Bayesian Based Security Detection Method for Vehicle CAN Bus Network

Abstract

AbstractThe existing on-board can bus network security detection methods have low accuracy in detecting the number of abnormal messages and long detection time. Therefore, combined with Bayesian algorithm, this paper designs a on-board can bus network security detection method. For data preprocessing, establish a classification model, classify network attacks, calculate the information entropy of CAN bus, establish a Bayesian model, learn Bayesian network, continue to reason, and realize the security detection of vehicle can bus network based on Bayesian. The experimental results show that the proposed security detection method has high detection accuracy and short detection time, and can find internal security vulnerabilities in time, providing technical support for the safe operation of vehicles.KeywordsBayesCAN busNetwork securityClearClassificationJudgment