the cooperation efficiency between vehicles is greatly reduced.This paper mainly studies the trust management mechanism of the Internet of vehicles.The main work is as follows:Firstly,a data filtering scheme EDFS(An Effective Data Filtering Scheme)is proposed,and a data centric trust management mechanism is established based on proposed EDFS.The proposed mechanism makes a detailed design of the system composition,traffic flow model and communication mode.EDFS,as a core component,is independently deployed on each vehicle and does not rely on any third-party facilities(such as roadside units).The EDFS scheme uses outlier detection algorithm to eliminate outlier data with large deviation.The clustering algorithm will filter a large number of malicious data according to the data characteristics combined with the context and the clustering changes according to the time series comparison.In the face of Sybil attack,that is,in the presence of a large number of false data,the vehicle can still achieve efficient filtering,so as to establish a trust relationship between nodes.Secondly,a trust management simulation platform based on Veins is designed and implemented to expand the related security components and functions,and the proposed trust management mechanism is simulated and evaluated on this platform.The platform integrates traffic flow loading,key management,encryption and decryption,parameter visualization configuration,multi message monitoring and trust management.The trust management module implements outlier detection algorithm and clustering algorithm.During the simulation,the Sybil attack component is designed combined with false message injection,which verifies the effectiveness of the trust management mechanism based on Veins.Users can configure the platform by themselves,and the modules can be plugged in and out,which greatly makes up for the deficiencies of the IOT simulation platform in safety related applications. |